Riars AS DISASTERS: 
AN EXPLORATORY CASE STUDY OF SELECTED ASPECTS 
OF THE CIVIL DISTURBANCE IN WASH INGTON, D. C., APRIL, 1968 
by 
Richard Guy Sedlack 
,,, 
Dissertation submitted to the Faculty of the Graduate School 
of the University of Maryland in partial fulfillment 
of the requirements for the degree of 
Doctor of Philosophy 
1973 
i r ,~;, 
APPJWV AL SHEET 
Title oi' Thesis: Riots as Disasters: An Exploratory Case Study of 
Selected ABpects of the Civil Disturbance in 
Washineton, fJ. C., April, 1968 
Name ol' Candidate: Richard Guy Sedlack 
Doctor oi Philoso~hy, 1973 
, .-, ' 
Thesis and Abstract Approved: I /l:~.~rf(t:/ .~ ;U ~ l<lrj 
D:t. Robert W. J an7 s 
Professor 1 / 
Department oi' Sociology 
Date Approved, (j)vc I0  /r  12-
ABSTRACT 
Title of Thesis: RI ITTS AS DISASTERS: AN IXPLORATORY CASE STUDY 
OF SELECTED ASPECTS OF THE CIVIL DISTURBANCE 
IN WASHINGTON, D. C., APRIL, 1%8 
Richard Guy Sedlack, Doctor of Philosophy, 1973 
Thesis directed by: Dr. Robert w. Janes 
Professor, Department of Sociology 
Following some of the more recent sociological literature 
which has been critical of research into riots, the topic of 
this thesis addresses itself to a hitherto neglected aspect of 
riots. It is an initial exploratory effort into the ecologica l 
dimensions of official statistics, utilizing the relevant 
temporal and spatial conceptualizations suggested by the socio-
logical disaster literature. 
The data sources were the offense and arrest r ecords of the 
District of Columbia Metropolitan Police Department and the fir e 
data on the Daily Conununication Log of the District of Columbia 
Fire Department. The offense and fire data were treated as 
partial indicators of the situation reported to the police with 
the arrest data as partial indicators of the response made by 
the police to the riot. The data were conceptually organized 
along three dimensions. First, the type of criminal violation 
was classified into six general categories: crimes against 
persons, crimes against property, traffic violations, crimes 
without victims, crimes related to fires, and miscellaneous 
crimes. For a more detailed analysis, the total crimes falling 
into any one of these general categories were subclassified into 
more detailed s ubcategories within each general category. Second, 
the spatial dimension was trichotomized into three locational 
specifications: the riot areas of major destruction, the corridor 
areas of sporadic destruction, and the non-riot areas of minimal 
or no riot destruction. Third, the temporal dimension was 
dichotomized into the total riot period of organized response and a 
representative normal time period, so that the latter could serve 
as a benchmark against which to compare the former. 
Two specific ~uestions were posited: what degree of difference 
existed between the defined riot period and the representative 
normal time period in terms of crimes and spatial location as 
reflected by the official statistics and what kinds of differences 
were evident. Three specific hypotheses were evaluated: (1) the 
offense and fire data hypothesis which suggested that the degree 
of association between the offenses reported and the selected 
riot-normal time period varies directly with the degree of con-
centrated riot damage, (2) the arrest data hypothesis which 
suggested that the degree of ass ociation between the police's 
response and the selected riot-normal time period varies directly 
with the degree of concentrated riot damage, and (J) the comparative 
hypothesis which suggested that the degree of association between 
the police's response and the selected riot-normal time period is 
less than the degree of association between the offenses reported 
and the selected riot-normal time period . 
Utilizing the lambda proportionate reduction in error statistic, 
the data were inconclusive relative to the first hypothesis and 
generally failed to support the second and third hypotheses, 
although the magnitude of the data indicated that there were some 
differences. The nature of the differences indicated that the 
incidence of fires and burglary violations increased substantially, 
while larceny, false fire alarm reports, and the degree of 
violence in crimes against persons decreased in the reported 
offenses during the riot. The police response was dominated 
by arrests for disorderly conduct and curfew violations with 
burglary arrests ranking second. While there were decreases 
in larceny and traffic arrests, the latter were still substantially 
represented during the riot and no meaningful numbers of arson 
arrests were made. Further, it was concluded that substantial 
numbers of offenses reported and arrests made occurred in the 
non-riot areas. 
It was concluded that the disaster literature provided 
relevant conceptualizations for the analysis of the spatial and 
temporal dimensions of riots, that further analysis of these 
dimensions is warranted , and that other dimensions of the disaster 
approach appear to be useful when applied to riots. 
PREFACE 
Introduction 
Presented below is a brief and selected descriptive account of 
the riot in tbe District of Columbia, which followed the assassination 
of Martin Luther King, Jr. in April of 1%8. The pur pose of this 
chapter i s firs t to assist the reader in his re coll ection of t hese 
events and second to partially elucidate the variety of behaviors and 
act i ons taken by the people in the Washi ngton metropoli tan area . It 
is hoped that the reader will develop an appreciation for t he magni -
tude, complexity, and variability of the ac t ions which accompany a riot. 
The first section is a chronology of events so that t he reader 
will again appreciate the flow of the riot period itself. Mos t of the 
events listed here deal with the principal and most visible actors eE-
gaged in the riot situation itself: the firemen, the police, the 
rioters, and the military personnel. In other words, this section 
focuses on those who started and perpetuated the riot and those whos e 
responsibility was to attempt to control the effects of the riot , The 
second section is concerned with the less dramatic response made by 
public agencies, private organizations, religious groups, and involved 
public citizens--those people who attempted to render aid to the vic-
tims of the riot. 
The Chronology 
At approximately 6:20 P.M. on April 4th, 1968, Martin Luther King, 
Jr ., was shot in Memphis, Tennessee. At 7:JO P.M., WTTG-TV, a local 
i i 
iii 
Washington television station, announced the shooting and reported 
that King was in serious condition. 1 Dr. King was subsequently pro-
nounced dead at 8:06 P.M. 2 In Washington, the news brought a crowd 
of several hundred persons to 14th and U Streets , N.W.--an intersection 
which is one of the major public trans portation hubs in the District. 
At 9:25 P. M., the first window breaking at the Peoples Drug Store next 
to the Southern Christian Leadership Conference offices at 14th and U 
Streets, N. W. was heard, and the Washington r iot began. 3 
At 12:JO A.M., April 5th, the first fullscale fires were set on 
14th Street, although the first fire call was received about an hour 
and a half earlier. Fire Chief Henry Galotta instituted Plan F, 
which split the existing fire companies in two, thereby doubling the 
response capability of the District of Columbia Fire Department , at 
about 11 :00 P.M., April 4th. 4 At 1:20 A.M., April 5th, the police 
asked and were granted permission to use tear gas on 14th Street. 5 By 
J :OO A.M., the last major confron1ati on on 14th Street was over, and 
200 stores had windows broken with looting occurring in 150 of them. 
Seven fires had occurred and 200 people had been arrested with thirty 
people injured, including fiv e policemen and one fireman. 6 Dawn saw 
many policemen on 14th Street, N. w., with this street then quiet. 
The police were to be kept on overlapping twelve-hour shifts with 
the special Civil Disturbance Unit to be activated in force late Friday 
afternoon. 7 
A riot commission study of twenty-four disorders had 
indicated that major violence in all those cases had oc-
curred during the night hours, and that, as a general rule, 
violence at night was followed by lulls during the daytime. 
However, there had been a few notable exceptions to this 
pattern. Washington was to be another. 8 
:v 
Tne d;:iylight hours during the morning of 1\pril 5th indicatE::d more thi::m 
the normal nwnber of idlers on the street. Tbe District officfals felt 
tbc.1t if the sc.hool system could operate norrm-.illy, t.he chances of a 
peaceful Friday would be enhanced. Several factors worked against keep-
ing the city's 150,000 pupils in classes. The Washington Teachers Union 
re~uested the closing of the schools so that teachers could attend mem-
orial services for Dr. King. Concerned pi::lrents wanted their children 
rele,.rned, and the Student Non-Violent Coordimi ting Conunittee rey_uested 
closure in memory of Dr. King. Finally, the pupils themselves mani-
fested a more than usual restlessness. Many students walked out of 
their classes during the morning hours or did not return after the 
luncheon peri. od  ? 9 At l:JO P.M., the school officials, unable to retain 
the students , dismissed the cJchools on a staggered basis. 10 By 2:00 P.M. 
reports of intensive burning and looting were being received from 
various parts of the District . 
Some policemen tried to persuade looters to go home, but the police-
men, as a whole, were badly outI1umbered and virtually powerless to in-
stitute any meaningful response to the situation about them. Looters 
just ran away from the police , knowing that there wa s little the police 
11 
could do, 
A well-dressed, matronly-looking woman was one of several 
who assured newspaperman Maynard that they were not looting. 
"Look, 11 she said . "I come here to shop every Friday 
night. Where else am I to find food? All the stores are 
closed or looted. It's not that the food is free. It's 
just that there's not going to be much food around here for 
a long time. 11 12 
After they departed, with loaded arms, a man, who had been 
standing on the sidewalk, watching the spectacle, removed 
a gasoline-filled cola bottle from under his coat, lighted 
it, and tossed it through the shattered doorway. It was a 
perfect hit . ?.. 
V 
Witnesses said the man had done the same thing at two 
other stores down the street. They were struck by th~ fact 
that he showed no interest in looting, only burning. 13 
These quotations should give the reader some feeling for the variety of 
behaviors manifested by the rioters themselves. What is inter esting 
here is that the r iot participants themselves did not s eem to represent 
any one consistent pattern of riot behavior. Rather different people 
performed different kinds of actions during the course of the riot. 
At 1:00 P.M., the first fire was reported on 7th Street, N. W., 
and soon thereafter some of the civilian employees at police head-
quarters and the District Building began to go home. After some mini-
mal harassment from small groups of blacks, the Hecht Company and the 
Woodward and Lothrop Company--Washington's two largest department 
stores--decided to close, which led quickly to many smaller businesses 
closing for the day. Large numbers of customers and employees "flooded" 
the streets, and the main shopping arteries ?were soon clogged with 
automobiles which could hardly move. 
All over town, office workers more or less simultane-
ously reached for their telephones. The heavy volume of 
calls, coming at ?the usually busy time before the start of 
a weekend, swamped the city's exchanges. 
Downtown office workers--both federal employees and 
those on private payrolls--then did what came naturally. 
They began piling into their automobiles to head for home. 
The trek, as it picked up momentum, created one of the 
most massive traffic jams in the city's history. 
Deputy Mayor Fletcher asked the Civil Service Com-
mission to stagger the release of the many tens of thousands 
of typists, clerks, and officials who wanted to go home, 
but it was too late. Their flight could not be slowed. 
Just as the schools had failed earlier in their effort to 
keep pupils in class, the government now was unable to pre-
vent its work force from fleeing the clrea. The general dis-
order, to which the released school children were already 
contributing, spiralect.14 
The District of Columbia Public Works Department reported that the 
vi 
evening rush hour traffic started one hour earlier than norma l, before 
the traffic signals could be switched into their rush hour patterns, 
thereby exacerbating the traffic congesti.o n . 15 
At 12:55 P.M., the situation on the streets was rapidly deterior-
ating and the police began calling for the District of Columbia 
National Guard. 16 At 2:48 P.M., the National Guard was placed on 
standby for weekend duty, and the call ordering them to report to their 
armories was issued at J:29 P.M. 17 At the same time, the Cherry Blos-
som festivities--Washington's first major tourist attraction of the 
spring--were cancelled, as was all leave for the District of Columbia 
policemen and firemen. At 4:40 P.M., the first federal troops began 
18 
moving into t he District. At 5:20 P.M., Arlington and Prince Georges 
Counties sent fire companies into the District to aid the District of 
Columbia Fire Department. 19 
Under a plan that had been worked out in advance with 
suburban fire departments, the Washington firemen were also 
aided by sixty engine companies from the Maryland and Vir-
ginia suburbs. Vol unteer fire departments came from farther 
away, one unit driving down from Lebanon, Pennsylvania,20 
This was the first major deployment of non-District f ire fighting 
equipment since 1927. It was reported that some inner-city residents 
helped the firemen by holdi:ig the fi re hoses, bringing them chairs to 
sit in, coffee, and sandwiches as they labored throughout the night to 
21 
extinguish the fires. While there reports of harassment of firemen, 
the Fire Department reported only nine instances during the riot when 
fire equipment was delayed through interference by crowds of people. 
Further, while rocks were sometimes thrown, there were no instances of 
sniping at the firemen, and the total damage to fire equipment was 
.$ 1. 0 , 000 --d e f. ?i ne dby  o ff i' ci.a  1 s as mi?n i?m a 1 ? 22 Deputy Chief Huntington 
vii 
summari zed. his feelings toward the burning saying, "In my 36 years, it's 
never been close to this, and I hope that it never will be again. 1123 
At 8:58 P.M., it was reported that th? White House and the Capitol 
were surrounded by mac i. h ne guns . 24 By midnight, 6600 troDps were pa -
t rolling the streets of the major riot areas and enforcing a curfew. 
Thirty-six hours after the initial entrance of f ederal troops into the 
District, there were a total of 13,600 federal soldiers and District 
of Columbia National Guardsmen on duty in Washington. 25 
The appearance of military personnel sharply curtailed the rioting 
in the major shopping areas. However, the looting and burning continued 
in some of the areas unprotected by the troops. 26 While the military 
did not have the authority to make arrests, they were empowered to de-
tain persons until the police arrived, and the latter formally charged 
those who were apprehended . 27 With the appearance of the military, a 
rigid policy of enforcement of the curfew restrictions was begun. At 
1:20 A.M., Saturday, April 6th, the District Court began processing 
1200 arrest cases, and the police cell blocks were reported full at 
2:51 A,M. 28 At 6:14 A.M., the Office of Civil Defense monitored a 
telephone call that Hecht's, Woodward and Lothrop, Jellefs, Raleighs, 
D. J. Kaufman, Kanns, Lansburghs, and Garfinkles would not open for 
business . 29 At 9:26 A.M., the District of Columbia Transit Company's 
buses were reported running but not servicing the bur ned-out area s, 
and at 9:50 A.M., all meetings and movies in Washington were cancelled.JO 
And so it went on Saturday with many business establishments 
closed, others curtailing their services, troops pouring into the 
District, the peak number of arrests made, and the fire and police de-
partments working to their capacities. 
viii 
Sunday, April 7th, was re lcl tively calm. Announcements wer e ma de 
that District stores were to be open on Monday unt i l 4:00 P.M. , t he 
schools would close at l:JO P.M., and District and Federa l empl oyees 
would work as usual with the business day ending one and one-half hour s 
earlier than norma1. 31 Many of these closing times were s e l ected to 
conform to the curfew restrictions placed on the area. Monday, April 
8th, continued calm, and the schools and many businesses were closed 
Tuesday, April 9th, in observance of Dr. King's funeral. At 10:20 A.M . , 
the Office of Civil Defense reported that police and fire calls were 
near normal. 32 District of Columbia Transit announced that their regu-
lar schedules would be maintained on Wednesday, April 10th. Finally, 
at 5:00 P.M., April 12th, after several days of calm, the Office of 
Civil Defense terminated its monitoring of the communication network. 
During the course of the riot twelve people had lost their lives 
in some riot- connected way. Two whites were reportedly beaten to death 
by blacks, although only one such death could be definitely linked to 
the riot. Two blacks were shot and killed by the police. One fifteen 
year old black youth was accidently shot when he da s hed from a looted 
store and bumped into the service revolver held by a policeman who had 
apprehended and was covering a third party . 33 A twenty year old janitor 
was shot and killed when he turned around quickly with a "shiny object" 
in his hand. 34 Eight persons died as a result of looting activities. 
Seven people were trapped in buildings which had been fired with the 
remaining person being killed when the remnants of a plate glass window 
crashed down upon him as he was climbing through the shattered window. 
Some of the more interesting general statistics reported in the 
Washington Civil Disorder Survey35 relative to the privately owned real 
ix 
properties and business establishments36 which were damaged during t he 
riot period are briefly mentioned below. A total of 1199 real proper-
ties and 1352 business establishments sustained damage during the riot 
with the total estimated value loss in dollars placed at 57.6 mil-
lion. 37 These businesses employed prior to the riot a total of 14,593 
full-time employees, a figure which decreased by 4916 to a post-riot 
figure of %77. 38 The riot was acccmpanied by a net loss of 2116 
people in residential or housing units with a net loss of such units 
placed at 403, 39 Of all the privately owned properties damaged during 
the riot, glass breakage was experienced by 89 per cent of the real 
properties and 94 per cent of the business establishments, and fir e 
damage occurred in 4J per cent of the real properties and 37 per cent 
of the business establishments. Theft of merchandise occurred in 85 
per cent of the business establishments which experienced riot damage. 
Only eight public buildings were damaged during the riot with all of 
them being outside the major areas of riot activity and with the value 
of the damage b ei? ng m?i ni?m a 1 ? 40 Wh?i  1 e the do 11a r va 1 ue of pri. vately 
owned real properties which incurred damage or loss was in the millions, 
this value was but 0.4 per cent of the estimated market value of all 
taxable properties in the District of Columbia on July 1, 1968. 41 Study 
of the insurance payments made by the American Insurance Association 
and other sources of information reveals that the private owners of 
damaged property and the insuring companies shared about equally tne 
value of the material damaged . 42 Among the reasons given for the ra-
ther large proportion of expense shared by the private owners were 
those owners with the wr ong type of coverage or no coverage at all, 
those who were under-insured, those who had deductibles or coinsurance 
Y.. 
pena l t i es , c.rnd some pol ic i es which paid actmil cash va l ue of damaged 
or lost ma terials ruther t han r epla cement costs . 43 Asked ci bout their 
fu t ure bur.:dne s s pl ans , I_/) per cent or 66 8 own er s of damaged bu:siness 
establis hments s aid they would not return t o t he ir pr e-riot l oca tior,s . 
Of these, t he owners of 245 b 1Jsinesses sa id t hey planned t o relor:ate 
outs ide their former neighb orhoods , while 423 own er s said they pla nned 
t o e ither clrn,e or s ell their businesses. 44 
Robert N. Gold, Assis t ant Di rector for Social and Economic Re -
s earch of the National Capital F'la nning Commis sion, pr esented three 
ba s ic characteristics of the targets oft.he ri ot in his tes t imony be -
fore a Senate subcommittee . 
1\bout 95 percent [ sicJ  of the busines s establi shments wer e 
in retail trade and services, This means that they ha d 
merchandise. 
The second point i s that a very high proportion, and 
I believe the figure is close to 90 percent [sic] , s erved 
primarily neig~1borhoods. These were not unfamiliar es-
tablishments in thos e neighborhoods , 
The third feature which stands out is that Lhe types 
of damage or loss that occurred with highest frequency wa s 
[ sic] glass breakage and theft of merchandise . 45 
Fur?ther, Edward C. Hromanik, chief of a special task force of the 
National Capital Planning Corrunission assigned to prE:pare urban renewal 
pla ns for the civil disturbance areas reported that roughly 50, 25, 
and 15 per cent of the material damaged or lost in the District of 
Columbia was suffered on 14th Street, N. w. , 7th Street, N. W., and H 
Stre et, N. E., respec t i.v e 1 y . L,.6 
The Response 
??? Members of both races--2 ,000 in all, including many 
white suburba nites-- volunteered to help. By Saturday, they 
had operating thirty- eight collection points at suburban 
churc::hes and forty- Lwo distr ibution centers at inner- city 
churches. L,.7 
xi 
As of 12:JO P.M., Thursday, April 11th, the Office of Civil Defense 
reported that 1,436,750 pounds of food had been distributed and 79,820 
people had been provided with three mea ls a day for a six-day period. 48 
Some of the military assisted in the distribution of food, while some 
corrunercial stores (such as the Giant and Safeway supermarket chains 
and the Crusty Pie Bakery), the United States Department of Agriculture, 
and private individuals donated food, labor, and equipment for the mass 
distribution of food supplies. 49 The Salvation Army itself dispensed 
over 250,000 cups of coffee, a thousand gallons of soup, 51,000 sand-
wiches, 775 dozen doughnuts, 13,000 usable items of clothing, 800 furni-
ture items, a limited amount of rent money, and provided some emergency 
shelter. 50 The District of Columbia Chapter of the American National 
Red Cross operated a blood bank (although the need was below normal), 
distributed 1600 disaster-type comfort kits, operated a canteen for one 
week, provided motor service for medical personnel, and assisted in the 
coordination of volunteers. 51 Several agencies set up information cen-
ters, such as the Interracial Council for Business Opportunity of Great-
er was h i. ngt  on, 
52 or rumor con t ro1  cen t ers to qu'i et rumors and to di' rec t 
people to the proper places for donations, such as the Montgomery County 
co.rnmission on Human Relations. 5 3 The District of Columbia Citizen In-
formation Telephone Answering Service processed approximately 5300 tele-
phone calls between 5:JO P.M., April 6th, and J:00 P.M., April 9th. 54 
This is not to say that all contingencies were met without con-
fusion and inefficiency. While the District of Columbia Department of 
Public Welfare had developed a plan to feed and house masses of people 
in the event of a disorder, 
It was impossible to put the Plan into effect on Friday, 
April 5, 1968, when the disturbance began in Washington. 
xi i 
Communications broke down so that contacting individuals 
to activate the Disorder Plan was impossible.5 5 
On Monday, April 8th, the Office of Civil Defense monitored a call at 
10:lJ P.M. that some food distribution centers had exhausted their 
supp 1i. es . 56 At 1:40 A.M., April 9th, the Department of Public Welfare 
needed personnel to assist in handling food at the District of Columbia 
Village Warehouse . 57 Again, at 12:,35 P.M., it was reported that two 
loading docks and twelve trucks were standing empty as thirty-eight 
people assigned to the District of Columbia Village Warehouse were all 
eating at the same tirne--the essential problem being that of the lack 
58 
of an expeditor. The Chesapeake and Potomac Telephone Company re-
ported that their unofficial data indicated a sharp increase in tele-
phone calls early Friday afternoon, April 5th, during which the number 
of calls reached its businest point in the history of the company. 
There were some breakdowns in equipment, and temporary overloading 
caused delays in getting a dial tone. 59 
Prior to the April riot, s~ne individuals involved in the District 
of Columbia judicial Sjstem recognized the need for judicial contingency 
plans in the event of a massive civil a?sorder. Dr. King's proposed 
Poor People's March, which wa s to terminate in Washington, was viewed 
as a situation which might require an extraordinary effort on the part 
of the court system. Keven P. Charles of the Young Lawyers Section of 
the Bar Association of the District of Columbia wrote to Joseph F. Hen-
nessey, the Chairman of the Committee: 
Again each of the organizations involved in this march 
feels it is doing its job and making its plans. Yet we are 
still like the Hindues driving the elephant. SCLC (the 
Southern Christian Leadership Conference] hasn't the foggiest 
idea what the court system is planning, the courts don't have 
any idea of the extent of the protest, the Justice Department 
xiii 
has no idea of the plans of either the court of [ sic J 
SCLC, and doubtless the same can be said f or the legal 
staffs of the District Government, the Inlerior Department, 
and HEW [ Health, Education, and Welfare]. O 
Qui?te simply, plans were being formulated but there was no integration 
or coordination of the various organizations and agencies involved. 
When the riot began, basic policies had not been worked out rela-
tive to apprehending or handling large numbers of arrestees. Those in-
volved in the drafting of the curfew regulation decided aga inst listing 
excepti. ons, encouragi. ng po 1 i? ce o ff.i cers t o exerci. se t h e1. r own J. udgement.6 1 
When the curfew was placed on the District, there was no consensus re-
garding the penalty for conviction. On Monday, April 8th, it was de-
cided that a curfew violator would be allowed to post and forfeit a 
$25 collateral. But the courts pursued an overnight detention policy, 
probably because of the rumor that curfew violators, who had been re-
leased by the court, were reported to have been seen on the streets re-
turning to the riot areas the same evening as their arrest. Hence, for 
all practical purposes, the penalty for a curfew violation was $25 and 
a ni? ht i?n ? 
62 
g Ja?1i.  
The total impression of the investigators is of an 
absence of guidelines and reliance within the Department 
on the individual discretion of the officers. 
? ? ? ? ? ? ? ? ? ? ? ? 9 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 
The interviewsof curfew arrestees tended generally to 
confirm an impression of lack of uniformity in curfew en-
f.o r.c e.m ent. . . . . . . . . . . . . . . . . . . . . " . . . . . 
. . ? Many defendants spent more than one night, or even 
24 hours, in custody before they appeared before a judge . 63 
Most of the officers interviewed after the riot reported that 
they had received no specific instructions regarding looters. Some 
officers arrested every looter; others only those looters with sub-
stantial amounts of merchandise. Following the suggestion of the 
xiv 
United States Attm?ney's Off ice , some poli'.!emen charged all looters 
with second degree burglary, as the former's position was to seek the 
mos t s t r ingent charge which t he typical looting violation will normally 
s upport. I t was envisioned tha t when t he r i ot s had been concluded, 
each sec ond degree burglary case would be r eviewed for a possible re-
duction to a less er charge--us ually a mis demeanor--should the lack of 
evidence suggest difficulty in attaini ng a c onvicti on on the former 
charge . 64 Since s econd degree burgl ary is a f elony char ge, there was 
a s ubstantial increas e i n paperwork required by t he court of t he ar-
resting officer. Some police off icers fe lt that i t wa s more important 
for them to be on the streets rather than filing formal f elony charges 
so they arrested looters for curfew violations, when the la t t er were in 
force, or disorderly conduct, when they were not. 65 Fina l l y, some of-
ficers only arrested looters when they were apprehended insi de business 
establishments. 
But wllat of the court itself? The District of Columbia Court of 
General Sessions--the primary judici al organization functioning d~ing 
the riot--had no de t ailed emergency plan for processing arrestees . It 
wa s placed on a twent y-four hour operating basis, and the only deci-
sion made during the riot was the condition of release of tlle char ged 
66 
offender. 
The Bail Reform Act provides that, instead of money 
bond, the defendant shall 11 ??? be ordered released 
pending trial on his personal recognizance or upon the 
execution of an unsecured appearance bond in an amount 
specified by the judicial officer, unless the officer de-
termines, in the exercise of his discretion, that such a 
release will not r easonably assure the appearance of the 
person as required. 1167 
On friday afternoon, April 5th, most defendants were releas ed on per-
sonal bond. However, when the rumor t ba t releas ed defendants were 
xv 
returning to Lhe rioting areas reached the court, Chief Judge Greene 
called a meeti ng of all the judges hearing riot casgs an~ suggested a 
$1.000 bond in all looting cases unless a responsible person in the 
court could vouch for the accused. The ana.lysis of statistics, trans-
cripts, and interviews with the judges themselves indicated that there 
was little uniformity in practice from the bench.
68 
? ? ? The treatment of defendants varied great l y from 
judge to judge, from day to day, and between day and 
night.69 
Further, the normal initial charging process broke down because 
officers, being needed on the s treets, were not in the court with the 
persons they had arrested. 70 The District Court's criminal calendar 
was seriously clogged prior to the disorder , a fact which did not allow 
a maximal processing of riot period defendants. 71 Finally, simplified 
arrest forms, which a previous demonstration had shown were needed, had 
been designed but not yet printed. 72 In sum, the emergency justice 
system during the riot lacked forethought and planning, detailed in-
structions and guidelines were absent, and justice was differentially 
applied by both the police and the court. 
Summary 
It i.s hoped from the foregoing that the reader will come t o see 
the complexity and variability in the events which surround a riot. 
After perusing the literature on riots, the reader will also realize 
that many of the dimensions of riots suggested above have not been sub-
jected to systematic analysis by sociologists. While the topic select-
ed for ana.lysis in this thesis can in no way be construed to embrace 
the totality of the riot in Washington, it does present an initial, 
xvi 
exploratory effort t.oward a dimension of riots which has to date re-
ceived little attention by sociologists. 
1-'rimary gratitulie is offered to my advisor, Dr . Robert W. Janes, 
who was a constant source of ideas and encouragement. Further, I wish 
to thank Dr. Robert K. Hirzel, Dr, Nils Bateman, Dr. W, T. Simons, Dr. 
Joseph J. Lengermann, and Dr. Walter E. Schlaretzki for their careful 
readings, suggestions, and constructive criticisms of the manuscript. 
I am i ndebted to Mr. James Flynn of the District of Columbia Fire 
Department, Dr, Gary A, Massel of the Institute for Defense Analysis, 
Capt. Miller, Sgt. Fralin, and Mrs. Taylor of the District of Columbia 
Metropolitan Police Department, and Mr . Etheridge of the Office of 
Crime 1\nalysis, who gave of their time and who made available to this 
researcher the data sources used herein. 
Finally, my dee pest appreciation is offered to my wife, Marilyn, 
whose interminable putience and understanding were major contributions 
to the completion of this work. 
--
TABLE OF CONTENTS 
Chapter Page 
PI{EF ACE , ??????? e ????? t't ????? e ????????? e ???? e ?? 1' ??????????? " ?? fl ? ii 
LIS'r OF TABLES ..?. ? ? ? ? ? ? ? ? ? " ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? " ? ? ? ? ? ? ? ? ? ? xxi 
LIST OF FIGURES . e ? ? ? ? ? ? " ? .. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?? ? ??????? " ????????? XXV 
LIST OF Mi\PS . " ?? ? ? ? .. ? ? ? ? ? ? ? ? ? ? ? * ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?? ? ... ... ? ? ? ? xxvi 
I. STATEMENT OF THE PROBLEM ??e???????????"?????????11ee- .. 111? 1 
A. Introduction ? ?????????????????????? ,, ????????? It .... 1 
B. The Major Concern of the Sociological Literature 
on Riots ............... "" ....................... . 1 
c. Specific Problematic of this study 2 
1. Assumptions ,. .................... 111 .............. . 3 
2. Questions ?????????111??????t1?1t??t11o??11teee"?????? 4 
D. SllIIUilary? " .............. " .. e ???? " ...... e ?? ,, ???? ,. .... . 5 
II. REVIEW OF THE LITERATURE 9 .................... 9 ........... 8. 7 
Introduction " ?? fl ???? "' ??? 7 
B. Major Current Conceptualizations in the Study 
of Riots .. " ft ??????????????????? "' 7 
c. Recent Critical Assessments of the Previous 
Riots Literature e?e?e?e???e?e?""????????,.??????&? 8 
D. The Disaster Literature .............................. "' lJ 
1. The Temporal Dimension .. ................. . fl "' .. lJ 
The Spatial Dimension """?????"'????????????&? 18 
E. Summary and Specific Hypotheses ??e???????e??????e 20 
xvii 
xviii 
Chapter Page 
III. METHODOLOGICAL PROCEDURES ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? , 24 
A. Introduction ............... .... .,.................... 24 
B. The Data .... " ? ? ? ? ? ? ? ? ? ? ? ? ? " ? .. ? ? ? ? ? ? ? ? ? ? ,. ? ? ? ? ? ? ? ? 24 
1. The Fire, Offense, and Arrest Records ? ? . ? . ? 24. 
2. Validity Problems with Official Sta-
ti.stics . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 
c. conceptualizing the Data: A Problem in Sim-
pl.ifica ti on ..?........... " . . . . . . . . . . . . . . . . . . .. . . . 32 
1. Classification of the Police and Fire 
Data .?? "' . ei ???????????? " .. ........ ., ? ? ? ? ? ? ? ? ? .. J3  
2. Definition of the Riot Period of Organized 
Response and a Comparable Normative Time 
Period ......................... fl ............ ., ,39 
J. Definition of the Riot, Corridor, and 
Non-Riot Spatial Areas ???????????.????????? 43 
D. Statistical Analysis of the Data "??? ??? ??&e"?? ?? 
1. Tabular Presentation of the Data ?.?? e ?????? 45 
2. The Lambda Statistic ............... " ........ . 45 
D.. SU1l1Inary ........................ " ? ? ? ? ? ? ? ? ? . ? ? ? .. ? .. .. 50 
IV. THE RESULTS ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? . . ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 52 
A. Introduction .... ,. ...................... ,_ ........ ., 52 
B. The Offense and Fire Data ????? " ??? fl ????????????? 53 
1. The Offense and Fire Data for the Riot 
Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . 54  
2. The Offense and Fire Data for the 
Corridor .Areas ....?. t'I ????????????? " ? ? ? ? ? ? ? ? 57 
J. The Offense and Fire Data for the 
Non-Riot Areas ?????????????.???????? o. ? ? ? ? ? 59 
xvix 
Chapter Pa ge 
4? Comparison of the Offense and Fire Data 
for the Riot, Corridor, and Non-Riot 
A:r'eas ................. . .................. .... 6J 
5. Evaluation of the Offense I)q ta 
Hypothesis ..... ,. ............................. 71 
c. The Arrest Data . ................................. . 75 
1. The Arrest Data for the Riot Areas ? ? ? ? ? e ? ? ? ? 76 
2. The Arrest Data for the Corridor Areas . ..... 80 
3. The .Arrest Data for the Non-Riot .Areas .. " .. " 82 
4. Comparison of the .Arrest Data for the ..........R iot, Corridor, and Non-Riot Areas " ..... 85 
5. Evaluation of the Arrest Data Hypothesi.s 93 
D. Comparison of the Offense and Fire Data with 
the Arre st Dd ta .. " .............. " ... " ........... ,. . . % 
1 . Evaluation of the Comparative Izy'pothesis ?.?? 103 
E. SUillT!lary ................. e ...................... " ? ? ? ? ? 105 
V. THE CONCLUSIONS ? ? ? ? ? ? ? ? . ? ? ? . ? . ? ? ? . ? ? ? ? . ? . . ? ? ? ? ? . ? . ? ? . . 108 
A. In?troduction ..................... ~."?????????????? 108 
B. Major Results of this Study...................... 108 
c. Suggested Further Resea:rch ???????????????.?.????? 112 
1. The Spatial Varlable ? ? ? ? ? ? ? ? ? ? ? . ? ? . ? ? ? ? . ? ? ? ? 112 
2. The Temporal Variable ??????"???????????,.?e? ? 115 
3. Inferred Propositions for Further Testing 117 
D, Methodological and Theoretical Implications 
of this Study ??????????.?????. , ? ? ? ? ? ? ? ? ? ? ? ? ? . ? ? ? ? 123 
APPBND IX A. NaI'ES ????? ? ? ? ? ? ? ? ? , ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? , ? , ?? , , ? ? ? ? 129 
APPENDIX B. T.ABl,E;S ????? fl ??? fl ?? fl le ? .. "' ? ? ? ? ? ? ? ? ? ? ....... ". ? ? ? ? ? ? ? ? ? 142 
---
xx 
Chapter Page 
APPE:NDCT C. FIGURES ? o ?????? ? ???? ? . ............................ . 195 
AP}>ENDI X D. JvfAPS " ...... " ??? ? ? ?..? ? ................. . ........... . 198 
SELECTED BIBLIOGRAPHY ??? ? ? ? ? ? ? ? ? ??????? , ????????? , ????????? ? ?? 200 
LIST OF '1' 1\BLES 
Table Page 
l. The General Code .,ind Description of that Code for Arrests 
and Offenses Recorded by the Distrfot of Columbia Metro-
politan Poliee Department ???????????.?..???.??.????.??.?.?? 143 
2. Number of Disorderly Conduct and Curfew Violations 
During the April, 1%8, Riot in Washington, D. C. by 
Date ??????. , , ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? , ? ??? , ????? , ??? , ?? , ?? , ??? , ??.? , 144 
J. Percentage of Offenses by General Category and by Time 
Period for Riot Area Locations ?????????.??.?.?..?.?? , ? ? . ? ? ? 145 
4. Percentage of Offenses by Collapsed Category of Crimes 
Against Persons and by Time Period for Riot Area 
Locations ..... ?. ,. ? ...?.... " "' .... " ....................... " . . . . . 1.46 
5. Percentage of Offenses by Category of Cr~nes Against 
Property and by Time Period for Riot Area Locations ? ? ? ? ? ? ? ? 147  
6. Percentage of Offenses by Collapsed Category of Crimes 
Without Victims and by Time Period for Riot Area 
Locations ......... "" ...... ,. ............ " ........................ ,. 148 
7. Percentage of Offenses by Category of Crimes Related to 
Fire and by Time Period for Riot Area Locations ???.???????. 149 
8. Percentage of Offenses by Collapsed Category of Miscellan-
eous Crimes and by Time Period for Riot Area Locations ?.??? 150 
9. Percentage of Offenses by General Category and by Time 
Period for Corridor Area Locations ?????????.??.? ?? . ???...?. 151 
10. Percentage of Offenses by Collapsed Category of Crimes 
Against Persons and by Time Period for Corridor Area 
Loe a tions .... " ? " ? ? ..... " . ~ ............. fl .......... ,. ?? " " ? " ? " ? " ? ? ? 15 2 
11. Percentage of Offenses by Ca tegory of Crimes Against 
Property and by Time Period for Corridor Area Locations ???? 153 
12. Percentage of Offenses by Collapsed Category of Crimes 
Without Victims and by Time Period for Corridor Area 
Locations ?????? ? ? ? ? ? ? ? ? ? ?? ? ? ? ? ? ? ? ?? ? , ? ? ? ? ? ?? . ? ? ? . ? ? ? ? ? ? ? ? ? ? 15!+ 
13, Percentage of Offenses by Category of Crimes Related to 
Fire and by Time Period for Corridor Area Locations ?.?????? J55 
14. Percentage of Offenses by Collapsed Category of Miscellan-
eous Crimes and by Time Period for Corridor Area Locations. 156 
xxi 
,:x.ii 
Ti:lble Page 
15. Perr.!eritage of Offenses by General Category and by Time 
Period for Non-Riot Area Locations ?? , , ?? , ? , ?? , ??.??.??.?.?? 157 
16 . Percentage of Offenses by Collapsed Category of Crime3 
Against Persons and by Time Period for Non-Riot Arec1 
Locations ......... ? ? . ? ? ? ? ? ? ? . .. . . . . . . . . . . .. .. . . . . . . . . . . . . . . . . . . ]_58 
17. Percentage of Offenses by Category of Crimes Against 
Property and by Time Period for Non-Riot .Area Locations ??.. 159 
18. Percentage of Offenses by Collapsed Category of Crimes 
Without Victims and by Time Period for Non-Riot .Area 
Locations ...... " .. ? ? ? ? ? .................... ,. ........... " . . . . 160 
19. Percentage of Offenses by Category of Crimes Related to 
Fire and by Time Period for Non-Riot Area Locations ?????..? 16] 
20. Percentage of Offenses by Collapsed Category of Miscellan-
eous Crimes and by Time Period for Non-Riot Area Locations ? 162 
21 , Lambda Values for the Offense Data by General Category and 
Collapsed Subcategory and by Spatial Area ?.??? ..? ?.???????? 163 
22, Percentage of Arrests by General Category of Violation and 
by Time Period for Riot .Area Locations ?????..?.?...????.??. 164 
23, Percentage of Arrests by Collapsed Category of Crimes 
.Against Persons and by Time Period for Riot Area Locat ions ? 165 
24, Percentage of Arrests by Category of Crimes Against 
Property and by Time Period for Riot .Area Locations 166 
25 . Percentage of Arrests by Collapsed Category of Traffic 
Viofotions and by Time Period for Riot .Area Locations 167 
26. Percentage of Arrests by Collapsed Category of Crimes 
Without Vfotims and by Time Period for Riot Area Locations ? 168 
27. Percentage of Arrests by Category of Crimes Related to 
Fire and by Time Period for Riot Area Locations ????..?.?..? 169 
28. Percentage of Arrests by Collapsed Category of Miscellan-
eous Crimes and by Time Period for Riot Area Locations ?..?? 170 
29. Percentage of .Arrests by General Category of Violation and 
by Time Period for Corridor Area Locations ??...?.????..??. , 171 
JO. Percentage of Arrests by Collapsed Category of Cr i.mes 
Against Persons and by Time Period for Corridor l1rea 
Locat;ions ............. ,. .... .., .... ,, . " . .. ........ . " . .. ........ .., . ... . 172 
xxiii 
'fable :Page 
31. Percentage of Arrests by Category of Crimes Against 
I'rop0rty and by Time Period for Corridor Acea Loi:!ations ???? 173 
32, Percentage of Arrests by Collapsed Category of Traffic 
Violations and by Time Period for Corridor 11rea Loca-
ti ons ..... .. ,. ... . ....... ,. ...... " ....... ,. ......... .. . e ..... ft " ? ,. ? ? ? ? J. 7 Lt 
JJ , Percentage of Arrests by Collapsed Category of Crimes 
Without Victims and by Time Period for Corridor Area 
Loe a ?tions ..... ft ... .. .... 9' ............. ,. ............... ti ???? .,, ? ? .. ? .. ? 175 
34. Percentage of Arrests by Category of Crimes Related to 
Fires and by Time Period for Corridor Area Locations ?.????? 176 
35. Percentage of Arrests by Collapsed Category of Miscel-
laneous Crimes and by Time Period for Corridor Area 
Lo at ions . . . . ? ? ? . . . .. . . . . .. . . . . . . . . .. .. .. . .. e .. ? .., ??? ,. ,,, .. " - " ? .. .. ? .. ? ? ? 177 
J6. Percentage of Arrests by General Category of Violation 
and by Time Period for Non-Riot Area Locations .??.?.??????? 178 
37. Percentage of Arrests by Collapsed Category of Crimes 
Against Persons and by Time Period for Non-Riot Area 
Locations ..... ,. . " .. " .. ... .... ., ......... ,. .... 9 .... " e ?? " ?????? fl" ..... 179 
J8. Percentage of Arrests by Category of Crimes Against 
Property and by Time Period for Non-Riot Area Locations .??. 180 
39, Percentage of Arrests by Collapsed Category of Traffic 
Violations and by Time Period for Non-Riot Area Loca-
t j ,. ons l'I ft, ? ? f fl, ... 9 e e ? ,- ? fl f ? ? l'I e ft ? l'I II e 9 9 Ill f 9 l'I fl 9 e tll ? fl ? " e ft ? ? 11 $ f Ill- e f e, e e It ? 181 
40. Percentage of Arrests by Collapsed Category of Crimes 
Without Victi1ns and by Time Period for Non-Riot Area 
Loe a +,ions . ,,. fl' . ..... " & .... ,. ? fl 9 ,. ??? fl ? ??? ? ??? , " .. .... fl ???? fl " ? ? ? ? ? ? 182 
41. Percentage of Arrests by Category of Crimes Related ?to 
Fires and by Time Period for Non-Riot Area Locations . ? ? ? ? ? ? 183 
42. Percentage of Arrests by Collapsed Category of Miscel-
laneous Crimes and by Time Period for Non-Riot Area Loca-
t ions 9 9 e ? 9 l'I ? ... fl fl fl 9 l!I fl ?? i, " 9 fl I.I ? ,i fl ? ft Ill 9 fl It 9 f e 9 ? fl ? e It fl 9 fl ? II f fll fl fl ? ft fl 9 e 184 
43, Lambda Values for the Arrest Data by General Category 
and CDllapsed Subcategory and by Spatial Area ?? ? ??.?.?.???? 185 
44. Percentage of Offenses Reported to the Police by Time 
Period and by Locational Area .???? ? .???.??????????????? , ? , ? 186 
45. Percentage of Arrests Made by the Police by Time Period 
and by Locational Area ??????? , ? ?. ? ? , ?. ? ? ? ? ? ? . ? . ? ? ? ? . ? . . ? ? ? . 187 
xxi v 
Toble l'age 
46. Tol,al Number of Offenses by Category of Viola tion and 
by Date for the Normal Time Period ????.?.??.?????????.??..? 188 
1/1. Total Number of Offenses by Category of Violation and 
by Date for the Riot Time f'eriod .?????..??????.....?..?.?.? 189 
48. Totdl Number of Arrests by Category of Violation and 
by Date for the Normal Time Period ????.?.?..?....????.?.??? 190 
49. Totol Number of .Arrests by Category of Violation a:1d 
by Date for the Riot Time Period ????...........?....??.?... 191 
50. Total Number of Offenses by Date and by Location ?.?.?.???.. 192 
51. Total Number of Arrests by Date and by Location ????.?...?.. 19J 
52, Total Number of Fires by Date and by Location ?????????????? 194 
LIST OF FIGURES 
Figures Page 
1. First General Classification of Arrests and Offenses, 
Indicating the Code and the Descrip?~:i.on of the Code . ? ? ? ? . . 196 
2. First General Classification of Arrests and Offens es, 
Indicating the Subcategories and the Codes and Their 
De s cription .. . ,. " ,. ,. . ? "' " .... e ... . . .. ... . ..... ,. ?? 19 ??? ?? ,. ? ? ? ? ? ? ? ? 197 
XXV 
LIST OF :MAPS 
Map Page 
1. Indicating the Riot, Corridor, and Non-Riot Areas ???...???. 199 
xxvi 
CHAPTER I 
STATEMENT OF THE PROBLEM 
Introduction 
One of the most visible characteristics of riots is the 
volume and variety of criminal activity, manifested primarily in 
the burning and l ooting of property within the affected riot areas. 
One of the major concerns , then, for the social system is to try 
to minimize the effects of the riot. Since the police are the 
primary agency involved in responding to this aspect of the 
crisis situation, the study of police activity is a logical and 
as yet poorly explored aspect of riots. 
Three objectives are accomplished in this chapter, First, 
a brief statement is presented which suggests that the current 
sociological literature on riots is predominantly concerned with 
causal analyses at various levels of investigation. Second, it 
is maintained that s001e aspects of riots have not received at-
tention by sociologists and that this thesis is an initial s te p 
toward the investigation of these neglected aspects. Finally, 
the objective of this study and the specific ~uestions to be 
answered are posited. 
The Major Concern of the Sociological Literature on Riots 
The majority of the sociological riot literature is concerned 
with causal analyses and directed toward the investigation of the 
\ 
2 
relationship between selected independent and dependent variables. As 
such, this literature treats a riot or some aspect of a riot as the de-
pendent variable, McPhail argues that much of the recent riot litera-
t ure has been biased toward causal analyses and that it has failed to 
rP.cognize both the complexity and the sequential aspects of riots. 1 
The position taken i n this study is t .hat riots may also be treated sui 
generis or as things in themselves. It is not argued that the factors 
which cause riots are unimportant, but that riots evidence a complexity 
and a variety of activities apart from tbe factors which cause them. 
Specific Problematic of this Study 
The problem under investigation in this study addresses itself to 
some of the complexities of riots during one of the posited sequential 
stages of the Washington, D. C. riot of April, 1%8. Two of the prin-
cipal factors in a riot situation become the rioters themselves and 
their activities, on the one hand, and the control agents and their 
responses to the actions of the rioters, on the other hand. 
The purpose of this study is to present a descriptive, ecological 
analysis of criminal activity during one specific time stage--the or-
ganized response stage--of the Washington riot of April, 1968 and to 
establish social indicators of the situation reported to the po.lice 
and the police response to that situation. The organized response 
stage of a riot is defined as the time during which formal organizations 
systematically respond to the needs of the affected community. The 
focus herein is upon the District of Columbia Metropolitan Police De-
partment, as one such formally established agency whose responsibility 
in part is to stop or minimize the criminal actions of the rioters . 
') 
.,) 
The officiul stutistics of the District of Columbia Fire Department 
and the police department are used as partial indicators of the crim-
inal activity reported to the police ,rnd the response made by the po-
lice to that act.ivity. Social indicators are here defined as the ma.jor 
criminal violations which occur and set the riot apart from the non-
riot time periods. This objective makes it necessary to contrast the 
selected riot period statistics against a comparable set of statistics 
from a non-riot time, which herein are designated, respectively, as the 
11riot" and "normal" time periods. The ecological variable of space i.s 
trichotomized into: (1.) the riot areas, where the greatest concentra-
tion of riot damage occurred, (2) the corridor areas, where more spo-
radic riot related damage and destruction occurred, and (3) the non-riot 
areas, where little or no destruction occurred. The sociological disas-
ter model fa used as the source of thE relevant conceptualizations of 
the ecological variables of time and space. 
~umptiops. Two major assumptions are made. The first is the 
rather obvious assumption that activities during riot periods differ 
from activities during non-riot or normal periods. That differences 
exist is hardly debatable, but the real question becomes one of the de-
gree of difference between riot and non-riot periods. The second as-
sumption is that the official police and fire department statistics are 
partial indicators of both the riot and normal time period situations. 
The police department offense record and the fire department fire re-
cord are utilized to describe the situation and are partial indicators 
of the riot situation in the sense that they reflect only the situation 
as reported to and officially recorded by these agencies. Obviously, 
this limitation is more apropos for the police than the fire data. The 
4 
arrest record is used to describe the response which the polic e mad e 
and is a partia l indicator bec.:ause it cont ai'."ls ,,,hat the polic e them-
s elves officially reported and recorded. 
Questions. Two questions are posited for investigation herein. 
First, what degree of difference in criminal activity existed between 
the riot period of organized response and a representative normal time 
period and between differing spatial locations, as reflected by the 
official police and fire department statistics? The degree of differ-
ence is measured by the lambda proportionate reduction in error sta-
tistic, in which large values will be interpreted to indicate that 
differences do exist. Focusing, then, on the org,rnized response stage 
of the riot and on the riot, corridor, and non-riot spatial areas, the 
offense, fire, and arrest data will permit a partial assessment of the 
degree of difference between the riot and the normal period with refer-
ence to the situation reported to and the response made by one agency 
of social control in the Washington community. 
Second, what kinds of differences were evident? Once we have 
measured the degree of difference, we may proceed to the second ques-
tion concerning the nature of the difference. It is here that :indica-
tors will be selected which best describe the riot situation and the 
police response to that situation. In other words, we will select those 
specific types of criminal violations from the large number of legally 
defined criminal statutes which best illustrate the nature of the dif-
ferences between the riot and non- riot time periods. It is hoped that 
this analysis will indicate that certain crimes are important ones dur-
ing riots and that only sane crimes change between the two time periods ? 
.As such, this study will function in part to simplify some of the data 
5 
which is available for the subsequent study of riots. 
Summary 
The problematic in this study is two-fold. Using official statis-
tics as reflective of the criminal activity during a riot and the re-
sponse made to this activity by one agency of social control, we will 
measure the degree of difference between the riot period of organized 
response and a representative normal period by category of crime and by 
spatial zone, describing the nature of these differences, and develop 
indicators which best surrunarize these data. As such this study will be 
a descriptive, ecological assessment of the official police and fire 
department data and an ini.tial step toward the investigation of the 
neglected aspects of riots in that it will focus on the variety of 
criminal activities reported to the police and the responses made by 
this agency to this criminal activity, emphasizing the differences be-
tween riot and normal periods and among various spatial locations with-
in the Washington area. 
The concern of this study, then, is with some of the ecological di-
mensions of riots as suggested by the sociological disaster literature 
and with the development of indicators of riots utilizing official sta-
tistics. It is maintained that this thesis is an initial exploratory 
investigation into one part of the neglected aspects of riots. The 
social indicators which are developed will be useful to the further 
study of riots in two ways. First, they will direct attention toward 
only those criminal offenses and arrest responses which are paramount 
to riots, thereby simplifying the m~ss of available official data. Sec-
ond, these indicators of riots may be used in the subseq,l?nt study of 
6 
riot s in other corrununities, as they are data which are collected by 
all police jurisdictions and classified in a consistent manner as sug-
gested by the Federal Bureau of Investigation's manual for the uniform 
repor t i.n g o f cri.m e . 2 
CHAPTER II 
REVIEW OF THE LITERATURE 
Introduction 
The first section of this chapter begins with a brief discussion 
of the major current conceptualizations used to organize the study of 
riots. r-t is shown that the predominant concern of this literature is 
with the causal relationships between selected independent and depen-
dent variables, with the latter being the riot or some aspect of the 
riot itself. The second section reviews the recent literature which 
has been explicitly or implicitly critical of the bulk of the riot lit-
erature. The critical literature maintains either that the empirical 
support for the hypotheses suggested by these perspectives is somewhat 
deficient or that other aspects of riots have been neglected by the em-
phasis on causal relationships. It is sh01,m how the problem treated in 
this thesis is a response in the direction of study suggested by this 
critical literature. The third section begins with a brief overview of 
the disaster literature and then focuses specifically upon those as-
pects of the disaster perspective which relate to the ecological dimen-
sions of time and space. 
Ma.jor current Conceptualizations in the Study of Riots 
The deprivational approach is one major theoretical perspective 
and involves some variant of the frustration-aggression bypothesis. 1 
Here, a perceived unfair distribution of rewards is seen as the cause 
of violence. Some recent studies have focused on variables which 
7 
P.,: 
intervene between the initial state of frustration and the final out-
come of aggression. 2 Other studies focus on some variation of the 
frustri;j tion-aggression theme, such as abso.1.ute deprivation, 3 relative 
deprivation, 4 .:md the revolution of rising expectations. 5 
Another major set of theoretical approaches to riots reflect the 
in~bility of some groups to identify with some of the normative pat-
terns of the existing social system or the inability of some agencies 
to restore these norms. Among the studies oriented toward this theme 
of community disorganization or lack of integration are those whi(:h 
6 
pos?i t po1  ?t.ic a.L,  a 1 i?e na t.io n, 1 ow s t atus, 7 J. cri?m i?n a 1 ? 
8 
1.s t 1? .c t en d enc1. .es, 
and the criminal riff-rciff theory9 as independent variables which cc:Juse 
riots or specify who participates in riots. 
The third major theoretical perspective currently utilized in t,he 
interpretation of riots is the group conflict theme. Some sociologists 
argue that riots are mechan.isms of political and economic protest, 10 
while others perceive a more gener ic relationship of conflict between 
blacks and wh ?t 11 1 es. 
In all of these studies independent variables are selected as fac-
tors which cauae riots or specify which individuals or social groups 
are likely to participate in riots. Therefore, the riot itself or some 
aspect of rioting is treated as the dependent variable. In other words, 
none of these studies addresses itself to the reaction of the community 
to riots. 
Recent Critical Assessments of the Previous Riot Literature 
Some of the recent literature on riots has been c:ritical of the 
above mentioned conceptual orientations. Two types of criticisms have 
9 
been explicitl y presented : f i r st , those which s uggest t hat the degree 
of suppor t offered in s ubstantiation of the hy potheses presented in 
the earlier r esear ch has been too mi ni ma l to war rant acceptance of 
t hese hypotheses and, sec ond , those which suggest that the previous 
s tudies have neglected aspect s of r i ots and have over simplified other 
aspects of riots. 
Spilerman has examined a number of pers pect ives which offer a 
variety of independent variabl es us ed to expla i n t he l ocation of 
riots . 12 Among the independent variables associated with the inci-
dence of riots whi ch were as sessed in t hi s research were : (1) high 
s ocial disorganization, (2) absolute deprivation as indicated by the 
black population's material conditions of life, (3) relative depriva-
t ion where blacks compare t heir situation with that of whites , (4) t he 
rising level of expectati ons, where blacks perceive the gap between 
wha t they expect and what they have really attained, (5 ) politi ca l 
alienation caused by existing political structures which are unrespon-
s i ve to the needs of the black population, (6 ) the negati ve reinf orce-
ment thesis, which posits that any disorder decreases the l i keli hood 
of subsequent disorders, (7) the positive reinforcement thesis , which 
ma i ntains that any disorder increases the likelihood of subsequent dis-
orders by leaving some sort of polarizing residue, (8) the ge ographic 
c ontagion hypothesis, which argues that riots are di rect l y r elated to 
proximity, and (9) that riots are random events in t he s ense t hat all 
communities have an equal disorder-proneness. 
Spilerman's research generated either no support or at bes t mini-
mal support for all of these independent "causal" var i ables when used 
t o explain the incidence of riots . Two variables, the number of blacks 
10 
in t he community and the dummy variable of region dichotomized i nto 
South and non-South, were t he best indicators of t.he locc1tio11 of dis-
orders . He w~ites : 
Yet, the crucial point i s not tha t non-white popu-
l a t ion i s s o important f or explaini ng t he distribution 
of dis orders--the number of Negroes woul d appear to be 
a basic r esource f or Negro upr isings --but tha t , after 
the effect of this conceptually prior variable ha s been 
removed , t he ot her community characteris t ics account 
for s o little. 
? ? ?T?he? c?o?nc?lusi?on* ?fr?om ?t?his? ?an?a?l y.s,is ?is? ?th?a t? t?he ?r a?- ? 
cial disturbances of t he 1960' s were not res pons es to 
conditions in the l ocal community. Disorder-prone ci t-
i es do dif f er from t heir less trauma t ized neighbors in 
many significant res pects. 
? ? ? ? ? ? ? ti ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 
However, these condi tions have little to do with a com-
munity being prone t o disorder, and are instead t he i n-
cidental characteristics of cities with large Negr o 
populations .13 
Spilerman concludes: 
?? . I would argue that although different communities 
are not equally prone to racial disturbance, the sus-
ceptibility of an i ndividual N~gro to partic ' pating in 
a disorder does not depend upon the struct ural charac-
teristics of the c orrununity in which he resides . 14 
Af ter pr esenting a general version of the "deprivati on-frustr ation-
aggression" hypothesis, McPhail analy zes the adequacy of the data pre -
s ented in suppor t o f' tl1 '1 s t hes i' s in recent J' ourna 1 ar t i? c1e s. 15 He 
concludes that while the majority of the independent variables asooci-
ated with riot participation in these previous studies are s ta tis ti-
cally significant and consistent wit h the direction originally 
hypothesized, the magnitudes of the associations are overwhelmingly l ow . 
He writes: 
In view of these r esults concerning individual riot pa r t ici-
pation, and the results of Spilerman (1970) and others con-
cerning the occur rence of riots, there i s considerable 
reason for re,jecting the sociological and popular cliche 
11 
that absolute or relative deprivation and the ensuing 
frustration or discontent or despair is the root cause 
of rebellion.16 
The impact of these recent studies by Spil.erman and McPhail is to 
question the validity of the interpretations whi.ch have recently dom-
inated the sociological literature on riots. What these researchers 
are suggesting is that the studies analyzed have been too quick to ac-
cept empirical support and/or that they have overinterpreted the data 
to some degree. In sum, these authors argue that some rethinking is 
necessary on the sub,ject of the causes of riots as our dominant inter-
pretations are without any firm empirical support. 
This critical posture also moves into a s econd direction, sug-
gesting implicitly or explicitly that some aspects of riots have been 
either neglected by sociologists or subjected to oversimplification. 
Cohen writes: 
??? the dominant bias in American sociology has been 
toward formulating theory in terms of variables that de-
scribe initial states ., on the one hand, and outcomes, on 
the other, rather than in terms of processes whereby acts 
and complex structures of actions are built, elaborated, 
and transformed . 17 
Grimshaw maintains that: 
There are more complicated dimensions to this issue 
[riots] than to any other I have examined in my role as 
a sociologist ?18 
Rainwater remarks: 
Riots are difficult to control precisely because of this 
voluntary division of labor among the participants. Be-
cause their many different sorts of activities require 
different sorts of responses, the riot becomes a highly 
complex event that can be brought under control only by 
a mass show of force. 19 
Finally, Q,uarantelli and Dynes state: 
On the contrary, if we have learned anything from otW 
studies of these situations, it is ?that the behaviors 
12 
and part i c ipants i nvol ved are f ar mor e heterogeneous 
t han is i mplied in a statement that "violence" broke out 
in this ghetto or t hat the Ne groes in a particular com-
munity "rioted." Snipi ng and looting, ars on and vandalism 
and other behaviors are not the same kinds of acts; dif-
f er ent part i cipant s take part in t hese activities , the 
action t akes place at differ ent locations and at differ-
ent time periods of the di s turbances.20 
McPhail ar.gues tha t many recent s t udies f ocus on the dependent 
var i able of riot part icipa t ion operationa l ly measured by arrestee sta-
t us, the r espondent's reported participati on status, the r espondent's 
witness of others' behaviors , and/or some combi nation of the above . 21 
Whil e t hese operational defi nitions vary, McPhail bel ieves they all 
have the ccmmon characteristic t hat t hey fail to sample t he res pondent's 
behavi or cont ent through t he t ime duration of the riot . 22 In other 
words, he f eels that any one ind i vi dual may qui t e l ikely be engaged in 
r i ot ing, non-r ioting, and counter- r iot ing acti vi t i es a t differ ent times 
t hroughout t,he duration of the r i ot itself and t hat the r ecent litera -
ture has failed t o adequately deal with this poss i bilit y. He writes : 
Rather, he [a rioter] is likely to be intermittentl y en-
gaged in a wide range of routine and "illega l " act ivities 
during the course of his presence in the area of the dis--
order . Unfort unately, measures of civil disorder parti -
cipation have failed to recognize ., record, and attempt to 
account for this differentiat ion in behaviors through 
space and across time. Perhaps, unwittingly, participa-
tion in civil disorder has been conceptualized as a mono-
lithic phenomenon and measured accordingly.23 
McPhail concludes this article with an alternative focus, s ug-
gesting s ever al major questions f or stuay. 24 Among these is the ques-
tion which refers to 11 , ? , the variety of i ndi vidua 1 and Joined 
per formances in which people engage during the course of the dis -
order . 1125 Finally, McPhail concludes: 
Civil disorders are complex and differentiated 
phenomena . Attempts to account for their occurrence 
and individual participation therein have failed to 
13 
acknowledge this complexity, theoretically and opera-
tionally .26 
Most of the above sociologists suggest that more attention be 
placed upon the variety of behavior whi.ch occurs during civil dis -
orders. McPhail argues for a conceptualization of this variety which 
encompasses the temporal clDd spi:ltial dimensions of the riot. This 
thesis is ,rn initial step toward the investigation of this complexity 
as suggested by the riot literature itself. In view of these specific 
suggestions as well as the lack of conclusive empirical evidence which 
is supportive of the previously investigated hypotheses, it seems 
quite justifiable to begin to move in the direction suggested by 
McPhail and others. 
The Disaster Literature 
The sociological disaster literature appeared principally during 
the middle 1950 1 s and early 1960's. Among the variety of dimensions 
of disasters studied have been the definition of disasters, 27 the clas-
sification of di. 29 sas t ers, 28 cti? sas t ers ro1e s, pani. c th eory, JO con-
vergenee behav.i or, Jl di. sas t er commun.i ca t i.o n ne t wor k s ., 32 vari?o us method-
ological issues related to disaster research, 33 and types of organized 
b e h avi. or i. n ct.i sas t er. 34 Since the spatial and temporal classifications 
provide in part the relevant categorizations used in this thesis ., i.t is 
to these that we now turn. 
The temporal dimension. Most descriptive studies of disasters 
have reported and organized the data in terms of the variables of time 
and space . Killian writes: 
Functional time-phases and spatial zones have been identi-
fied in most disaster studies a~d these concepts prove high-
ly useful in ordering the data.35 
14 
Killian suggests four chr onological time phases. 36 First, the warning 
per iod is the time during which information is available concerning 
the probable danger of the approaching disaster . Second, there is the 
period of impact, during which the destructive agent is affecting the 
community. Third is the period of emergency, when there is a rela-
tively unorganized response to the disaster made by the affected popu-
l ation. During this per5 cct, res~ 1Je operations, first aid, and emergen-
cy medical care are offered to the victims of the impact period, 
Finally, there is the period of recovery during which the long term 
activities designed to restore the system to a functioning entity are 
carried for th. 
Fritz presents a partial listing of time sequences, which are 
quite similar to those mentioned above, but he adds the initial period 
of preparation for disaster. 37 This period is only evident in those 
conununities which have experienced frequent exposure to disasters , and 
it refers to the preparations made by the community in anticipation of 
future disaster experiences. For example, i t is likely that most min-
ing communities experience this initial time phase and during it devel-
op a fatalistic attitude relative to the incidence of a disaster, as 
the possibility of cave-ins and explosions becomes something one lives 
with in the performance of this occupational role. Fritz also sub-
divides Killian's period of warning into, first, the period of disas-
ter warning in which information is received concerning the probability 
of a disaster, and second, the period of threat in which the probable 
danger is perceived as an actual danger which has not yet affected the 
community but whose effect is inevitable. 38 For example, the period 
of warning would begin with the issuance of a statement by the mas s 
15 
media conc erning t he possibility of, say, a tornado. The per i od of 
threat begins when t he warned community perceives that the tornado is 
not only a possibility but also sees it as de f initely going to strike 
them. 
Barton suggests a further subdivis ion of Kill ian 's per iod of 
emer gency int o t he period of relatively unorgani zed r esponse during 
which s urvivors search for their own family member s and administer aid 
in a somewha t hap.hazard fashion and the period of organized social re-
s pons e where formal organizations specifically tra i ned for emer gency 
r elief operations are activated and present i n the stricken commun-
i t y.39 The former refers to a comparatively short period of t ime dur-
i pg which emergency assistance is rendered on a one to one individua l 
basis as opposed to the longer latter period during which forma l organ-
izations more systematically provide relief functions. 
In sum, the above sources suggest the total range of poss ible 
chronological time phases in disasters: (1) the period of pre para -
tion--which is only present when the defined community ha s experienced 
frequent exposure to disasters; (2) the period of warning--during 
which information is received that a disast er might occur; (J) the per -
iod of threat--during which the possibility of danger is seen as ac tual 
danger, an actual danger that has not yet affected the communit y but 
is going to affect the community; (4) the period of impact--when the 
destructive agent is actually at work in the cOllUllunity; (5) the period 
of relatively unorganized response-- during which surviving individuals 
randomly provide aid to casualties; (6) the period of organized re-
sponse--during which formal organiza ti.ans more systematically and fully 
provide aid ; and (7) the period of r ecovery--which begi ns when the 
16 
crisis situation is defined as over and the restoration of the com-
munity begins. 
When investigating any one disaster situation, one might find 
that some of these time phases are missing or that some of them are 
truncated or elongated in comparison with the amount of time devoted 
to these phases in other disaster investigations. Other variables af-
fect the likelihood of the occurrence of some of these phases. For 
example, the nature of the stress agent can affect both the period of 
warning and the period of threat. In the past the weather bureau had 
neither the resources nor the sophistication to spot tornado dangers, 
hence there was no period of warning. Further, if the destructive 
agent is an explosion in a chemical factory, it is quite likely that 
its suddenness would eliminate both the periods of warning and threat. 
Again, in the case of a tornado, the impact period is likely to be no 
more than a matter of minutes, while this same per iod in the case of a 
flood might take several days . Further, in the former the periods of 
relatively unorga nized and organized response will be likely to follow 
the conclusion of the impact period, while in the latter unorganized 
and organized response will quite likely overlap with the impact period. 
Therefore, the probability of occurrence of the seven previously men-
tioned time phases is not equally likely as one moves from one specific 
disaster to another. Nor will the amount of time encompassed by any 
one specific phase be congruent as you move to the same time phase in 
another disaster situation. Finally, some pha s es may be fairly dis-
tinct from others or they may overlap with others depending upon the 
specific nature of the disaster agent. 
The temporal sequence discussed above suggests that the community 
17 
affected by a disaster moves through a number of qualitatively dif-
ferent stages which are ordered and which call forth different kinds 
of responses during the various stages. It is argued here that some 
of these temporal stages are useful to an analysis of civil disorders 
and that when so applied to the t~~e dimension in riots, a certain 
pattern seems applicable. If the period of preparation exists, it is 
likely that it will only be present in those communities which have 
experienced riots or similar massive demonstrations in the past. There 
is evidence that some segments of the M:ishington community had initi-
ated preparatory measures prior to the April, 1968 riot, for example, 
the implementation of a simplified arrest form by the District of 
Columbia Metropolitan Police Department40 and contingency planning for 
future scheduled demonstrations by several c ommunity organizations. 4l 
The periods of warning and of threat seem to have little appli-
cability to civil disorders . For the sake of argument here, let us 
assume that the independent variables mentioned briefly in the be-
ginning of this chapter as the causes of riots are in truth causal 
factors. It is logical to assume, for example, that relative depriva-
tion character izes all of the corrununities in the United States . But 
all communities in this country have not experienced civil disorders, 
and there f ore we must conclude that these conditions are too gener-
alized to indicate any kind of warning period. Some sociologists argue 
that riots are pr eceded by amongst other variables a "precipitating 
event. 11 42 But quite often this pr ecipit ating factor is of such a nature 
that it can be interpret ed as such only in retrospect af.ter the impact 
period has started. In other words, the "precipitating event" is often 
a quite common event, such as a raid on an after-hours social club 
18 
which began the Detroit riot of 1%7, 43 the rumored beating by polic e 
of an arrested cab driver which led to the Newark riot of 1967,44 or 
the arrest of a black man for a traffic violation in Watts in 1965.45 
These common, everyday events which only rarely result in riots can 
hardly be used as indicators of the period of threat, which posits 
that the danger i.s imminent. While hindsight mLght indeed i ndicate 
underlying causal variables and precipitating events, these latter 
factors are hardly sufficient as indicators of the outbreak of a riot 
in a given place and at a given moment in time. 
The period of impact and the periods of unorganized response and 
organized response overlap in the case of riots. The rioters them-
selves become the disaster agent and are operative at the same time 
that the formal and informal agencies both within and without the af-
fected community are responding to the riot. Finally, the period of 
recovery begins when the crisis is defined as over. There will be a 
qualitatively different type of response b?tween the periods of un-
organized and organized response, on the one hand, and the period of 
recovery, on the other. In the former, the community is directing its 
response to the immediate conditions creaued by the disaster agent, 
conditions which create immediate dangers and hazards to the affected 
population. In the latter period, these dangers have subsided, and 
the reconstruction and restoration of the community begins. 
In this thesis, we will concentrate on the time phase: of organ-
ized response as a whole, which overlaps with the period of impact. 
The spatial dimension. A second major dimension involves the 
spatial characteristics of the disaster situation. Killian46 and wal-
lace47 posit identical concentric zonal schemas. The total impact 
19 
zone is the innermost and the area where the danger and destruction 
caused by the disaster agent is the greatest. Second, the fringe im-
pact zone is the area where destruction occurs but is somewhat molli-
fied when compared to the total impact zone. Third, the filter zone 
is the area where personnel and material goods flowing into and out of 
the impact zones meet. Fourth, the organized community aid zone is the 
area wherein the local resources are marshalled. Finally, there is 
the extra-community aid zone where personnel and materials come to the 
stricken area from places not .oormally perceived to have any real or 
direct vested interest in the disaster-stricken area. 
Of all the conceptual dimensions of the disaster approach, the 
spatial dimension is the one most in need of an operational definition, 
empirical testing, and subsequent revision. As with Burgess' classic 
ideal typical concentric zonal scheme of spatial distribution and 
growth, which emerged from the empirical efforts of the Chicago human 
ecological perspective in urban sociology, the spatial zones in the 
disaster model are idealized constructions which have rarely been used 
in the empirical investigations of disasters. The obvious difficulties 
here are the methodological ones of operationalizing definitions of 
these zones, Probably, one of the reasons for the limited usage of the 
spatial. dimension is that students of disaster have tended to restrict 
their research questions so that the notion of space became at best a 
minor consideration in these past studies. 
Since the dimension of space has not been a major focus of re-
search into disasters, this aspect of the literature provides minimal 
suggestions for the study of riots. As riot behavior is not randomly 
or evenly distributed throughout the affected community, the idea that 
zones exist seems a use ful conceptualization for orga nizing the data. 
Since the f ocus of t hi s t hesi s i s upon one a gency of social con~rol , 
since that agency is l ega l l y restr icted to activity within its politi-
cally defined j uri sdi cti on, and s i nce the police are more oriented to-
ward mi nimizing and/or preventing the i mmedi ate haza rds created by 
r ioters, s ome of the possible disast er zones have minima l utility for 
t his speci fic problemati c. But t he total and fr i nge impact zones are 
quite useful here . The t otal impact zone woul d conta in the geographic 
area of greatest concentrated riot damage, whi le t he fr inge i mpact 
zone would contain an area of les ser damage c oncentrat i on . The r e-
maining area wi thin the District of Columbia would, in terms of damage, 
have to be defined as unaffected. 
Sum.mary and Specifi c Hypotheses 
McPhail and others have argued that t he current literature 11as 
oversimplified the s t udy of riots. McPhail specifi cally s ugges t s con-
cent ration of research effort upon the complexity and variability of 
riot behavior organized through time and acros s s pace , This t hesis 
pres ents an i nitial exploratory effort into the ecological dimensions 
of time and space, suggesting that the disaster literature provides a 
useful conceptualization of these variables. Specifically investi -
gat ed are the police and fire data for the organized response time 
period of the riot which are compared to similar polic e and f i r e da t a 
during a non-riot time period across the dimension of s pac e , c once pt u-
alized as riot, corridor, and non-riot areas. 
Dynes distinguishes four basic t ypes of organizations which ar e 
a ct ive in the period of organized resp onse during a disaster . 48 One 
21 
of these he identifies as the established type of agency, which has 
an already existing structure and which is called upon to perform reg-
ular tasks. He writes: 
Even during a major ommunity emergency these or-
ganizations attempt to adhere to regular activities as 
much as possible . 
.............. \tl!IA.,.,t.ll????? 
If a police or fire department is forced to engage in 
some search and rescue, there is an effort to revert 
back as quickly as possible to the regular work of main-
taining security or fighting fires. Whether intended or 
riot, such restri ction of activity helps prevent disaster 
demands on Type I groups from exceeding organizational 
capabilities . 
??? ,. ............ " .......... It. 
Whatever the reason, Type I organizations attempt to 
restrict themselves to traditional tasks even in an 
emergency and tend to use only their own personnel or 
almost identical personnel from similar groups else-
where. 49 
Therefore, it mlght be supposed that there will be a lesser degree of 
association between the normal and riot time periods in the arrest data 
than in the reported offense and fire data. Further, it may be posited 
that the degree of association will increase in boLh data sources as 
one moves from the non-riot areas to the corridor areas to the riot 
areas . 
Three specific hypotheses are evaluated. First, the degree of 
association between the offenses reported and the selected riot-normal 
time period varies directly with the degree of concentrated riot damage. 
Remembering that the police offense record contains reported crimes 
which are classified and recorded by the police and that the fire data 
contains all the fires reported Lo the fire department, it is suggested 
that when the offenses and fires during the riot period studied are 
compared to the offenses and fires during a non-riot time period, the 
association be t;ween them will increase as one moves from the non-riot 
22 
spatial areas to the corridor spatial areas to the riot spatial areas. 
In other words, the degree of divergence between the riot data and the 
non-riot data will increase the closer one gets spatially to the areas 
of maximum destruction, 
Second, the degree of association between the police's response 
and the selected riot-normal time period varies directly with the de-
gree of concentrated riot damage. Remembering that the police arrest 
rEcord reflects in part the responses made by this agency, it is sug-
gested that there will be a greater divergence in the responses during 
the riot period studied when compared to a comparable non-riot time 
period t he closer one gets to a spatial area of maximum destruction. 
Finally, the degree of association between the police's response 
and the selected riot-normal time period is less than the degree of 
association between the offenses reported and the selected riot-normal 
time period . In other words, when the associations in offenses and 
fires as reported are compared to the associations in arrests as 
reactions made by the police, we would expect greater association in 
the former than the latter. 
More simply, it is suggested, first, that the situation reported 
to the police and fire depar t ments during the riot time period studied 
will not only be different when compared to a comparable non-riot time 
period but also evidence an increasingly greater divergence or greater 
degree of association as one moves along the dimension of space from 
those areas not being burned and looted to those areas where the burn-
ing and looting are most severe. Second, it is suggested that the po-
lice response during the riot time period studied will not only be 
different when compared to a comparable non-riot time period but also 
2] 
evidenr::e an increasingly greater divergence or greater degree of as-
sociation as one moves along the dimension of space from an unaffected 
area to an area most severely affected by rioters. And thirdly, when 
we compare the situation reported to the police and fire departments 
against the responses made by the police, the greater divergence will 
be found in the offense and fire data rather than the arrest data. 
CHAPTER III 
METHODOLOGICAL PROCEDURES 
Introduction 
The first section of this chapter deGcribes the nature of t he 
fire, offense, and arrest records and discusses the validity problems 
concomitant with the use of official statistic::;. The second section 
deals with the need for a simplification of the raw data and posits 
a classification of the police data into socially more meaningful cate-
gories. Further, the operational definitions of the riot period of 
organized response and a comparable normal time period are presented 
as well as the operationalization of the spatial dimension into the 
riot, corridor, and non-riot areas. The third section describes the 
tabular presentation of the data in terms of the categories and the 
descriptive statistics utilized and concludes with a comparison be-
tween the traditional statistic of chi-square and lambda as well as 
the categories used in the interpretation of the selected lambda sta-
tistic . Finally, a summary section will emphasize that there are a 
number of crucia l methodological questions which must be answered be-
fore the degree and type of differences can be assessed. 
The Data 
The fire. offense , and arr?st records . The empirical data herein 
analyzed comes fr~n three sources: (1) the daily alarm log of the 
Communication Section of the District of Columbia Fire Department, 
which recor ds all reported fire alarms by time of day in minutes and 
24 
location by street address; (2) the offense record of the District of 
Columbia Metropolitan Police Department, which records all the sub-
stantiated a11d formally filed complciints of violations by hour of the 
day, location of the crime corrunitted by street address, and type of 
violation committed; 1 and (J) the arrest record of the District of 
Columbia Metropolitan Police Department, which reports all arrests 
made by hour of the day, l ocation by street address, and type of vio-
lation conunitted, 
Since the District of Columbia Fire Department data do n0t formally 
specify and only seldom i nformcilly indicate the nature of the request 
for assistance, it is not possible to differentiate real fires, false 
fire alarms, or ambulance calls on this record. While such a distinc-
tion might be interesting, it is hardly crucial in this analysis, since 
the fire depcirtment must respond to any call, be it one for a real fire 
or a false alarm. Hence, the information contained in this record in-
dicates the total activity reported to the fire department and to 
which the fire department was obliged to respond. These data will be 
used as an indicator of the incidence of fires as reported to the fire 
department. 
The offense record contains the total number of complaints re-
corded by the District of Columbia Metropolitan Police Department, for 
vhich there is substantial evidence that a violation has actually been 
corrunitted. 2 While the origin of the report is unknown, it will be fil-
tered through the police department's perception and recorded on this 
record in a legally defined criminal category. For example, if some-
one broke into a home while the residents were away and the ensuing 
report to the police claimed that a robbery had occurred, the police 
26 
would classify this crime as a burglary to conform to the legal defin-
ition of the violation rather than as a robbery which the victim has 
self-reported. These data reflect only what has been reported to the 
police and must be considered a partial indicator of the situation to 
which the police were obliged to respond. 
The arrest record summarizes the total number of apprehensions 
and chargings made by the District of Columbia Metropolitan Police 
Department, exclusive of non??moving traffic violations. Parking vio-
lations, for example, do not appear on this record. These data must 
s imilarly be interpreted as a partial indicator of the police 's re-
sponse as they do not contain response for assistance which are unre-
lated to criminal activities. Further, they contain only those appre-
hensions which result in formally filed charges. Finally, if an indi-
vidual is app~ehended and charged with multiple violations, he will be 
classified once and only once in the most serious category from amongst 
the multiple charges . Therefore, the arrest record corresponds to the 
total number of people arrested and charged, not to the total number of 
violations perpetrated. 3 
Two general points remain to be made with reference to the inter-
pretation of the data. First, an arrest reported does not necessarily 
coincide with an offense reported. In other words, it is not possible 
to determine which arrests match up with which offense, if . indeed, 
they match at all. s~cond, both the offense and arrest records r eport 
data by hours of the day. The poli ce classify any violation in any 
specific hour if it occurs JO minutes prior to or after the specified 
hour. Fo.c example, a classification of 1:00 A.M. would contain all 
those viol:3ti011s reported between 12:Jl A.M. and l:JO A.M. Hence, 
27 
d~ta recorded for any one day will not contain the first JO minutes of 
that day and will contain the first JO minutes of the subsequent day. 
For reasons of comparability, the fire data, which is reported in the 
time units of minutes, has been so adjusted as to correspond to the 
police department's time classification. 
Validity problems with official statistics. Numerous investigators 
have remarked about ?the difficulties surrounding the analysis of offi-
ci' a 1 cr1J?I 1e sta?t 1? s t1' cs. 4 First, since many crimes are not discovered or 
discovered but not reported, crime statistics hardly reflect the total 
criminal activity in any locality. Second, while what crimes are re-
ported are often interpreted as an index of the "true i:!rime rate," the 
relationship between this index and the true crime rate is not constant 
but subjec-t; to variation. Third, differential interpretations of any 
specific violation among differing jurisdictions as well as differing 
interpretations through time in any one jurisdiction make comparisons 
quite difficul?t. Fourth, since crime statistics are often compiled for 
administrative purposes, they are often biased by political and budge-
tary considerations. Fifth, some crimes, for example, white collar 
crimes, are not routinely compiled. Sixth, q_uite often nebulous defin-
itions, as in the case of juvenile delinquency, further compound inter-
pretations of these data. Finally, sane social characteristics of the 
offenders themselves lead to differential treatment by the regulatory 
authorities. For example, the possibility of actually being arraigned 
for a violation is inversely related to the offender's social class 
position. 
Are these criticisms of paramount importance to this study? The 
first four criticisms have in common that they arc most relevant when 
28 
a study encompasses a considerable time element . The data utilized in 
this thesis covers a total time period of fifteen contiguous days, and 
the effects of changes through time must therefore be considered as 
minimal. Whi.Le tht: first criticism is applicable, the problematic in 
this study is not the true crine rate but the degree and nature of the 
differences in what is reportc:i. We are willing to grant that bias 
exists, but we argue further t.1at the bi.as should be operative equally 
on the police statistics for t.1e time period covered with the exception 
of the effects of the riot itself, which is the principle concern here-
in. Of course, differential interpretations of the law between politi-
cal jurisdictions is iri?elevan~ here. The fifth and sixth criticisms 
are also irrelevant, becnu3e the white collar crimes are compiled by 
the police in Washingt,on, and juvenile data are not a part of these 
data. The final criticism i3 i?elevant, but again, because of the short 
span of time investigated, the data should be equally biased. 
The a oove, then, leads to a methodological assuinption of extrema 
importance. While we do assume some bias in these data, we further as-
sume a homogeneous distort.ion Jrom the true values. This makes the 
normal time period an absolute necessity as a benchmark to compare the 
riot period against, because the comparative differences between these 
two time periods should then reflect the real differenc es. 
While the above criticisms are relevant to official criminal sta-
tistics, Kitsuse and Cicourel present a number of criticisms relative 
to any official statistics.::, u,uoting Merton, these authors write: 
There is little in tlie history of how statistical 
series on the incidence of juvenile delinquency came to 
be collecteJ that shows t.hem to be the result of ef-
forts to identify either the sources or the contexts of 
juvenile delinquency. ThEse are social bookkeeping data. 
And it would be a hap[JY ~oincidence if some of them 
29 
turned out to be i n a f orm relevant for research . 
From the sociological standpoint, 'juvenile de-
linquency' and what i t encompasses is a form of devi-
ant behavior for which the epidemiological data, as 
it were, may not be at hand. You may have to go out 
and collect your own appropriately or ganized data 
rather t han to t ake t hose wliich are r eady-made by 
eovernmental agencies .6 
Kitsuse and Cicourel interpr et Mer ton to mean that quite often official 
statistics ar e not in a form which is suitable f or sociological re-
search. 7 It may be argued t hat while the police data utilized in this 
thesis are classified i nto l egally defined categor ies, t hese categories 
are quite amenable to reformuli zation i nto socia lly meaningf ul units 
of analys is. The very deta i led classification of location, time,, and 
t ype of viola tion, while too cumbersome f or sociologi cal analysis, is 
mani pula table into mor e gener al categories . 
The authors continue: 
Mert on also ar gues agains t t he use of of f i cial 
s tatis tics on quite di fferent grounds. He states t ha t 
s uch dat a are "unreliable" because "successive l ayers 
of error intervene between the actual event and t he r e-
corded event, between the actual rates of deviant behavior 
and the records of deviant behavior ." In t hi s s t ate-
ment, the argwnent is that the statistics are unreli able 
because some i ndividuals who mani fest deviant behavior 
are apprehended~ classified and duly recorded whi le 
others are not . 
And, 
From this point of view, deviant behavior is behavior 
which is organizationall y defined, processed, and t reated 
as "strange, " "abnormal,n "theft," "delinquent," etc. , 
by the pers onnel in the socia l system which has produced 
the rate . 9 
The argument here is that often the i ncidence of "deviant behavior " is 
defined as such by the personnel within the agency performi ng t he 
classifi cation. In a word, these statistics te l l t he res earcher more 
about the people making the classifi ca t ion than the people being 
... 
30 
classified. Since the focus of this study is upon the police as one 
agency of social control, then this particular criticism is irrelevant 
in the present context. 
But there are a number of positive aspects to the use of official 
data. One of the basic factors which delineates riot situations from 
normal situa ?tions is the general breakdown or change in the normative 
structure of the social system. Looting and burning of property and 
?the imposition of emergency measures are frequently found during civil 
disturbances. This is not to say that riots necessarily generate anomic 
behavior, only that they establish a normative pattern which differs 
from a normal tLme period. Hence, as investigators of riots, we are in 
part interested in what kinds of behavior are manifested during riots 
and what kinds of responses are made to this behavior. Further, 3ince 
part of the behavior manifested in riots is in violation of the law, 
the police may be assumed to be the one formal agency most familiar 
with this aspect of the riot and the one agency interested enough to 
collect such data. Apart from the validity of police statistics, the 
offense and arrest records provide the best single source of information 
available concerning crime during a civil disturbance. 
Amongst the available data on riots are a variety of reports by 
various agencies within the community. While various agencies within 
the District of Columbia provide infor mation, these reports all have 
in common a lack of systematic presentation. For example, the District 
of Columbia Department of Public Health surveyed the total casualties, 
tot.al admissions, and total deaths reported by eleven hospital facili-
t?i. es i.n  Washi.n g t on . 10 While this information is reported on a day t;o 
day basis for the five days from April 8th through April 12th, the 
; l 
firsi:, Lhrt::e full days of the riot, J\pril 5th through 11pril 7th, are 
collapsed into one tabular preseuta tion. On the other h~nd, the Dis-
trict oC Columbia Metropolitan Police Department data are very system-
atic. The units of reportage are consistent and small so tha 1; 
meaningful combinations of time periods or violations can be easily 
accomplished . 
While a field observer may attain a qualitative impression which 
ccinnot be gleaned from quantitative secondary analysis, the latter does 
provide the investigator with quantification . A perusal of the riot 
literature indicates that riot situations are very definitely quali-
tatively different from non-riot periods . Again, we mention the inci.-
dence of looting and burning as obvious differences. But very little 
of the l iterature indicates any quantification of these behaviors. As 
an instance of collective behavior and as an instance of, at best, 
differing normative patterns, it is understandable wlzy quantified data 
arc difficult to obtain. For this very reason, then, police statistics 
are valuable. After the researcher has distinguished qualitative dif-
ferences between riots and non-riots, it becomes necessary to specify 
how much of a change is manifested from one situation to the other. 
Finally, systematic arid quantified data are absolutely crucial 
for the fourth characteristic of police data--comparability. One of 
the ma jor purposes of this study is to quantitatively describe the dif-
ferences in officially conceived criminal activity between a riot period 
and a non-riot or normal period. One can only make cQmparisons in dif-
fering time periods from actual data if the data are quantified and if 
they are reported in a constant fashion. That the data are quantified 
is obvious, but what about consistency of classification? The police 
32 
classification is ba,.Nd on the legal definitions of criminal violations 
as stated in the District of Columbia Code and is, ?therefore, irrele-
vant to the particular characteristics of the social situation. 11 The 
preceding statement must be modified somewhat, when a new ordinance is 
enacted specifically for the existing situation, as was done during t he 
riot. A curfew was placed on Washington during the riot period and as 
such there was no provision in the classificatory schema for this vio-
lation. Therefore, all arrests for curfew violations were placed in 
the disorderly conduct category. Further, there is no ordinance against 
looting. The criminal code contains laws against burglary and against 
rioting as an offense. Therefore, all looters arrested and all looting 
offenses reported to the police were classified as burglaries. With 
the exceptions noted above, all other violations were classified in the 
12 
same categories prior to and during the riot.
In short, the police and fire data maximize quantification, system-
atiz.ation, and comparability within the limits specified by the sub-
stantive conc.:erns of ?this research. Of the existing sources of data, 
these are the best. This is a very important consideration. Riots 
contain elements of deviant as well as collective behavior, which ex-
acerbates the validity problem. The issue essentially is a very simple 
one: either we use these data recognizing their limitations or we do 
not research these aspects of riot situations. 
Concept,ualizing the Data: A Problem in Simplification 
The District of Columbia Metropolitan Police Department identifies 
any violation with a four digit code. For example, first degr?e murder 
is identified with the number 0101. Since every legally differentiatable 
jj 
erirne is designated with its own code number, the total number of e odcs 
i s 497, 13 which obviously creates a need for some kind of simplificati on 
of these data. This information can be generalized into any one of 
thirty more general categories . For example, first degree murder, 
second degree murder, manslaughter, negligent homicide, and homic ide 
are all placed in the more general category of homicide or the 0100 
series of codes. Again, robbery with no weapon (OJOO), robbery with a 
weapon (0310), and a ttempted robbery (OJOl) are all placed in the 
general category of robbery or the OJOO seri es of codes. 
The most specific reportage of violations herein is presented in 
terms of the thirty-three more general ca t egories . 14 In the analysis 
of a c ivil disturbance, it is important to know how many aggravated as-
saults occurred and relatively immaterial from a social point of view 
as to the pcirticular material object used in such an assault. There is 
one exception to the above statement. Because of the high incidence of 
fires and burning during the riot, false fire alarms (code 2649) have 
been separated from the general 2600 series of "other" violations. 
Table l indicates the general codes and the substantive violations. 
Classification of the Police and fire data. Figure 1 presents a 
more socially meaningful categorization of the data, conceptualized 
into six general categories: (1) crimes against per.s ons, (2) crimes 
against property, (3) trciffic violations, (!+) crimes without victims, 
(5) crimes related to fires, and (6) miscellaneous crimes. Crime s 
against persons are those violations which result in direct physical 
15 
harm to a victim. Crimes against property involve no such personal 
injury or threat of personal injury to one's physica 1 being. Traffic 
violations have been placed in a separate category because of the very 
34 
large volume of these violations. Crimes without victims are 
here used to indicate violations which involve the offender and 
which rarely, if ever, entail a vi.ctim directly. One might 
argue that the chronic alcoholic, who is continually on the 
police blotter for drunkenness, is performing behavior which 
is injurious to the welfare of his fami.ly. While this notion is 
readily accepted here, one may argue further that alcoholism 
does not involve the same kind of direct contact with a victim 
as would a homicide or assault. Nor does the alcoholic transgress 
directly upon some victim's property as in larceny or auto the f t. 
Crimes without victims, then, in this research refer to those 
violations in which the offender performs behavior unrelated to 
property or to persons who are unwillingly involved in the 
violation. 
Arson and false fire alarms obviously belong in the fire-
related crime category, but how can the incidence of fires 
reported to the fire department be so placed? While reporting 
instances of fire can not be conceived as criminality, such 
reports do generate actions on the part of the fire department 
in much the same way that ?the police must respond to a ca 11 for 
assistance. It is for this reason, coupled with the obvious 
substantive similarity, that the fire statistics are placed in 
this category. 
Finally, the remaining violations are lumped together in 
the miscellaneous category for the simple reason that these vio-
lations are irrelevant or appear to be irrelevant to riot 
behavior. Since the looting of liquor stores is so prevalent in 
35 
most civil disturbances, one might question why liquor law violations 
have been classified in this categury. Since the liquor laws involve 
the regulation of the manufacture, distribution, sale, and taxation of 
alcoholic beverages, these violations are pertinent to those individuals 
whose business interests are found in the area of alcohol rather th;m 
those individuals who avail themselves of the product. 
In sum, at this point, the 49? specific criminal violations have 
b-=?en reduced to thirty-three generalized headings and classified into 
one of six mutually exclusive general categories, as reported in Figure 
1. It, will be necessary to look within each one of these six gerieral 
categories for a more detailed descriptive analysis. But S:)Jne of these 
general categories contain a relatively large number of generalized 
headings. For example, the general category of crimes against persons 
contains eight headings and the miscellaneous category contains ten 
headings. Therefore, a further simplification of these generalized 
headings within the six general categories would be useful, and it is 
to this that we now turn. 
Crimes against persons seem to fall naturally into a trichotomy: 
(1) crimes which carry a future potential for violence, (2) crimes 
which carry an actual threat of violence, and (3) crimes in which vio-
lence or force is really exercised. First, weapons violations, which 
involve possession, sale to a minor, and unlawful sale, are placed in 
the s ubcategory of future potential for violence, because they involve 
a material object which could be used against a person but which at the 
time of the charge there is no evidence that it has been so used. 
Therefore , weapons violations connote possession of' an object which 
could lead to the second and third subcategories of crimes against 
36 
persons. Second, robbery has been placed in the astual threat of vio-
lence subcategory, because it involves theft accompanied by violence 
or imposing fear of violence in the relationship between the offender 
and the vic?~im. 16 From the def lnition just given, robllery might have 
been placed in the third subcategory of really exercised violence. It 
has been kept separate, because it is not possible in the data to dis-
tinguish those robberies involving threat of violence from th:)se in-
volving real violence. Also, it would logically appear that the main 
purpose of a robbery is to secure the victim's property, while the 
violations classified in the real violence exercised category involve 
only personal bodily harm. But this consideration implies that rob-
beries should be classified as crimes against property since they in-
volve theft . As such, a robbery could be placed in either category. 
IL is retained in the crimes aga inst persons division because of the 
assumed greater concern or priority with physical harm than with the 
loss of property. Third, homicide, rape, aggravated assault, other 
assaults, sex offenses, and offenses against the family and children 
are placed in the actual force exercised subcategory for obvious 
reasons. Parenthetically, all sex offenses involve some real sexual 
action and most all family and child offenses involve behavior which 
proct11ces directly or indirectly physical trauma. 
The general category of crimes against property has not been col-
lapsed, because these crimes are all substantively different from one 
another with the obvious exception that they involve property viola-
tions . Burglary-house breaking i nvolves unauthorized entrance and 
actual or intended theft. Lclrceny entails only theft and is separated 
from auto theft, which involves theft of a motor vehicle. If some 
J? 
item on or within a motor vehicle is stolen, this crime would be clas-
sified as a larceny. Therefore, auto theft refers only to the actual 
unauthor ized use of the motor vehicle itself. Stolen property applies 
basically to possession of pilfered goods. For example, if one indi-
vidual removed an article by shop lifting, he would be charged with 
larceny. If the stolen article wa s given to a ,,econd person, who was 
not involved in the action of taking the article, this second person 
would be charged with receiving stolen property. Finally, vandalism 
entails primarily destruction of another's property. Again, all the 
crimes against property involve no real or threatened physical harm to 
the owner of the property. 
The third general category of traffic violations has been collap-
sed into two subcategories of moving and equipment violations. The 
traffic series (JJOO codes) and the other traffic series (3500 codes) 
are synonymous and separated in the police stat:i.stics only because the 
number of possible moving violations exceed 100. Intoxicated driving 
is placed with traffic and other traffic violations because, first ?' it 
is a moving violation, and second, because of the minimal number of ar-
rests in this category, this violation is proportionate]y rather unim-
portant. Equipment violations are placed in a second separate 
subcategory, because these deal wi.th mechanical malfunctions of the 
automobile rather than human errors in the operation of a motor vehicle. 
Again, recall that non-moving violations (for example, illegal parking) 
are not considered important enough to be included in the District of 
Columbia Metropolitan Police Department statistics. In fact, while 
police officers write parking tickets, the police department's respon-
sibility ends with the issuance of the violation, 
W.ithin the general category c,f crimes without victims, prosth,u-
tion, narr::otics violations, and gambling offenses are generi:.illy ones 
whlcn are not logically related to riot situc ?~lons and have been placed 
in thr~ non-riot related subcategory. This logical assumption seem::; to 
be supported by the data which indicate a fairly constant number of 
these violations between thP. normal and the riot periods studied. 
Drunkenness is placed in a separate catebory because of the very high 
incidence of liquor store looting. Finally, since the DistriGt of Co-
lumbia Metropolitan Police Department classification contains no curfew 
violation category at the time of the 1%8 riots, all curfew violations 
were placed in the disorderly conduct category. Since curfe,r violatlonu 
are peculiar to civil disturbances, the disorderly conduct category is 
trea Led separately. It could be strongly argued that one sho11ld E,epar-
a-Le the curfew violations from the normal disorderly conduct violations. 
Table 2 indfoates the unofficial police stcltistics for these violations, 
Under the assumption that curfew violations were classified as disorder-
ly conduct violations, the reader will note thc:Jt these statistics are 
all possibilities with the exception of April 8th and 1\pril 9th. On 
both of these dclys, there were purportedly more arrests for curfew vio-
lations than reported disorderly conducts. Since i.t is not possible to 
verify which source is inaccurate, we assume the official statistics to 
be more representati.v e than the unof f1' c1. a 1 ones . 17 
The general category of crimes related to fires has not been col-
lapsed, because tbe violations within this category are obviously dif-
ferent. Finally, the miscella neous general category contains all the 
remaining crimes . It muy be argued that since fraud, forgery, and em-
bezzlement i uvolve theft, these violations are best treuted as crimes 
39 
against property. They have been placed in the subcategory of fraud, 
because they are thefts which suggest a more intellectual or mental 
manipulation when compared to larceny and burglary. In other vords, 
these thefts are ones which do not admit force or violence as does a 
robbery. Nor do they admit breaking and entering as does b1Jrglary-
house breaking. Further, larceny in general fa treated as qualitative-
ly different from these violations. 
Larceny-theft is the unlawful taking or stealing of 
property or articles of value without the use of force 
or violence or fraud ???? In the Uniform Crime Re-
porting Program this crime category does not include 
embezzlement, "con" games, forgery, and worthless 
checks. 18 
For reasons cited earlier, liquor laws are included in this category. 
other arrests ., vagrancy, and suspicion were grouped because of their 
similar number in the normal as well as the riot period and because of 
their irrelevance to any of the other miscellaneous categories and 
identified with the label "varied." Finally, all the "unknown" viola-
tions were grouped together. These latter violations appear to be 
basically miscoding errors in the data processing process within the 
Police department itself. 
In sum, the thirty-three generalized headings have been further 
collapsed ~nd grouped into six general categories, as indicated in 
Figure 2. This figure indicates the substantive criminal categoriza-
tions which are to be used in the interpretation of this aspect of the 
data. 
Definition of the riot period of or~anized response and a comparable 
n,ormative time period. As we argued in the preceding chapter, the tlrne 
Phases of unorganized response and organized response begin after the 
impact tim? period has started. It will be necessary to present a 
39 
aga inst property. They have been placed i n t he s ubca t egory of fra ud, 
because they are t hefts which s ugges t a more intell ect ual or mental 
mani pul ation when compared to l arceny and bur glary . In other wor ds, 
t hese thefts are ones which do not admit forc e or vi olence as does a 
robbery. Nor do they actmi?t breaking and entering as does b1Jrglary-
hous e breaking. Further, larceny in general is treated as qua litative-
ly different from these violations. 
Larceny-theft i s the unlawful taking or s t ealing of 
property or articles of value withou?~ t he use of forc e 
or violence or fraud ???? In the Uniform Crime Re-
porting Program this crime category does not include 
embezzlement, "con" games ., forgery, and worthless 
checks. 18 
For reasons cited earlier, liquor laws are included in this ca t egory. 
other arrests, vagrancy, and suspicion were grouped becaus e of t heir 
similar number in the normal as well as the riot period and because of 
their irrelevance to any of the other miscellaneous categori es and 
identi fied with the label "varied." Finally, all the "unknown" viola-
tions were grouped together. These latter violations appear to be 
basically miscoding errors in the data processing process within the 
Police department itself. 
In swn, the thirty-three generalized headings have been further 
collapsed ~nd grouped into six general categories, as indicated in 
Figure 2. This figure indicates the substantive criminal cat egoriza-
tions which are to be used in the interpretation of this aspect of the 
data. 
Definition of the riot period of organized response and a comparable 
normative time period. As we argued in the preceding chapter, the tlme 
Phases of unorganized response and organized response begin after the 
impact time period has s t arted. It will be necessary to present a 
40 
brief chronology of the beginning of the impact stage of the Washington 
riot so that we may select an appropriate beginning of the ped.od of 
organized response. 
At approximately 6:20 P.M. on April 4th, 1968, Martin Luther King, 
Jr., was shot in Memphis, Tennessee. At 7:JO P.M., WTTG-TV, a local 
Washington televisiou station announced the shooting and reported that 
Ki. ng was in serious cond1? t?io n. 19 Dr ? K1. ng wa s subsequent 1y  pronounced 
dead at 8:06 P.M. 20 In Washington, the news brought a crowd of several 
hundred persons to 14th and U Streets, N. W.--an intersection which is 
one of the major public transportation hubs in the District of Colum-
bia. At 9:25 P.M., the first window breaking at the Peoples Drug Store 
next to the Southern Christian Leadership Conference offices at 14th 
and lT Streets, N. w., was heard, and the Washington riot began. 21 At 
12:JO A.M., April 5th, the first full scale fires were set on 14th 
Street, although the first fire call was received about an hour and a 
half earlier. Fire Chief Henry Galotta instituted Plan F, which split 
the existing fire companies in two, thereby doubling the response capa-
bility of the District of Columbia Fire Department at about 11:00 P.M., 
April 4th. 22 At 1:20 A.M., April 5th, the police asked and were granted 
23 
Permission to use tear gas on 14th Street. By J:00 A.M., the last 
major confrontation on 14th Street was over, and 200 stores had windows 
broken with looting occurring in 150 of them. Seven fires had occurred 
and 200 people had been arrested with thirty people injured, including 
f i.v e policemen and one r i?r eman. 24 
While the first indicators of the beginning of the impact period 
were at 9:25 P.M., the fire department did not switch to Plan F until 
approximately 11:00 P.M., and the police department did not begin to 
41 
implen,ent riot control procedures untiJ 1 :20 A.M. , April 5th, when per-
mission was granted for t he use of tear gas. With the fire department 
activating its preplanned civil disturbance routine about ,:me hour b e-
f ore midnight and the police department beginnine to use tear gas a 
lit,tle over one hour after midnight, the riot period of organ ized res -
ponse is operationally defined as beginning on the day following the 
assassination of Martin Luther Kine, Jr . and ending on the beginning of 
the day following the lifting of the curfew against the free mover,ient 
of persons within the District of Columbia. It is clear that the lift-
ir1g of the curfew indicated to the public officials of the District of 
Columbia that the immedi.ate danger of the impact period had past. In 
terms of time , then, the riot period of organi zed response is oper,.ition-
ally designated as beginning at 12 :01 A.M., April 5th and ending at 
12:00 P.M., April 12th. Hence, this time period covers eight full days. 
Once the organized response time period has been defined, one must 
make a variety of decisions relative to a comparable normal time period. 
While the annual report of the Federal Bureau of Investigation is c on-
cerned only with a limited number of variables in its analysis of crime, 
t?h i. s repor t i.
25 
n di. ca t es month ly f 1 uc t ua t?.i ons i. n se 1 ec t ea'  cri.m es. Sta-
tistics are also presented indicating variations in the number of slain 
26 27 police officers by day of the week and by hour of the day. The an-
nual r eport of the Dist rict of Columbia Metropolitan Police Depar?tment 
shows variations in offenses reported to the police by month, 28 by day 
of the week, 29 and by hour of the day. JO Similar fluctuations occur 
w"i th arres~.  d a t a. Jl 
The incidence of major crimes was greate3t on week-
ends ?.?? The peak was on Saturday, declining Sunday 
through Tuesday, and beginnine to cli~b toward the Sat-
urday peak. 
Generally, the daily incidence of serious crime 
was highest during the hours from 8 P.M. to midnight, 
when approximately 70 percent [ sic ] of the daily of-
fcnses wer? cleared by the police ? 32 
Comparisons of crime statistica for differing tii11e periods require 
c3reful attention to congruity of time of the year, day of the week, 
and hour of the day. Because of seasonal fluctuations in cr~ne statis-
tics, one must select a comparable time period very close to the selec-
ted riot period. The seven days prior to the selected riot period have 
been designated as the normal time period, because this period mini-
mizes seasonal fluctuations, differing inte:cpretations of law enforce-
ment thr ough time, and possible changes in law enforcement subsequent 
to the riot which might have resulted from the civil disturbance. 
Hence it was possible to compare congruent days and hours of the day, 
holding as constant as possible seasonal factors. 
One further assumption of significance was made. Since the selec-
ted riot period covered eight days, beginning on a Friday, this time 
span contains data for two Fridays (April 5th and April 12th). There-
fore, the canparablc normal t.ime period must also contain eight days of 
data. The first Friday of the selected riot period (April 5th) has 
been compared with Friday, March 29th or the day seven days previous. 
If the second Friday during the riot was compared to the day one week 
earlier comparisons would be be~ween one day at the beginning of the 
selected riot period and another day at the end of that same period. 
Therefore, the second riot Friday was compared to Friday, March 29th. 
In other words, the normal time period has been operationally defined 
as March 29th through April 4th with the first day of this period cnwn-
erated twice to compose a normal time period of eight days. If the nor-
mal time period had been defined as eight chronological days prior to 
43 
the stlected riot period, comparisons would have been betvreen differ-
ent days of the week. If the second Friday of the riot period was 
compared to the second Friday prior to the period of organized response, 
there would have been a time differential of three intervening weeks. 
Neither one of these alternatives appeared logical in view of the 
known fluctuations in crime statistics. 
Definition of the riot, corridor, and non-riot spatial areas. A 
second major concern of this thesis is a geographic analysis of the 
distribution of crime between the two defined time periods, The Na-
tional Capital Planning Commission and the District of Columbia Rede-
velopment Land Aeency made field surveys of the structural damage and 
classified this damage into three categories: (1) slight damage: 
0-10 per cent; (2) substantial damage: 11-50 per cent; and (3) ex-
tensive damage: over 50 per cent. 33 Since these estimates dealt only 
with building condition or structural damage, it has been as3urned that 
even in the 0-10 per 0ent category of window breaking and entering, 
there was probably rather heavy looting of inventories within the struc-
tures. Further, this report contains block maps of the major riot 
areas in which each structure is color coded into one of the three dam-
age categories noted above. Therefore, the reader is able to determine 
specifically the spatial distribution of structural damage for each of 
the major riot areas. One further source presents a complete map of 
Washington with the incidences of structural damage identically 
treatect. 34 
Both reports agree that the major riot areas were 14th Street, 
N. W.; 7th Street, N. w.; and H Street, N. E. Since the 8th Street, 
S. E. riot damage was similarly concentrated (although not as great in 
dollar value), this researcher has interpreted this area as a fourth 
major area of destructiou. All four of these areas have been collapsed 
l.nto what is herein conceptually defined as the "riot areas.'' A per-
usal of the damage maps indicates other areas of lesser concentration 
of damage along several major transportation routes within the Dis-
trict of Columbia: Rhode Island Avenue, N. w. and N. E.; Benning Road, 
N. E. ands. E.; and Good Hope Road, S. E. These areas of sporadic 
damage have been collapsed into what is conceptually defined as th~ 
If ? ? 
corridor areas." All the remaining areas within the politically de-
fined area of the District of Columbia have been conceptually defined 
as outside the areas of riot activity and identified as "non-riot 
areas," 
The four riot areas and the three corridor areas were operation-
ally defined by including two blocks of space as one moves through the 
specific blocks and streets of damaged areas as defined by the struc-
tural damage reports. A radius of two blocks was selected, because it 
'Was assumed that the general area of confusion, smoke, tear gas, and 
other riot characteristics could have been sensorily perceived by any 
Person within that area. While this delineation seems somewhat arbi-
trary, and is, it also appears fairly logical. This is to say that 
the Physical characteristics of a civil disturbance can not be as-
sumed to end thirty feet from the center line of the major road of 
activity. But at the same time, these cha~acteristics can not extend 
indefinitely into space. 
Given these operational boundaries, all streets and their block 
numbers were established using a street address map of Washington, 
D. c. 35 Each offense, arrest, and fire report was locationally 
classified and collapsed into one of the above mentioned three spatial 
ca -.i ;egor.l .es. 36 
Statistical Analysis of the Data 
Tabular presentation of the data. Each table contains three di-
mensions. First, the time dDnension has been dichotanized into the 
"riot" period of organized response and the "normal" period of a com-
parable span of tDne. Second, the substantive category of criminal 
violation contains either the six general categories, as noted in 
Figure 1, or the collapsed subcategories within any one of these six 
general categories, as noted in Figure 2. Finally, the spatial dDnen-
sion, trichotanized into riot, corridor, and non-riot areas, is held 
constant within any one table. Each table reports three types of 
descriptive summary statistical data: (1) the frequency of items fal-
ling into each cell, (2) ~he percentage of the number of items in a 
specific cell computed from the total number of items contained within 
the specified time period, and (J) the percentage change from the 
normal to the riot time period. 
The tableG are presented for analysis according to the following 
guidelines. The offense and fire data appear first, followed by t,he 
arrest data. The offense and fire data are ordered by pre3enting 
these data for the riot, C!orridor, and non-riot spatial areas, re-
spectively. Finally, these data are ordered further by present:i.ng the 
six general categories first and following with the more detailed col-
lapsed s ubcategorizations. The arrest data is presented according to 
the same guidelines as posit?d for the offense and fire data. 
Tlle l~mbda statistic. The chi- square statistical test may be 
used to test differences between two or more independent distributlons 
under the null .hypothesis of no differences. Should a "statistic'1.lly 
significant" r esult occur, t he null .hypothesis would be rejected, and 
ote would fail to reject t he alternative hypothesis of difference. 
While the chi-square t est would be t he traditional choice for a con-
tingency table analysis wher e the data are reported at the naninal 
le?1el of measurement, the syrrunetrical, proportionate reduction in error 
measure of lambda has been selected. Hays remarks: 
When the value of chi square turns out significant one 
can say with confidence t hat the attributes A and B 
are not independent. Nevertheless, the significance l evel 
alone tells almost nothing about the strength of t he as-
sociation ???? Statistical relations so small as to 
be almost rwnexistent can show up as highly significant 
chi square results, and this i s especially likely t0 occur 
when sample size is large ???? The lambda indices do, 
however, suggest Just how much the relationship found 
implies about real predictions, and how much one attribute 
actually does tell us about the other. Such il1dices 
are a most important corrective to the experimenter's 
tendency to confuse statis?t;ical significance with the 
importance of results for actual prediction. Virtual l y 
any statistical relation will show up as hi ghly sig-
nificant; given a sufficient sample size, but it takes 
a relation of considerable strength to enhance our 
ability to predict in real, uncontrolled, situations. 37  
Chi square's descrip?~ive interpretation, then, i s limited to one of 
association and neglects the strength or degree of as socia tion pre-
sent. It is also very much influenced by the size of the sample and 
also by the number of expected frequencies per cell. 38 Finally, and 
probably most important of all, is the te:ndency to interchange sta-
tis t. ical significance with substantive significa nce. Hays continues: 
On the other hand, chi-square tests are always approx-
imate, and the evidence at hand suggests that the good-
ness of the approximation varies with a mnnber of 
factors, not all of which can be taken into acc ount in 
a simple rule of thumb.39 
Blalock remarks: 
A nun,ber of other measures of assocfation which 
can be used with contingE::ncy table~ have been presented 
by Goodman and Kruskal. Most of these measures, only 
une of wbich will be discussed in this text, involve 
what have been referred to as probabilistic interpre-
tations. Since they have an intuitive meaning enabling 
one to interpret values intermedia t.e between zero arid 
one, these measures would seem to be superior to those 
based on chi square,40 
For these and other reasons, chi-square has been disregarded in favor 
of lambda. Hays continues: 
One of the oldest problems in descriptive statis-
tics is that of indexing the strength of statisti~al 
associntion between qualita?t.ive attributes. 
"???tt????????e,.??????11111"??1'! 
As we have seen, most of our notions of the strength of 
a statistical assor!iation rest on the concept of the 
variance of a random variable ???. However, when the 
independent and dependent variables are each categorical 
in nature, ?the variance p?1' se is not defined, Something 
else must be used in specifying h~w knowledge of the A 
category to which an observation belongs increases our 
ability t.o predict the B category. 
Three somewhat different approaches to this problem 
will be discussed here. The first rest.s directly on the 
iJOtion of statistical independence between two attributes, 
defined as p (Aj' Bk) -= p (A.) p (Bk). In this approach, 
the strength of association !s basically in terms of the 
difference the extent to which the probability of a joint 
occurrence differs from the probability that would be true 
if the attributes A and B were independent ???? 
Another and much more recent approach deals with 
,m:edictiye association. Association between categorical 
attributes is indexed by the reduction in the probability 
of error in prediction afforded by knowing the status of 
the indi victual on one of the attribures. This way of 
defining association makes intuitive good sense, but is 
not as directly tied to tests of association as the first 
approach. 41 
The proportionate reduc ti.on in error measures, then, may be used in 
tests of association. Since the value of a PRE measure falls some-
where between zero and one, it facilit~tes interpretation. A value 
close to one may be interpreted as evidencing a strong association 
48 
between the variables, while a value close to zero would suggest a 
weak associatiorJ. While the lower limit of chi-square is zero, it 
contains a variable upper limit, c:Jnd this latter characteristic makes 
the interpretation of the chi-square statistic difficult, especially 
when one considers the variety of factors which mcJy alter this statis-
tic . Hays continues: 
Finally, in some contexts it may be desirable to 
have a syrr?netric measure of the power to predict, where 
neither A nor Bis specially designated as the thing 
predicted from or known first. Rather, we act as though 
.sometimes the J\ and sometimes the B informc:1tion is given 
beforehand. 42 
While most _[-JRE measures assume the direction of prediction, lambdc.1 
itself does not. Since the basic concern of this research is to test 
the hypothesis of association between normal and riot periods in terms 
of r]riminal violations ., the direction of the prediction is unimportant. 
In other words, it is herein immaterial whether the violation is used 
to predict the time period or the time period is used to predict the 
violation. Should this be our desire, the former may be assessed 
using lambda band the latter using .lambda a. Since the value of 
lambda falls between lambda a and lambda b, lambda itself is a more 
conservoti.ve measure of both lambda a and lambda b. If one is not 
s?Jre what his directional assessment should be, then the synunetric 
measure is the better one. This is to say that for consistency's sake, 
lambda has been chosen rather than fluctuating back and forth between 
lambda a and lambda b. 
Hays concludes: 
In the light of its somewhat complicated statistical 
character, a significant Pearson chi-square test may 
mean next to nothing, but an apparent predictive re-
lationship in the data is usually ?.1orth looking into. 1t3 
49 
In sum, if tha lambda measures indicate a predictive relationship, a 
small value can be interpreted as a minimal. reduction in one's error 
Prediction, with the knowledge of how any one variable is distributed. 
Therefore, a small value indicates little association or a minimal 
amount of difference. 
Finally, any of the lambda measures are useful in contingency 
tabular analysis because of the leniency of the assumptions under-
lying a proper meaningful interpretation of the statistic. 44 One must 
have two "pol.ytomies " or classifications, in which there is no under-
lying continua or natural ordering of interest. In the case of lambda, 
as opposed to lambda a and lambda b, one does not have to assume asym-
nietry or that one classification precedes the other causal.ly., chrono-
logically, or in any other way. While it may be argued that some of 
the collapsed tabular presentations show ordina l properties (for ex-
ampl~, the crimes against persons category), other's do not. Given 
the nature of the data collected and the interpretation desired in 
the analysis of these da ta, lambda seems the most appropriate statis-
tic . 
One final topic remains relative to lambda--its utilization with-
in this study. If lambda is used as a measure of the association be-
tween the riot period of organi zed response and a comparable normal 
time s pan, it is necessary to specify certain ranges in the lambda 
Values for interpretive purposes ., and four s uch ordinal ranges are 
POsitect : (1) minimal association ('t;be computed lambda value falls 
between 0.0000 and 0.1000), (2) weak association (the cooiputed lambda 
value falls between 0.1001 and 0,2000), (J) moderate association (the 
computed lambda value falls between 0.2001 and 0,5000), and (4) strong 
50 
association (the computed lambda value falls between 0.5001 and 1.0000) . 
While such an ordinal categorization of the lambda values is an arbi-
trary one, it is governed by one very llnportant property of this sta-
tistic. Lambda is a weak measure of association. This means two 
things relative to its interpretation. First ., a relatively low lambda 
value indicates more association in the data than a similarly low va l-
ue i11 a more powerful measure of association. And second, because of 
this first consideration, the cat,egorizations of possible lambda val-
ues have been divided more toward the zero end of the continuum rather 
than being eg_ually spaced throughout the entire possible range of this 
statistic. 
To summarize the statistical analysis of the data ., the lambda 
values will be used to evaluate the degree of association between the 
selected normal and riot tllne periods and as such are restricted to 
the evaluation of the hypotheses presented in the preceding chapter. 
The descriptive statistics are used to indicate the nature of the dif-
ferences in the cr1minal violations ., thereby facilitating the selec-
tion of the indicators. 
summary 
There are a number of crucial methodological questions with 
which we have dealt i n this chapter and which must be answered before 
the degree and type of differences between the normal and riot time 
Periods selected for analysis can be assessed, apart from the ~bvious 
questions relative to the ecological variables of time and space. 
First, what data will be used? As indicated above, the data must be 
continuously collected so that the normal time period prior to the 
51 
:ciot can be used as a benchmark or control against which to compare 
the riot d11ta. Second, are these data recorded in a form which is 
socially meaningful? While the violations are reported in a form 
which best corr~sponds to ?t;he legal definitions of crimes, these cate-
gories may be easily simplified into many fewer socially meaningful 
categories. The additional data dimensions of time and location are 
sufficiently precise that, again, meaningful simplifications are pos-
sible. Third and finally, are these data valid? While these data 
are, surely, not without some bias, they are utilized because in view 
of 'the other requirements (the nature of the questions asked, the 
focus on the police department, the need for continuously cc,llected 
data, which can be manipulated into meaningful categories), they are 
the most valid data of that which is available. As instances of col-
lective behavior, riots are atypical situations, and the student of 
these occurrences can not expect the same degree of sophistication in 
data collection which we have cane to expect under normal circum-
stances. 
CHAPTER IV 
THE RESULTS 
Introduction 
In this chapter the results of the data analysis a re pre::;ented 
c1n,i discussed. There are four major sections which encompass the dis-
cussion of (1) the offense and fire data, (2) the arrest data, (3) a 
'.!Dmparison of the offense and fire data with the arrest data, and 
(4) a brief s ummary indicating those violations which are indicative 
of riots. Each of these first three mojor sections is subdivided into 
five pc1rts which include the discussion, respectively, of (1) the riot 
area data, (2) the corridor area data, (J) the non-riot area dc.Jtc1, 
(4) the comparison of the three spc1tial areas, and (5) the evoluation 
of the relevant hypothesis. Further, within ony one of the spatial 
areas discussed, the tubular presentations are ordered beginning with 
the classification by general category, as indicated in Figure 1, and 
then proceeding to the collapsed subclassifications of the data, as 
indicated in Figure 2. The particular data are analyzed by type of 
violc.i tion and by time period in terms of both number and percentage 
of the specific category of violation to the total number of viola-
tions within the specified time period. Finally, each spatial sub-
section concludes with a listing of the major findings of thc1t sec-
t~ion. 
The analysis of the first three subsections of the .spc1tial 
variable is descriptive, showing which criminal categories change 
52 
5] 
and which do not as well as the magnitude of the change between the 
Lwo s e lec ted time periods . 11lso, the violations which dominate and 
Lhe magnitude of the dominance are not ed, with emphasis placed on the 
r;hanges within each spatial area. Finally, the fourth subsection 
deals with the comparative differences among the spatial areas, 
The Offense and Fire Data 
Before the analysis of the offense and fire data is presented, 
the reader is referred to Table 50, which reports the total nwnber of 
offenses reported by day and by location for both the defined normal 
and riot time periods. The last colwnn of this table refers to those 
offenses for which no l ocat ional data appeared on -i;he police offense 
record . During the normal time period, 291 of 1566 offenses or 18.6 
per cent of all offenses recorded by the police did not contain loca-
tional data. During the riot period selected, 345 of 1689 offenses 
or 20 .4 per cent did not present any locational data. While the normal 
and riot percentages of offenses reported of "unknown location" were 
very constant, thes e percentages do indicate that approximately one-
fifth of the offenses were lost in the analysis of the loGational 
1 
variclble. Two points need to be made here. First, the consistency 
of the percentage of items missing locational information between the 
normal and the riot time periods suggests that the police handling of 
repor ted offenses was not appreciably affected by the riot, Second, 
since one-fifth of the items lack addresses, it is impossible to soy 
which items fall into which category of the locational variable. 
Quite simply, one-fifth of the offense data had to be treated in the 
locational analysis as if it never happened. 
The offense and fire data for the riot areas. The tables analyzed 
i n thls s ection represent the summation of the offenses r eported within 
the riot areas of 14th Street, N. w.; 7th street, N. W.; H Street, 
N. E.; and the small cluster in S. E. Washington . In other words, 
this s~ction analyzes the offense data by category of violation and by 
t ime period for only the riot areas themselves. 
Table J shows the riot area reported offenses by general category 
and by time period. The lambda value of 0.1288 indicates weak as-
sociation between the riot and norinal ti.me periods. Since t,here were 
no reports of ?traffic violations made to the police and since the 
number of reports in the general categories of crimes without victL~s 
and of miscellaneous crimes were qu:i.te low, these subcategorizations 
are not discussed. Crimes against persons decreased in absolute num-
ber during the riot and manifested a proportional decrease of 19.2 
per cent. Crimes against property, while increasing in number, de-
creased proportionately by 11.8 per cent, The fire-related category 
was the only one which manifested a proportionate increase during the 
riot. While crimes against property represented somewhat less than 
one-half of t he offenses reported during the normal time with crimes 
against persons and fires respectively accounting for one?-fourth of 
the offtnses reported, the fire-related category dominated the riot 
time by accounting for almost 6 out of every 10 offenses reported. 
During the riot, reported crimes against persons accoun-ted for less 
than 10 per cent and 0rimes against property decreased to slightly 
over JO per cent. While the riot areas were the locations of massive 
riot destruction, the general public reported primarily instances of 
flr e , 
55 
Table I-,. shows the collapsed category of crimes agains t persons 
for t he rio-L area locations. The lambda va lue of 0.1818 indicates a 
weak association between the riot and normal t:i.Jr,e periods . The 
reader is r eminded that Table 3 indicates an overall increase in of-
fenses r eported during t,he riot, while crimes aga ins?t persons de-
creased proportionate ly and in absolute number from about one-quarter 
during the normal ti;ne to less than 10 per cent dur ing the riot time . 
Within this category and within the four zones of maximum destruction, 
r eported cri.11es involving potentfa l threat of physical harm greatly 
i.ncrea3ed proportionately from a norma l of J . l per cent to a riot per-
centage of J0.9. Crimes involving an actual threat to one's physical 
well-being decr?eased proportionately from a normal percentage of 62. 9 
to a riot percentage of 25.5. Finally, crimes resulting in real phys-
ical harm increased proportiunately by 9.6 per cent, although they de-
creaserj in absolute number during the riot period. While the reported 
crimes involving actual personal injury proportionately i ncreased 
during t he riot, a check of the absolute numbers in Table 4 suggests 
t hat the increase is really due to the fewer munber of crimes report-
ed during the riot period in this general category. 
Table 5 indicates crimes against property within the riot areas, 
and the lambda value of 0.1468 displays a weak associa tion in the 
data. Proportionately, auto theft, vandalism, and stolen property 
remained fairly constant between the normal and riot time periods. 
Reported larc'3ny violations dropped from a normal proportionate per-
~entage of 38.4 to a riot percentage of 6.0. Finally, burglary-house 
breaking proportionately increased frcm 40.9 per cent during the nor-
r1al time to 7J.6 per cent dur i ng the riot. Further, burglary-house 
56 
breaking and l arceny were propor tionat~ ly t he same during the normal 
time per iod, but the f ormer not onl y increased in mwber during the 
riot time bu-t also clearly dominated t he entire category of crimes 
against pr operty. The decrease i n lar!'! eny violations indic:ates that 
the the f ts occurred when t he businesses were closed. 
Table 7 r eports ?the fire-related items r eport ed wi thin the f our 
maj or destruction zones. The lambda value of 0.0164 indicates minimal 
assoc iat ion i n the data . In both the rormal and r iot t ime per iods, 
the re por t ing of the incidence of f i re virtually monopolized t his cat-
egory. Fi res accounted proportionately f or 85.9 per cent dur i ng the 
norma l period and 96,2 per cent during t he r iot period. The riot 
per iod showed a proportionate decrease in fals e fire alarms of 10. 8 
per cen t, but the absolut e numbers indicate that this decrease was 
rea l ly a function of the tremendous riot increase i n the number of 
fire s reported within the riot areas. Clearly, ars on wa s not a fa ctor 
i n t he publi c reportage during the riot. Aga i n, Tabl e 3 shows t he 
domi nation of fires during the riot with a percentage of 25 .1 for the 
normal time and a large increase t o 58.8 per cent during t he riot time . 
The following surnmary stat ements are descri ptive of the major 
findings relative to the offenses aud fires repor ted within the de-
fined riot areas of 14th Street, N. w.; 7th Street, N. w.; H St r eet, 
N. E.; and the small riot area in s. E. Washington. 
1. Within the riot areas themselves during t he normal t ime 
period, the general public r eported primarily crimes 
against property, with lesser proportionate reportage of 
cri.111es against per sons and fire-related items . (Table J) 
2. Within the riot areas t hemselves durlng the r i ot time period , 
57 
flre-related items dominated the offense re8ord with pro-
perty viobtions ranking a distant second and crimes against 
persons ranking a miuimal third, (Table J) 
J. While reportage of crimes involving actua l physical harm in-
creased slightly, reported crimes threatening harm decreased 
proportionately and reported crimes carrying a potential f or 
harm increas ed proportionately. (Table 4) 
4. During the riot time period, reported instances of fires and 
burglary-house breaking accounted proportionately for 56 .6 
ar.d 24,2 per cent of all recorded offenses within the riot 
areas. (Tables 3, 5, and 7) 
The offense and fi~e data for the corridor areas, The tables 
ar,aly zed in this section reflect the summation of the offenses repor t-
ed along the major lines of transportation (Rhode Island Avenue, Ben-
ning Road, and Good Hope Road) where some sporadic riot destruction 
occurred within the District of Columbia. 
Table 9 indicates the violations by general category and by time 
for the corridor areas of sporadic destruction, The lambda value of 
0,0532 indicates minimal association within the data, Traffic viola-
tions, -::rimes without victims, and miscellanGous crimes together to-
taled but eleven offenses for both t~~e periods and are not discuss ed 
further. Crimes against persons proportionately decreased by lJ.5 
per cent from a normal percentage of 23.1 to a riot percentage of 9.6, 
Crimes against property and fire-related items increased during the 
riot somewhat in number but remained proportionately constant between 
the normal and riot times, respectively accounting for approximately 
JO per ceDt and 50 per cent of all the offenses reported during the 
r i oi, . In those areas which experiE:::1ced some riot damage, public re-
porl..ing of total crimin1:1lity differed very little between the nor mal 
and the riot time periods. Fire-related items again dominated both 
t:.he normal and riot periods propo!'tionate\y accounting for about one-
half of the offense data. As in the riot areas, report~d crbnes 
against persons decreased during the riot period in the corridor areas. 
Table 10 reports t,he corridor area offenses by time for crimes 
against persons. The lambda value of 0, 3095 indicates moderate asso-
ciation in the data. Since the total numbers are small, very careful 
inte1?pretations of the percentages are necessary. Perhaps it is saf-
est to say merely that reported crimes involving actual personal in-
jury remained fairly constant, while crimes involving a threat of 
bodily harm (robbery) decreased during the riot and crimes fovolving 
a future potential for personal harm (weapons) increased. 
Table 11 reports the corridor area crimes against property_, and 
the lambda value of O.l.233 indicates weak association between the 
normal and riot time periods, While the total numbers for the normal 
and riot periods were again small, Table 11 clearly shows that burg-
lary-house breaking was most reported during bo(,h time periods. The 
total number of c,ffenses rose from a normal figure of 38 -~o a riot 
figure of 52, while vandalism and auto theft remained constant and 
larceny decreased during the riot period sanewhat. No stolen proper-
ty violations were reported during either time period. 
Table 13 reports fire-related items within the corridor areas, 
and the lambda value of 0,1684 shows weak association between the 
normal and riot periods, The incidence of fire d~ninated both the 
normal and riot periods as evicten~ed by the percentag1.;;s of 70.8 and 
59 
8
7,4, respectively within this category. False fire alarms decreased 
markedly during the riot by 25.8 per cent to a riot low of J.4 per 
cent. Arson reports increased from a zero nwnber during the normal 
time period to eight during the riot. 
The following swnmary statements are descriptive of the major 
findings in the corridor area offense and fire data. 
1. Public reportage of riot period offenses within the corridor 
areas remained reasonably constant when compared to the 
normal time period data. (Table 9) 
2. Reported fire-related items and crimes against property re-
mained proportionately constant between both time periods, 
respectively accounting for approximately JO and 50 per 
cent of all reported corridor area offenses. (Table 9) 
3. The incidence of fires and burglary-house breaking dominated 
the riot period offense data in the corridor areas. 
(Tables 11 and lJ) 
4, Within the corridor areas during the riot period, reported 
crimes involving bodily injury remained constant, while the 
remaining crimes against persons evidenced a movement to-
ward less personal violence. (Table 10) 
Ihe offense and fire data for the non-riot areas. In this sec-
tion the offenses reported for non-riot ar ea locations are analyzed, 
A non-riot area is defined, again, as any area within the District of 
Columbia which does not fall within two city blocks of the four major 
streets of massive riot destruction or within two city blocks of the 
three major lines of transportation which experienced sporadic riot 
damage. In other words, the non-riot areas contain the statistical 
60 
::;wnma tion of all the offenses which have not been classified as riot 
r.irea, C!orridor area, or for which the location data was miscoded or 
cJbsent from the offense record. 
Table 15 reports the non-riot area offenses by general category 
and by time period. The lambda value of 0,0419 indfoates minimal 
as3ociation between tne normal and riot periods. Miscellaneous crimes, 
traffic violations, and c rimes without victims were proportionately 
low in both time periods, reasonably constant between the normal ,:rnd 
riot Limes, and will not be further discussed . Crimes against proper-
ty showed little fluctuc1tion between the normal and riot times, but 
these violations accounted for cilmost 40 per cent of all the offenses 
reported during either time period. Crimes against persons dropped 
proportionately by 6, l per cent during the riot changing from a normc11 
[Jercentage of 15 .1 to a riot percentage of 9, 0. The only general cate-
gory to experien ea riot increase was the fire-related items, whi h 
i ncreased by 9.9 per cent from a normal percentage of 43,1 to a riot 
percentage of 5J . O. Again, fires dominated the offenses reported dur-
ing the riot with crimes against persons not only proportionately de-
creasing but also decreasing in absolute nwnber. If one looks at the 
riot and normal totals , public reporting of crimes as a whole outside 
the riot and corridor areas remained constant. In sum, over 80 per cent 
of all offenses reported and recorded by the police in the non-riot 
areas dealt with crimes aga inst property and fire-related reports in 
both the normal and riot time periods, with the fire category receiv-
ing some ascendancy during the riot time . 
Table 16 focuses on non-riot area reported offenses relative to 
crimes against persons. The lambda value of 0.1575 indicates weak 
61 
ass ociation between the normal. and riot periods. The only category 
t o incr eas e during the riot in absolute number was the potential 
t hreat ca tegory or the weapons violations. Vi olations which res ult in 
':l '..: tua l bodily harm proportionately decreased slightly by 5. 3 per c ent 
a lthough there was a marked reduction in absolute number during the 
riot. Also evidencing both a large reduction in nwnber and proportion 
was the threat of bodily harm category during the riot. While reported 
robberies accounted for about 60 per cent of all crimes against pers ons 
during the normal time period, this percentage dropped to about JO per 
cent dur i ng the riot. Weapons violations increased from a normal per-
centa ge of less than 3 per cent to a riot percentage of about 35 per 
cent. Again, the data clearly indicate a movement toward less real 
personal violence reported during the riot period. 
Table 17 reports non-riot area crimes against property offenses 
where minimal association exists between the normal and riot time 
periods as indicated by the lambda value of 0.0970 , While the total 
nwnber of offenses reported during the normal and riot periods was 
fairly constant, burglary-house breaking reports increased by 18 .8 
per cent from a normal period percentage of J7. 4 to a riot period 
percentage of 56,2. Larceny violations decreased during the riot 
by almost the same amount changing from a normal perc entage of 35.1 
to a riot percentage of 15.2. Auto thefts, stolen property, and 
vandalism remained proportionately constant between both time periods 
with auto theft accounting for about one-fifth of all the offenses 
reported in the crimes against property category. Again, looting 
which was classified as burglary dominated the riot period 
62 
proportiuna tely accounting for over one-half of all non-riot a.l'.'ea of-
fenGes withln this general category and was the only violation to 
manifest an increase in absolute number. 
Table 19 shows fire-related items for the non-riot area and the 
lambda value of O. 0213 shows minimal associcltion between the riot and 
normal periods. Again, Table 15 indicates that the fire -re lated items 
accounted for 43 .1 per cent of all non-riot area reported offenses dur-
ing the normal period and for 5 J . 0 per cent of all non-riot area of-
fenses reported during the riot period. Fires represented the most 
numerous category within th e offense data in the non-riot areas during 
both time periods. Table 19 shows that fires reported to the Fire De-
~artment dominated the fire-related category, accounting for 86.0 per 
cent of the normal period offenses and 89.6 per cent of the riot period 
offenses. The proportional distribution of fires reported remained 
constant, although the absolute number increased during the riot period. 
False fire alarm reports decreased during the riot, while arsor: reports 
increased, although the latter were proportionately minimal during 
either time period. 
The following summarizes the major results of this section. 
1. In the non-riot areas, the total number of offenses reported 
to the police were relatively constant between the two time 
periods. (Table 15 ) 
2, During the riot period in the non-riot areas, fires dominated 
the reported offenses accounting for about one-half of all re-
ports with burglary- hous e breaking accounting for about one-
fifth of the total offenses, each showing some increase over 
the normal time period. (Tables 15 , 17 and 19) 
63 
J , Within the non-riot arec:1s during t.i'ie riot period, reported 
crimP.s involving bodily in.jur.Y remc1inerJ con:;tarit, while the 
remaining crimes against persons evidenced d movement toward 
less personal violence. (Table 16) 
Comparison of the offense and fire data for the riot, n,orridor, 
,rnd non-.,,.iot a.,,.eas. When Tclbles 3, 9, and lJ are compared, a nwnber of 
interesting points are evident in the generc:11 categori~tion of the of-
fern:;t?s b.,1 time anri by location. Traffi8 violations in all three loca-
tiorn:il <:lreas accounted for not one single offense reported in either 
time period. Miscellaneous crimes and crimes without victims were min-
imi.:11 proportionately and in absolute number in the offense data of all 
three locational areas during both time periods. The riot and the cor-
ridor areas evidenced an equcll public conern with crimes against per-
sons during the norma 1 time period, as these offenses accounted for 
about one-quarter of all the offenses reported. However, in the non-
riot areas crimes against persons reported during the normal time 
period accounted for onl,y 15 .1 per cent of a 11 the of fens es. During 
th8 riot period, rE:porting of crimes against persons dropped in al.1 
three locationcll areas to cl relatively constant proportion of <:lll the 
offenses. Further, els one moves from the riot to the corridor to the 
non-riot areas, the riot time period dEmonstrated a steady proportion-
ate decrease of 19,2 per cent, 13.5 per cent, and 6.1 per cent, re-
spectively, in rEported crimes against persons, It would seem, then, 
that the riot effectuated a decrease in reported crimes clgainst persons 
no matter which particul<:lr locational area is studied. 
Directing attention to crimes against property, the data in 
Tables 3, 9, and 15 become a little more difficult to interpret. If 
one looks at t.he corridor area proportions for the normal and riot 
Peri.acts c1nd then at the non-riot area proportions for bot.h time periods, 
ihe data clearly i.ndiccltes little change between both time periods, al-
though the non-riot area proporti.ons were slightly higher. The riot 
areas, however, demonstrated a marked proportionc1te decrease of 11.8 
Per cent .from a high during the normal time period of t,4, 7 per cent to 
c1 low of 32. 9 per cent during the riot. Interestingly, within the riot 
areas themselves, where property damage was at a comparative mc1ximum, 
the Proportion of crimes against property reported to the police was 
c1lmost at its lowest value, although the absolute number of these of-
fenses increased somewhat. Not only was the proportion of crimes 
c1gainst property reported the greatest in the riot areas during the 
normal time period when compared to the corridor and non-riot areas, 
but also the greatest proportionate reduction in crimes against pro-
Perty reported during the riot occurred in the riot areas. 
The fire data represented proportionately the major category of 
PUblic concern during the riot period in all three locational areas. 
Fire-related items accounted for almost 6 out of 10 offenses reported 
during the riot within the riot areas and slightly over 5 out of every 
lo Offenses reported durir1g the riot within the corridor and non-riot 
i:lreas. While this would be c1 logical supposition for any riot period, 
the normal time period indicated proportions of 25.1 per cent, 48.5 per 
cent, and 43.1 per cent for fire-related items within the riot areas, 
the corridor areas, and the non-riot areas, respectively. In other 
words, while there were increases in the absolute number of offenses 
reported in this category in all of the locational areas, the corridor 
areas remained relatively constant with the non-riot areas experiencing 
ubout a 10 per cent increase and the riot areas experiencing about a 
JJ per cent lncrease. One would expect a great inr~rease in reports 
of fire within the riot ..ireas, which the data support;:,. Interesting-
Jy though, fires incre,:1Sed in the non-riot areas as well, although 
this latter increase was not as dramatic in magnitude, Equally curious 
was the relatively constant figure during both time periods in the cor-
rid0r areas of sporadic destruction. To put it another way, fi:!:'es 
dr:Jmatically increased in the riot areas but also were quite in evi-
dence in the non-riot areas as well, while remaining constant in the 
corridor areas. 
Finally, if one looks at the total numbers of offenses reported 
during the normal and riot periods in each of the three locational 
areas, it was obvious that the greatest increase during tbe riot oc-
curred in the riot areas with some increase in the corridor areas and 
a relatively constant figure in the non-riot areas. 
Tables 4, 10, and 16 report the offenses for crimes against per-
aons for the three locational <e1reas . We may c1ssume that crimes against 
persons demand sufficient public c1ttention that changes in the social 
situation will affect this category of crirninali ty as a whole to a 
l'::sser degree than the other general <::ategories . Remember, however, 
in terms of absolute numbers reported crimes agc1inst persons decrecisect 
during the riot in a ll three locational areas. If we look within the 
categories of these tables in terms of the absolute numbers, crimes 
involving actual b odily harm and crimes involving a threat of bodily 
harm decreased during the riot, with the latter showing very dramatic 
proportional decreases of 37,I+ per cent, 58.J per cent, anri 27.8 per 
ce nt for the riot, corridor, and non-riot areas respectively, 2 Only 
66 
the potential threat of bodily harm category (weapons violations) evi-
denr:::.ed an increasing number during the riot period. Comparatively, re-
ported crimes involving actual bodily in jury remained proportionately 
the most corisisten~ category between the normal and riot periods in 
all three locatioual areas . Since the absolute numbers in these tables 
are relatively small, the percentages are somewhat misleading. Perhaps 
after a perusal of the absolute uumbers, it is safest to say that the 
riot caused a dimunition in the level of violence in reported crimes 
agair,st persons. The riot seems to have been accompanied by less real 
or actual personal harm than found in the normal period. Further, re-
ported crimes suggesting a future potential for personal violence in-
creased, indicating a greater public concern with the less violent end 
of the continuum of violence against persons. These conclusions are 
applicable to all three locational areas. 
Tables 5, 11, and 17 specify offenses related to crimes against 
property for the three ma ,Jor locational areas. Lookiug at the riot 
and normal time period totals in these three tables, it may be r~on-
cludE::d, obviously, that the riot period was accompanied by a large in-
crease within the riot areas, a moderate increase within the corridor 
areas, and a relatively constant number in the non-r:iot areas of re-
ported crimes against property. Reported auto thefts, stolen property 
violdtions, and vandalism remained reasonably constant in flll three 
locational are~s between t he normal and riot periods. Of the three 
r_:rirnes , the incidence of auto t beft appeared most prevalent with van-
dalism and stolen property comparatively and rtJ spectively ranking a 
poor second and third. What is interesting here is that the offense 
data do not indicate any real public ~eporting of stolen property or 
67 
vandalism in any of the three locational areas during the riot period, 
The normal time period reports of la1?ceny violations propartiona t e-
ly accounted for roughly one third of all offenses in this general cate-
gory in all thn,e l ocational areas. Not c,nly did the absolute number s 
decrtase during the riot in all three areas, but also the proportionate 
distribution of reported larcenies decreased to 6.0 per cent, 11.5 per 
cent, and 15.2 per cent for the riot, corridor, and non-riot areas, 
respectively. When many businesses were closed, the possibilitie s for 
larceny violations were severely restricted, Tha?t we may make this 
supposition is further reinforced by the data in 1able 47, which re-
perts offenses by specific catego:ry and by date during the riot period 
itself. April 5th was a Friday, the first full day of rioting, and 
businesses began the day in a somewhat normal fashion. The number of 
reported larcenies on this day was 36 from a riot period total of 146. 
April 6th and 7th were weekend days at the peak of the rioc peri od with 
April 9th being the day of Martin Luther King, Jr. 1 s fune1?al, on which 
many businesses closed, During these four days, reported larceny vio-
lations totaled but 28 out of 1~6. The last three days of the riot, 
April 10th through April 12th, manifested an increase of 82 reported 
larceny violati.ons, when the peak of the riot impact period was over. 
Burglary-house breaking is legally defined as a theft accompanied 
by breaking and entering or by unauthorized presence. With many busi-
nesses closed or restricting their hours of commerce during the riot, 
one would expect this category of crimes against property to increase. 
Tables 5, 11, and 17 show that during the riot period, burglary ac-
counted for 7J.6 per cent, 75,0 per cent, and 56.2 per cent of all re-
ported offenses against property in the riot, corridor, and non-riot 
68 
areas, respectively. These facts are what one would logically suppose 
during any riot period. It is interesting to note that in terms of 
the ab3olute numbers of burglaries for which locational information is 
available, the total riot period increase in reported burglaries was 
251 of wnich 111 occurred outside the riot or corridor areas, Put an-
other way, of the totaJ number of 581 reported burglaries during the 
riot for which locational information is available, J58 or well over 
half of these burglaries were not in the riot or corridor areas? These 
statistics must be tempered with the fact that an additional 98 burg-
laries were reported during the riot period for which there is no lo-
cational data. However, if we assume that all of these 98 additional 
violations occurred in the riot and/or corridor areas, then a ?total of 
J21 bu:cglary violations occurred in areas which experienced heavy or 
~oderate riot period damage, while J58 violations occurred elsewhere 
in Wa shington. The conclusi.on must be reached that looting was fairly 
prevalent outside the riot and corridor areas. In other words, looting 
was not as localized in the ghetto areas ~f major destruction and the 
corridor areas of lesser destruction as one might suppose. 
Tables 7, 13, and 19 report the fire-related offenses for the 
three locational areas. Looking first at the totals of the absolute 
numbers in the normal and riot time periods, the riot period showed an 
increase in the number of items reported in all three locational areas. 
However, only the riot areas ther1selves experienced a dramatic increase, 
while the corridor and non-riot areas increases were comparatively 
modest. False fire alarms decreasJd during the riot in all three lo-
cational areas to a prop~rtionately minimal amount in the riot and 
~orridor areas and to a proportionately modest level in the non-riot 
69 
areas. Further, wbile the number of arson reports increased J11ring the 
riot in all three locational ar8as, proportionately these offenses ac-
counted for minDnal amounts during both times periods. During the 
normal time period, report,ed fires accounted for 35,9 per cent, 70. 8 
per c~nt, and 86.0 p<3r cent of all fire-related offenses reported in 
the riot, corridor, and non-riot areas, respectively. During the !'iot 
time period, tbese proportions increased to 96.2 per cent, 87 .4 per 
cent, and 89.6 per cent within the riot, corridor, and non-riot loca-
tions, respectiv~ly. Put in other terms, the absolute number of fires 
repo~ted in the riot areas during the riot period was over five times 
the normal time period number. In the corridor areas, the number of 
fires reported during the riot was slightly over one and one-half times 
the normal time period number, and the non-riot areas experienced an in-
crease in reported fi.res during the riot of less than one and one-half 
times thG normal number. That the incidence of fires dominated the of-
fenses reported during the riot situation is obvious from t hese data and 
from the proportional figures presented in Tables J, 9, and .15 . And as 
one moves from the areas of ccxnparatively maximum destruction to the 
areas of comparatively mir:imum destructiori, the riot period experieuce 
with increasing numbers of fires diminished. All of the above is to 
be expected. What is not expected was the number of fires reported 
during the riot in the non-riot areas (Table 52). During the normal 
period, the non-riot areas accounted for slightly over 80 per cent of 
all fires reported in the District of Columbia. While this pe:cc enta ..,.e 
0 
decreaserl during the riot period, the non-riot areas still accounted 
for slightly over 60 per cent of all reported fires. We may conclude 
) 
then, that while the major areas of destruction showed dramatic rio~ 
70 
period incr':lases in t he incidence of fires, the non-riot areas experi-
er;~e.J a high incidt=mce of fires themselves du.cing the riot period. 
The following s ummary statements highlight t he major findings of 
the analysis in this section. 
1. Reported traffic violations, crimes without victims, and mis-
cellaneous crimes were minimal in proportion during both time 
periods in all three locational area J . (Tables 3, 9, and 15) 
2, The riot period studied was characterized by dec-reases in re-
ported crimes against persons in all thre~ spatial areas with 
the decreases varying directly wi?th the concentration of riot 
damage and a concanitant diminution in the reported level of 
violence. (Tables J, 4, 9, 11, 15, and 16) 
J. Reported fire-related items proportionately d~ninated the of-
fense data in both time periods and all locational area~ with 
the exception of the riot area normal t~ne period with maxi-
mum reportage of the incidence of fires and minima l reporting 
of arsonist activity, as well as a reduction in riot period 
reported false fire alarms in all three locational areas. 
(Tables 3, ?, 9, lJ, 15, and 19) 
4. When the three locational areas were compared, the greatest 
proportion of reported crimes against property during t he 
normal time period and the smallest proportion of reported 
crimes against property during the riot time period both 
occurred within the riot areas themselves. (Table.::,~ 3 , 9 , 
and 15) 
5. The degree of in::rease in absolute number of offeno::,es reported 
between both time periods varied directly with the 
71 
conoentration of riot damage. (Tables 3, 9, and 15) 
6. During the riot time period, reported larceny violations de-
creased and reported burglary offenses increaJed in all three 
locational areas, while auto theft, stoler. property, and van-
dalism remained proportionati!ly relatively constant. (Tables 
5, 11, and 17) 
7. While reported burglaries and incidences of fire dominated the 
offense data in all three locational areas, the majori.ty of 
the re ported offenses in these two categories occurred in the 
non-riot areas. (Tables 5, 7, 11, 13, 17, 19, and 52) 
~valuation of th~ offense data hypothesis. Before the analysic of 
the lambda values is presented, we must digress somewhat to put these 
data into sane more reasonable context. One of the as sumptions which 
guided the selection and definition of the spatial categories was the 
idea that the situation reported to the police and the response made 
by the police would both be affected by prox:!..mity to t he areas of major 
riot damage. After making such an assumption , it would appear logical 
to posit a number of ideas. First, since the defined rio~ areas are 
those areas of maximum concentration of riot activity, one wculd ex-
pect the police to concentrate a majority of their resources within 
these areas. Second, since the damage surveys indicated sQne sporadic 
destruction in the defined corridor areas, one would expect some second-
ary concentration of police effort within these spatial areas. Finally, 
since the non-riot areas experienced cooiparatively little riot damag':? , 
then police arrests would be minL~al in these areas compared to the 
riot and corridor areas. Quite simply, it 3eems logical to suppose 
that police response would indicate the greatest number of riot period 
72 
:Jrrests within the riot areas wiT,h fewer arrests in the corridor areas 
and with very little attention paid to the non-riot areas. It waa 
further ~soumed that the offense and fire data reported to and recorded 
by the police would follow a similar pattern. 
While such a series of aesumptions appear logical, they must be 
tempered somewhat by one additional factor. A check of the map defin-
ing the major riot areas and the corridor areas will definitely indi-
ca t.e that t.hn geographic area enccmpassed by both of these spatial area.::, 
is quite small when compared to the remaining areas classified as non-
riot. 3 Therefore when the reader looks at the absolute numbero in the 
three spatial categories, he might very well expect to see a larger 
number of offenses and fires, on the one hand, and of arrests, on the 
other hand, in the non-riot area s than the assump~ions presented above 
might suggest. That such is true, the data will support. 
Yet, the real curiosity in the offense and fire data exists in th':! 
corridor areas. If one looks at the total numbers of normal and riot 
~ime period offenses and fires reported in the corridor areas when 
compared to the s ame data in the riot and non-riot areas, H is obvious 
that the corridor area data are quite small in number. For example, 
Tables 3, 9, and 15 contain the total number of offenses anJ fires re-
ported for the riot., corridor, and non-riot areas, respectively. Dt.r-
ing the normal time period, the riot, corridor, and non-riot areas 
accounted for, respectively, 16.6 per cent, 6.1 per cent, and 77.4 per 
cent of all the reported offense data for which locational information 
is available, During the riot period, this riot area proportion in-
c-reased to 2$ .J per cerrt and the non-riot area proportion decreased 
to 65.5 per cent, as one m:i..ght logically expect. But curiously, the 
73 
corridor aren proportion re:nained e;ons urnt increasing during the riot 
pericJ by ouJy 0.1 per cent. This iG to say that in the areas of ::,ome 
0poradic destruction during the dot period, reportage of fires 1md 
offer,ses proportionately did not change when compared to the other 
spu tial are'3s. 
Furthr;r, if one focuses on the collepsed subcategorizations within 
the corridor spatial areas (Tables 10 through 14), two charac~eristlcs 
of the corridor area data become obvi ous. First, some subclassifica-
tions evidenced very small numbers of items. For example, the crimes 
against persona subcategory contained only 16 offenses reported during 
the riot period, the miscellaneous crimes subcategory contained but 
one item for both time periods, and there was not one offense recorded 
in the traffic oubcategory during either time period. Sec0nd, some of 
the subclassifications not only contain min~nal numbers of offenses 
reported but also indicate that all or most of the items reported 
~lustered within one cell of the table. For example, the subcategory 
of crimes without victims contains all 10 items in the riot period non-
riot related cell. In view of these considerations, the analy3is of 
the lambda values for the c orridor area data are not particularly 
meaningful and are not discussed in relation to the posited hypothesis. 
To summarize the above, it is suggested that a variety of facto1?s 
make any meaningful interpretation of lambda difficult i.n the corridor 
area data. However, this does not dismiss the very interesting ques-
tion of why the numbers of offenses reported in the corridor areas are 
so small. It merely means that the lambda statistic will not assist 
us toward an answer to ?this question. 
Table 21 reports the lambda values for the offense data by general 
74 
catP.gory and collaps?d subcategory and by spatial area. It has been 
hypothesized that the degree of association between the cffenGes re-
porl.ed and the selected riot-normaJ. time period varies directly wlth 
the degree of concentrated riot damage. In other words, it is sug-
gested that the closer one moves toward the major riot areas, the 
higher the lambda value will be, indicating an increasing degree of 
association. 
The calculated lambda values for the general categoriz~tion and 
the subclassification of crimes against property indicate increasing 
::3ssociation as one moves from t;he areas of least damage concentra t.ion 
to the areas of greatest damage concentratiun. The3e data indicate 
Lhat the hypothesis as stated should fail to be rejected. However, the 
lambda values for crimes aga i nst persons and fire-related i?Lems evidence 
a relatively constant. degree of ass ociation as one moves from the riot 
to the non-riot areas. Finally, the collapsed subclassifications of 
traffic violations, crimes without victims, and miscellaneous crimes 
contained so few i~ems that the evaluation of the lambda values for 
any spatial area would be misleading. In view of the minimal degree of 
change from one locational area to another for some offense categories, 
it must be concluded that the data warrants neither a rejection of nor 
a failure to reject the above hypothesis. Rather, the data appear 
inconclusive. 
However., it is interesting to note that with the exception of the 
corridor area crimrJS against persons, all the other lambda values fall 
in the minimal to weak range of association, indicating not only that 
there were few differences in reported offenses between the defined 
organi zed response period of the riot and a representative normal time 
75 
pl'::riod b..it also that what differences exi.st are of a rather small mag-
nitude, This i0 to say that given the modal marginal category on an.v,  
un i? variable, thE::: knowledge of thP. modal marginal category on the other 
variable uoes not sufficiently reduc e the er ror in one's prediction 
of the dist.ribudon of the data . 
In s 1.rn, while the data do :iut indicat'3 clear support or denlal of 
t he hypothesiG, they do indl":!ate a very small degree of dif feren'.:!e in 
r i ot period reportage of violations uhen compared to a normal time 
period , 
The Arrest Data 
Before the analysis of the arrest data is presented, it is necess ary 
t o digress into the issue of t h ose arrests for which no locational in-
f ormation appeared on t he official arrest record , The reader is refer-
red to Table 51, which reports the total number of arrests by locational 
are a and by date for both the normal and riot periods. The column iden-
t ified as "location unknown" reflE:cts the number of arrests for which 
n o locational information appeared on the arrest record. The collunn 
identif ied as "miscoded" r e ports the arrests for ? which the 1.o,;ation, 
while specified, was uninterpretable due to supposed miscoding or key-
punch errors. In terms of the analysis in t his thesis, it does not 
matter whether the locational information is deficient because it is 
missing or inaccurate. Ra ther, both categories are combined as the 
total number of arrests which must be disregarded in the locational 
analysis. Hence, during the normal time period, 522 of 3387 arrests 
(15 .4 per cent) had to be dropped from the locational analysis, During 
the riot period, J.805 of 7J08 arrests (24.7 per cent) had to be dropped. 
76 
It appears that the riot period arrest data were ~haracterized by an 
Jncrea3c in the numoer of item.:; for which locational information was 
missing or inaccurate. Again, there is no way to assign i n any mean-
ingful fashion thes~ arrests to one or an~ther of the locational areas. 
W,uite simply, they have been deleted . 
The arrest data fer the riot are::is. In this section the arrests 
are analyzed for the four major riot areas by time period. The reader 
is reminded that the arrest data are a partial indicator of the re-
sponses na de by the District of Columbia Metropolitan Police Depart.ment 
to the normal and riot periods. 
Table 22 shows the arrests made by general category and by t~~e 
period for the riot areas. The lambda value of 0 .1584 indicates weak 
association. Arrests for miscellaneous crimes, violations related to 
fires, and crimes against persons were minimal in absolute number and 
proportion durine; both the normal and riot time periods and will not 
be discussed further. Traffic violations during the normal time period 
accounted for s lightly less than one out of every two arrests made in 
the riot areas. However, during the riot period traffic citations 
dropped drastically from a normal high of 47 . 1 per cent to a riot per-
centage of 5. L Put another way, during the riot police arrests lr1-
creased almost four times while traffic arrests decreased by almost 
two?? t hirds. Crimes wi t:.hout victims claimed the next greatest propor-
tion of c1rrests during the normal time period with 35, 5 per ce11t, c1nd 
they radically increased during the ~iot to proportionaGely 74.7 per 
cent . Arrests for crimes against property increased during the riot 
p13riod I)roportionately to 17.6 per cent from the normal time period 
percentage of 8.J. Finally, in terms of absolute nwnbers, the police 
77 
"lrrssts greatly increased in the riot areas during the riot period. 
Table 23 shows the riot area arrests for crimes against persons. 
Th~ la:r:bda value of O. 2459 indfoates moderate association between the 
normal and rio~ time periods. The percentages here must be interprsted 
wi.th great care, since the absolute numbers are so very small. The 
reader should remember that as a whole, this category of arrests de-
".!rea.:,ed during the riot proportionately to a minimal percentage of less 
than 2 per cent of all arrests made (Table 22). Perhaps, we should say 
that arrests for crir.ies involving actual bodily harm and for crimes 
threatening bodily harm (robbery) remained relatively constant during 
both tir.le periods. There was a slight increase in nwnber (from 4 to 
18) in the arrests made for weapons violations or crimes involving a 
f11ture potential for bodily harm. On the whole, however, police re-
sponse to crimes against persons within the riot areas remained rela-
t.ively constant during both time pEriods. 
Table 24 indicates arrests for crimes again;:,t property in the riot. 
areas, and the lambda value of 0.2909 indicates a moderate degree of 
association in the data. Auto theft arrests were minimal durhg both 
time periods, while vandalism arres ts remained proportionately constant 
accounting for a normal percentage of 7 .9 and a riot, percentage of 6.J, 
St0len property arrests during the r iot period propor?t;iona tely increased 
from a normal percentage of O. O to a riot percentage of 7. 0, If one 
looks at the absolute numbers, the police arrested 40 persons for stolen 
property and vandalism during the riot period. This is to say that the 
poli ce did not utilize these categories of criminality to any degree 
within ~he riot areas. During the normal time period, larceny arrests 
do:ninated proportionately crimes against property accounting for 60. 5 
78 
P":r ci:nt, although the numbers were small. Durlng the riot period, 
la r~b1y arrests all but evaporated acGounting for only l. 0 per cent 
r,,rri[.J'Jrtionatcly. Burglary-house breaking proportionately accounted for 
20 . 9 per ctJnt of the normal period arrests and dramatically increased 
t.u 85. 0 per cent during the riot period within the riot areas, 
Table 25 indicates traffic citations within the riot areas, and 
,,be lambda value of 0, 0000 indicates minimal association in the data. 
'I'rir-; i:lb.:rnlute numbers indicate that police concern with any kind of 
i:,raff1c viola t.ion decreased within the riot areas during the riot per-
i vd, However, we may logically surmise that the opportunity for polir::e 
, rreGts in this category would be radically reduced by the quite severe 
l iml tations pla ed on vehicular movement by the riot situation itself. 
',/hat is ::rnrprising here is that the police made any arrests at all dur-
i ng the riot period within the riot areas themselves for traffic viola-
r, ir:.m~;. Further, the proportionate distributions of moving and non-moving 
?riolations between the normal and riot time periods were the same, which 
lndicates that the riot did not appreciably change the police response 
t,o different kinds of traffic violations. 
Table 26 reports arrests for crimes without victims within the 
r.iat areas, and the lambda value of 0.170J indicates weak association 
in the data . Arrests for non-riot related crimes (prostitutlon and 
vice, narcotics and drugs, and gambling) proportionately decreased from 
a normal percentage of 6.7 to a riot percentage of 0.5, While the num-
b~r of arrests for drunkenness during the riot period increased some-
what within the riot areas, this category was proportionately decreased 
by '.J6.9 per cent by the tremendous increase in "disorderly conduct" 
'Jrrests. The reader will recall from an earlier discussion that curfew 
'/9 
r~strictions were placed on, among other things, personal movement dur-
ing selected hours of the riot period days, Arrests for curfew viola-
tions were classified as disorderly conducts with no real possibility 
of practically separating these two violations from the arrest record. 
During the normal time period, disorderly conduct arrests accounted for 
about one out of every four arrests made within the category of crimes 
without victims. During the riot period, disorderly conduct (plus cur-
few violations) accounted for nine out of every ten arrests. Further, 
in terms of absolute numbers, within the riot areas this category in-
creased from 43 arrests in the normal period to 1144 arrests in the riot 
period . 
The following descriptive statements highlight the major findings 
relative to the arrest data for the riot areas . 
1. During the normal time period traffic arrests (principally 
moving violations) dominated the arrest data in the riot 
areas ., while crimes without victims (principally disorderly 
conduct and curfew violations) dominated the riot time 
period in these areas. (Tables 22, 25, and 26) 
2. The riot period indicated no appreciable change in the riot 
areas in arrests for crimes against persons, crimes related 
to fires, and miscellaneous crimes. (Table 22) 
J, Within the riot areas in the crimes against property category, 
the normal time period was dominated by arrests for larceny 
violations, while the riot period was dominated by burglary-
house breaking arrests. (Table 24) 
4. No arrests were made during either time period within the riot 
areas for arson or reporting a false fire alarm. (Table 27) 
80 
The arrest data for the corridor i;\reas. In this section the arrest 
rJ[;Jta for the ci r ea s of sporadic riot destruction along the major lines of 
Lransporta tior1 are analyzed by time period and type of violation. 
Ti::lble 29 reports the arrest data by generi::ll category and by time 
[Je.ciod . The lambda value of 0.2274 indicates moderate association 
amongst the data. Arrests f or crimes against persons, crimes related 
Lo fir es , and miscellane ous crimes were minimal in terms of absolute 
mm1ber s and proportions during both ti.me periods within the corridor 
areas . Arrests for crimes against property increased proportionately 
by 16. 5 per cent from a normal percentage of 1. 1 to a riot percentage of 
17.6. More dramatically, arrests in this category increased in abso-
lute numbers from one during the normal time period to eighty-two during 
the riot period. Traffic arrests decreased proportionately from 71.6 
per cent during the normal time period to 9, 0 per cent during the riot 
period. Arres ts for crimes without victims increased during the riot 
from a normal percentage of 18. 2 to a riot percentage of 70. l. All of 
these percentages must be interpreted carefully because of the compara-
tively small total nwnber of arrests during t he normal time period. In 
other words, the nwnber of arrests for crimes against property very 
greatly increased during the riot, but the percent age distribution is 
moderated by the tremendous ire rease in arrests for crimes without 
victims. Finally, traffic arrests dominated the normal time period 
accounting for seven out of every ten arrests in the corridor areas. 
During the riot period, arrests for crimes without victims accounted 
for seven out of every ten arrests wi th propert;y violations accounting 
for somewhat .less than one out of every five arrests. Finally, in t erms 
of absolute numbers the corridor area arrests increased quite 
8.l 
3ub::,t::.i n tia lly during the riot period. 
Table Jl indicateG arrests within the corridor areas for crimes 
again.st property by time period. The lambda value of 0,0000 indicates 
minimal assocfa tion between the normal and riot periods. Strangely 
1:-;nough, there was only one arrest during the normal time period for any 
r1rime against property. The only category to appreciably increase dur-
ing the riot period was burglary-house breaking. 
Table 32 reports arrests in the c orridor areas for traffic viola-
tions by time period, and the lambda value of 0,0000 indicates minimal 
association within the data. There was no measurable distinction be-
tween the proportions of arrests for different types of traffic viola-
tions between the normal and riot periods. In terms of absolute numbers 
the total number of arrests for traffic decreased during the riot period 
by approximately one-third. 
Table 33 reports corridor area arrests for crimes without victims 
~y time period, and the lambda value of 0.0455 indicates minimal asso-
ciation in the data. Arrests for non-riot related crimes were nonex-
istent during either time period. Because of the very few nt.nnber of 
arrests in tne corridor areas during the normal time period, the dis-
tributional percentages are somewhat misleading. Suffice it to say 
that during the riot period, there was a very large increase in the 
number of disorderly co:nduct (and curfew) arrests with a slight in-
crease in the nwnber of arrests for drunkenness . Clearly, disorderly 
conducts dominated the arrests during the riot period within the cor-
ridor areas. 
The following statements are indicative of the major findings of 
the analysis of the arrest data in the corridor areas. 
82 
1. Within the corridor areas the normal time period arrest data was 
dominated by traffic citations, while the riot period arrest 
data was dominated by crimes without victims, predominantly 
disorderly conduct and curfew arrests. (Tables 29, 32, and 33) 
2. During the riot time period arrests for burglary-house breaking 
dominated the crimes against property general category and 
evidenced a large increase when compared to the normal time 
period within the corridor areas. (Table Jl) 
The arrest data for the rion-riot areas. Analyzed within this sec-
tion are the arrest data for all those remaining areas within the Dis-
trict of Columbia which do not fall into the defined riot areas (14th 
Street, N. w.; 7th Street, N. w.; H Street, N. E.; or SE Washington) 
or into the defined corridor areas (Benning Road, Good Hope Road, or 
Rhode Island Avenue). 
Table 36 reports the non-riot area arrests by general category 
and by time period, The lambda value of 0,4595 indicates moderate 
association in the data. Since this lambda value is the highest at-
tained in the data analysis, we may conclude that the police response 
within the non-riot areas was least consistent between the riot and 
normal time periods. Arrests for crimes against persons, fire-?related 
crimes, and miscellaneous crimes were not only minimal but also rela-
tively constant during both time periods, Arrests for crimes against 
property increased proportionately during the riot period by 10.9 per 
cent from a normal percentage of 4 . 2 to a riot percentage of 15 .1. Also 
important here is a greater than quadrupling of the number of these ar-
rests during the r iot period . Traffic citations clearly dominated the 
arrests in the non-riot areas during the normal time period, as they 
prr.Jportionately accounted for 69. 5 per cent of all the arrests made. 
However , during the riot period, traffic citations dropped to 15.2 per 
r;8nt. Arrests for crimes without victims dominated the riot period, 
ucc:ounting proportionately for 6J.9 per cent of all the arrests made in 
the non-riot areas. This was a proportionate increase of 46.7 per cent 
from a percentage of 17.2 during the normal time period, Finally, if 
one looks at the total number of arrests for both time periods, he will 
3ee that there was a substantial increase during the riot period from 
a normal total of 2656 to a riot total of JJ28. Apparently, the police 
were quite active outside of the riot and corridor areas during the riot 
period . 
Table 37 shows the non-riot area arrests for crimes against persons 
by t.ime period. The lambda value of O. 2203 indicates moderate associa -
tion within the data. It is interesting to note that in terms of total 
numbers , the police made fewer arrests during the riot for crimes clclssi-
fied in this category than they normally did. Further, arrests for 
'..:rimes which really result in personal injurJ1 decreased proportionately 
from a normal percentage of 61.0 to a riot percentage of 33,6 . Also, 
Lhe absolute number of arrests in this subcategory halved during the 
riot. The number and proportion of arrests for crimes threatening 
bodily injury remained constant_, while arrests for crimes with a future 
potential for bodily harm increased proportionately by 28,0 per cent 
from a normal percentage of 26 . 0 to a riot percentage of 54.0. Clearly, 
arrests for crimes involving personal injury dominated the normal t:ime 
period, while arrests for crimes involving potential future bodily 
harm dominated the riot period within the general category of crimes 
against persons. Ther efore, we may conclude that the riot was 
u';?;c,rnpanied by a real reduction in the level of personal violence within 
,, 11(; non-riot areas . 
Table J8 reports non-riot areu arrests for crimes against property 
tw tirn8 period, and the lambdc1 value of 0.0923 indicates minimal asso-
'; la ti ori. The proporti.onc1te distribution be tween the norma 1 and riot 
time periods sllould be very carefully :interpreted because of the rela-
r,Lrdy ,3mall total number of arrests made in this general category 
rJ1 rlng the normal time period, Proportionately, burglary-house break-
irig and larceny equally dominated the normal period within the non-riot 
:Jreas, respectively accounting for J6.6 per cent and .34 . 8 per cent of 
the total arrests made. During the riot period, however, bu:cglary 
'; le~rl.Y dominated the arrest dc1ta, accounting for 82.5 per cent of all 
the arrests for crimes against property. While arrests for auto theft 
and stolen property increased during the riot when compared to the 
normal period , these increases were proportionately dwarfed by the 
tremendous increase in burglary arrests. 
Table 39 reflects the non-riot area arrests for traffic violatiom, 
and the lambda value of 0.0000 indicates minimal association in the 
r.i~ ta. In terms of absolute numbers, the riot caused a tremendous diminu -
tion in traffic citations which fell from a normal total of 1846 to a 
riot total of 505. However, the riot did not affect the proportionate 
dl3tribution of the two basic types of traffic violations. 
Table 40 reports the non-riot area arrests for crimes without vic-
tims by time period. The lambda value of 0.2818 indicates moderate as-
:,;ociat:i.on in the data. The reader will remember that arrests in the 
eeneral category of crimes without victims accounted for slightly more 
than 6 out of every 10 arrests made during the riot period in the 
85 
,,,Jn-?.d0t :Jf'fias ,rnd slightly less than 2 out of every 10 during the 
rirJrir1al pi:.:ri od (Table J6). Non-riot related arrests (pros titution and 
., i'''o, nur 1.! ot ics and drugs, and gambling) were minimal in number and 
r,rorjort l on during both time periods. Drunkenness proportionately ac-
; r../rnt,i:.:d f or 71. 3 per cent of the arrests during the normal ti.me period 
? ,nrJ der;reased during the riot to 11. 5 per cent . Disorderly conduct 
,~rr<1sts proportionately increased dramatically from a normal percentage 
rjr ~ L~ ? .3 to a riot percentage of 87. 8. That this increase was meaning-
Lul i s reinforced by the absolute numbers which show that disorderly 
r~c,ndur.::t acc ounted for 111 normal time period arrests and 1867 riot time 
01::::rlod arrests. 
The following summary statements highlight the major findings of 
this section. 
1. Within the non-riot areas, the normal tiD1e period was dominated 
by arrests for traffic violcltions, while the riot time period 
was dominated by arrests for crimes without victims, predom-
inantly curfew and disorderly conduct violations. (Tables 
J6 and 40) 
2. The increase in the number of arrests in the non-riot areas 
during the riot period indicates that the police were quite 
active outside the defined riot and corridor spatial areas. 
(Table J6) 
3. In the non-riot areas during the riot time period, the arres t 
data indicate a reduction in the level of personal violence, 
as well as a decrease in larceny arrests and an increase in 
burglary-house breaking arrests. (Tables 37 and J8) 
Comparison of the arrest data for the riot, corridor, and non-riot 
86 
tH?e::w. Tables 22, 29, and )6 report t.he arrests by general category 
:rnd by time period for the riot, corridor, and non-riot areas, respect-
iv,:dy. Comparatively, arrests for crimes against per?sons and for mis-
':ellaneous crimes represented minimal numbers and proportions during 
both tlme periods in all three locational areas, although both cate-
6ories showed slight proportionate reductions during the riot period, 
11rre::;ts for arson and reporting a false fire alarm totaled but three 
during both time periods in all three locational areas. Arrests for 
.:rimes against property increased proportionately and in absolute num-
bers during the riot period in all three locational areas, accounting 
for normal percentages of 8.), 1.1, and 4.2 which increased to 17.6, 
l? .6, c1nd 15,1 during the riot period in the riot, corridor ., c1nd non-
riot c1reas, respectively. What is interesting here is that arrests in 
this category were fairly constc1ntly distributed proportionately through-
out the three locational areas during the riot period itself. 
Traffic citations accounted proportionately for 47.1 per cent, 
71.6 per cent, and 69.5 per cent of all arreots made in the riot, cor-
ridor, and non-riot areas, respectively, during the normal time period. 
Further , these violations dominated the arrest data in the normal time 
period, accounting for 2317 of 378) or 61.2 per cent of the total ar-
rests made (Table 48). The normal period traffic citations were pro-
portionately the same in the corridor and non-riot areas, accounting 
for approximately 7 out of every 10 arrests while sli ghtly less tnan 
5 out of every 10 arrests fell into this category during the normal 
time period within the riot areas. During the riot period traffic 
,~itations were proportionately reduced to 5,1 per cent, 9.0 per cent, 
and 15 . 2 per cent within the riot, c orridor, and non-riot areas, 
8'? 
re.sp,~c tlvely. It appears that as one moves from the geographic areas 
ui' grea te::; t concentration of riot property damage to the areas of least 
c;0ncentra tion of property damage, the proportion of arrests for traffic 
violations increased, indicating that traffic ci tations varied inversely 
with the r;oncentration of property damage during the riot period. It 
~liould be noted, however, that in all three locational areas, the num-
bers of traffic citations during the riot period were dramatically cur-
Lciiled when compared to the normal time period, But even though the 
riut was accompanied by a radical reduction in traffic citations, these 
arrests s till accounted for 712 of a total of 7)08 (or 9.7 per cent) 
of the arrests made. Only burglary-house breaking (937) and disorder ly 
,:;onduct (4501.) accounted for a greater volume of arrests during the riot 
period (Table 49). 
Finally, arrests for crimes without victims accounted proportion-
ately for 35,5 per cent, 18.2 per cent, and 17,2 per cent of all the 
arres ts made during the normal time period within the riot, corridor, 
and non-riot locational areas. During the riot period, these arrests 
dramatically increased in nwnber and proportion to 74,7 per cent, 70.1 
per cent, and 63.9 per cent within the riot, corridor, and non-riot 
areas , re spec tive1,.v . Clearly, this general category of crimes domin-
ated t.i1e arrest data during the riot period in all ~hree locational 
areas . Further, there were slight proportionate decreases in this 
category of arrests as one moves from the areas of greatest concentra-
tion of the riot damage to the areas of lesser concentrated damage. 
Tables 23, JO, and .37 report arrests for crimes against persons 
by time period for the three locational areas. Since the absolute 
numbers in all three tabular presentations are small, the percentages 
38 
m1Js t be interpreted cautiously . Perhaps the sa rest conclusion here is 
r,ha t r~h? level of violence against persons decreased during the riot 
lJ8r iod in all three locational areas. Further, while Tables 22, 29, 
urvJ 36 show proporLiom1te decreases in arrests for the general category 
e;f 8rimes against persons, we may argue that these decreases were a 
f?ms tlon of the increases during the riot period in some of Lhe other 
generc:11 categories , for the total number of arrests for crimes against 
persons in all three locational areas during the normal and the riot 
pedod were 151 and 156, respectively. 
Tables 24, 31, and 38 report arrests for crimes against property 
by time period for the three locational areas. Interpretations are 
again complicated by the very small numbers, particularly in Table Jl, 
which represents the corridor area. Perhaps we should reiterate here 
that the corridor areas evidenced only one arrest for burglary within 
this entire category of crimes against proper ty during the normal time 
period . While this number increased to 82 during the riot period, only 
10 arrests in the corridor areas were attributed to non-burglary cate-
gories , Both the categories of stolen property and auto theft were 
proportionately min:irnal during either time period. In terms of nwnbers 
of arrests, vandalism increased somewhat in the riot areas and remained 
constant in the corridor and non-riot areas during the riot period . 
However, the riot period proportions of vandalism arrests were low in 
all three locational areas. It was only in the non-riot areas during 
the normal time period t hat vandalism arrests were f airly abundant, 
c1cr~ounting for almost 1 out of every 5 arrests within the crimes against 
fJroperty category, Larceny arrests showed dramatic proportionate de-
creases in the riot and non-riot areas during the riot period. However, 
t b u.;e p~rcE:ntagc decreases r eally over state the reduction of arrests 
for larr~eny i f one peruses t he a bsolute number s in these two t ables . 
?crha ps , ,,e may conclude her e tha t when the percent a ges f or stolen 
pruper t y , vandalism, auto theft, and l arceny ar e assessed wi t h an eye 
on the t otal number s from which the perc entages were derived, these 
arrests either r emained r el a tivel y constant or wer e mi nimal i n number 
durlng the normal and r iot periods , espec i a lly when compared t o t he 
burglary - house breaking arrests. Obvious ly and not s urpris ingly , bur-
glc.1ry arres t s dominated t he riot period both in number and pr oport.ion , 
accounting for proport ionately 85,0 per cent, 87.8 per cent , and 82 .5 
per cent of all the arrests ma de in this general category f or the r iot , 
corridor, and non-r iot areas, re s pectively. The increase in number of 
burgl ary arrests during the riot period in all three locationa l areas 
wa s s ubs tantil:ll from a normal period total of 53 to a riot period tot a l 
of ?43. Interestingly, if the burglary arrests within the riot and cor-
rid or areas are crunbined (328), it is surprising to find tha t a signif-
ica nt ly larger number of arrests (415) were made in the non-riot areas 
during the riot period itself. Perhaps, we may conclude, fr om this 
distribution of the burglary arrests during the riot per.;.od, that 
loot i ng acr,ivity wa s not as concentrated as one might initially sup-
pose . Several factors complicate the meaningful interpretation of the 
burglary arre s ts: (1) the United States Attorney' s Office dec i s ion to 
r:harge all looters with burglary with the intention of possibl y r educ-
i ng this charge to a lesser one as more evidence was gathered in each 
"bur glary" case; (2 ) t he self perception of some polic e offic ers 1 
social role during the riot as being ne eded on the s treet s rather 
than filing formal felony charges in the c ourt sys tem, which r es ulted 
ill <.Jome looters being charged with a misdeueanor such as curfci;.; or dis-
orderly conduct. which required less paperwork; and (3) the differential 
~rcesting practice applied to looters by some police officers who, when 
interviewed after the riot, c laimP.d that t hey received no specific; di-
rection:s for handling looters and followed an indivi dualized definit.i.on 
0f the situation. 4 Q.uHe simply, these three complicating factors re-
pre;-;ent contradictory attitudinal peTspectives on the part of those 
persons responsible for maintaining or trying to maintain order. On 
the one hand, the United States Attorney's Office advocated a very 
stringent stance toward looters , while., on the other hand, some members 
of the police force defined their roles less stringently. Finally, 
some officers were unclear about their duties wi. th regard to looting and 
r?eacted i n a variety of possible ways. Unfortunately, it was not pos-
sible to estimate or to determine how many police officers adopted the 
sec.:ond or third posture, because the sample from which the above in-
formation was obtained was at best a Judgemental, non-probability one. 
Suffice it to say that the response by the police to looting behavior 
during the riot period was surrounded by a certain amount of official 
c onfusion . 
Tables 25, 32, and J9 report the arrests for traffic violations 
by t ime period for the three locational area s . The most obvious fact 
in these data was that the riot was accompanied by a considerable re-
duction in traffic citations as a whole in the riot, corridor, and non-
riot locational areas. Secondly, when the two different types of 
trd ffic violations are compared, there were no meaningful proportional 
changes in any of the locational ar eas, no meaningful changes between 
time periods, and no meaningful changes between subcategories . Quite 
91 
.:;im{:Jly, tne riot did not influence traffic citations at all with the 
8x~eption of reducing their number. What was surprising in these data 
w1:J:J the fact that during the riot period ( given the supposed volume 
c,f luoting and arson), the police still resporided with 712 traffic ci-
tations, which accounted for 9.7 per cent of all the arrests made 
(Tuble 49). 
Tobles 26 , 33, and 40 report arrests for crimes without victims by 
time period for the riot, corridor, and non-riot areas. Non-riot re-
luted arrests within this category were minimal both in number and pro-
portiori in all three locational areas . Arrests for drunkenness dominated 
this category during the normal time period, accounting proportionately 
for 66.9 per cent , 56 , 2 per cent, and 71.J per cent of all arrests made 
within the riot, corridor , and non-riot areas, respectively, although 
the corridor area proportion is misleading due to the small numbers in-
volved . During the riot period, arrests for disorderly conduct domin-
1:l ted proportionately this category, accounting for 89,5 per cent, 94,2 
per cent, and 87.8 per cent of the totals for the riot, corridor, and 
non-riot areas , respectively . These statistics are all the more mean-
ingful when one considers that within the three locational areas there 
were a total of 6J5 arrests for crimes without victims during the normal 
period and that this cat egory increased during the riot to a total of 
)731 arrests. Again, this tremendous increase was pr obably due to the 
imposition of the curfew restrictions which, once filed, were classi-
fied as disorderly conducts. 
One very interesting point needs to be made. If we compute the 
perceutage of riot disorderly conduct arrests in the non-riot areas 
from the base figure of the total number of riot disorderly conduct 
92 
?Jr rE:: ~J ~s f or which locational i n f orma t..ion is specif ied , we will find 
1~lvi t, 56 . 3 per ce nt of the disorderly conduct arrests during the r iot 
[A::rlod were n ot made in the ar eas of c oncent rated or s por adi c r iot 
1Jdlnage . I f we c ompute the percentage of riot period arres ts for all 
c; rimes without victims in the non-riot area s using the tota l number 
uf' arrests in this category for which locational inf ormation is speci-
fied as t he ba s e figure, we will f ind that 57.0 per cent of t he arr es t s 
for cr in1es with out vic tims during the riot period wer e not ma de i n the 
areas of c onc entrated or sporadi c riot damage. Further, Tables 22 , 29., 
and 36 r e port collective ly 5503 arres ts made during the riot period f or 
which locational information is available with 3328 arrests (or 60 , 5 
per cent) having been made outside the riot and corridor areas. Fi n-
'.Jlly , i f 3328 arrests (Table 36) were made in the non-riot areas from 
a gra nd riot t otal of 7308 arrests (Table 49), the non-ri ot areas ac -
'! ounted f or a minimum of 45. 5 per cent of the tota 1 arrests. While the 
Qbove mus t appear somewhat redundant, it is equally obvious that the 
polic e responded to a tremendous amount of a ctivity during the riot 
period which was not within an area of concentrated or sporadic physi-
cal des truction--no matter how one compute s the percentage figure for 
Lh8 n on-riot areas . 
There wece two possible violations related to fires--reporting a 
false f ire alarm and arson. During the normal time period, there were 
no arr ests in any of the three locational areas for either arson or 
fals e fire alarms, During the riot period 7 total ars on arrest s wer e 
made (Table 49) w:i.th only 3 specifying locational information and all 
of t hes e three arrests were made in the non-riot areas, 
The following summary statements indicate the major findings of 
9J 
~cmI-,arutive analysis of the arrest data. 
L i:/hile traffic citations dominated the arrest data in all three 
locu tional areas during the normal time period, crimes ?,.rith-
out victims, primarily curfew and disorderly conduct viola-
tions, dominated the riot period arrest data wi"th arrests for 
erimes against property, principally burglary violations, 
ranking a poor second, i n all three locational areas, (Tables 
22, 24, 26, 29, 31, 33, 36, 38, and 40) 
2. The riot time period arrests for crimes against persons indi-
cated a reduction in the degree of violence in crimes which 
result in personal harm. (Tables 23, JO, and 37) 
.3. The looting activity w,rn not as localized as one might suppose 
during the riot time period selected for study. (Tables 22, 
29, 36, and 38) 
4. While the number of traffic citations issued during the riot 
period dropped considerably in all three locational areas, 
traffic citations still accounted for approximately about 
one out of every ten arrests made, although the number of 
citations var ied i ndirectly with the degree of concentrated 
damage. (Tables 22, 25 , 29, 32, 36, and 39) 
5. The police did not or could not respond to the increase in 
fires during the riot time period with any appreciable number 
of arrests for arson. (Tables 22, 27, 29, 34, 36, 41, and 
49 ) 
Evaluation of the a r rest data hypothesis, Before the analysis of 
the arrest data lambda values is pr esented, we must again digress to 
<.1 consideration of the corridor area, which we have ear 11? er argued i s 
94 
:< ?, ?...ri tn inLerpreta tive difficulties. Tables 22, 29, and 36 indi-
: 'J ,f, t-nc total nwnber of arrests for which locational data is available 
'H' the r lot , corridor, and non-riot spatial areas, respectively? Dur-
ing LnE. normal time period, the riot, corridor, and non-riot areas a c-
" :.;UffU::d for , respectively, 14.J per cent, 2 . 7 per cent, and 82 .9 per 
.:'..:: ri\, of all the arrest s . During the riot time period, the riot area 
, r-r.,::;ts increased by 16. 8 per cent to a percentage of Jl. l, the corri-
?.ic.,r area arrests increased by 5.7 per cent to a percentage of 8.4, and 
Uv..; non-riot areas decreased by 22 . L, per cent to a percentage of 60. 5 . 
Uri like the offense and fire data for the corridor areas, the arrest 
d~ r,d proportional changes in these areas follow the patt ern which was 
5 
lvgi~ally assumed in an earlier section of this chapter. However, 
with the exception of the general categorization, all of the collapsed 
~ubcategories of the corridor area arrest data report minimal numbers 
of items and/or the items are primarily found in one cell. Further, 
the general categories of fire-related crimes and miscellaneous crimes 
e videnced very few items in all three locational areas, For the reasons 
,Jiscussed earlier, the lambda values for the collapsed subcategories 
in the corridor areas and the lambda values for fire-related and mis-
c~llaneous crimes will not be discussed. 
Tables L,3 reports the lambda values for the arrest data by general 
Gategory and collapsed subcategory and by spatial area . It has been 
hypothesized that the degree of association between the police's re-
sponse and the selected riot-n ormal time period varies directly with 
the degree of concentrated riot damage. This is to say that the closer 
one moves toward the m1:1jor areas of destruction, the higher the lambda 
value will be, indicating an increasing degree of associat? 
l.OD. 
95 
Tne calculated lambda values for the general categorization indi-
r;a te de ,reasing a:.rnociation in the arrest data as one moves from the 
non-riot areas of least concentrated damage to the corridor areas of 
.sporadic damage to the riot areas of major concentrated damage, thus 
failing to support the .hypothesis. Also note that the degree of asso-
ciation decreases as evidenced by a reduction in error of 46.0 per cent 
for the non-riot areas, 22.7 per cent for the corridor areas, and 15.8 
per cent for the riot areas. Further, the collapsed subcategory of 
~rimes without victims indicates lambda values which do not support 
the hypothesis wi th a reduction in error of 28.2 per cent for the non-
riot areas and 17.0 per cent for the riot areas. 
We have remarked earlier that the riot period studied did not 
affect the traffic citations to any appreciable degree, saving the 
reduction in their number, and the lambda values surely indicate this. 
Similarly, the lambda values indicate an approximately equivalent de-
gree of association in the arrest data for the crimes against persons 
subcategory between the riot and the non-riot areas. 
The subcategory of crimes against property is supportive of the 
hypothesis as the riot areas evidenced a 29.1 per cent reduction in 
er :c?or, while the non-riot areas evidenced but a 9, 2 per cent reduction, 
Further, the absolute numbers in Tables 24 and 38 are surely not mini-
mal as suggested earlier in the discussions of the corridor area lambda 
values. This subcategory is the only one which seems to clearly sup-
port the hypothesis. 
Finally, if one looks at the magnitude of the lambda values dis-
r;ussed, there is a moderate degree of association in on?-hal.f of the 
lambdas, while two lambda values evidenced weak association and the 
% 
rEom1.llning four lambdas show rninimc.11 association, although three of these 
la t+,er four values are applicable to traffic citations. 
While there is not an absolutely clear pattern, with the exception 
of the subcategory of crimes against property, the remaining lambda 
?,alues discussed do not support the arrest data hypothesis if we fur-
tn(:;r disregard the traffic citation lambdas. 
Comparison of the Offense and Fire Data with the Arrest Data 
In this section the offense and arrest data are compared by loca-
tional area. The reader is reminded that the offense da t a are assumed 
to be a partial indicator of the normal and riot situa :.ion. Tne arrest 
dota are assumed to be a par tial indicator of the response made by the 
police to the situation . Further, the rec1der is reminded thc1t some of-
fenses and arrests did not specify locational information on their 
respective records, and therefore, these cases had to be eliminated 
from the locational analysis. 
Tables 3, 9, and 15 report the offense data by general category 
and by time period f or the riot, corridor, and non- riot areas, respec-
tively. T. bles 22, 29, and J6 report the arrests made by general cate-
gory and by time period for the riot, corridor, and non-riot areas, 
respectively . Both the offenses and arrests for miscellaneous crimes 
were minimal in number and proportion in both time periods in all of 
the three locational ar eas . The offense and arrest data classified in 
the general category of " fires " is difficult to compare, because the 
former reflects i n par t ca l ls made to the fire department for assis-
tance as well as reported fire-related criminal activity. The compar-
ison of items related to fire will be postponed until we focus 
,JfJE:cif.ir;ally upon the sUb(!c:1tegories with.in this general classifica-
t,ic).(J. The offem;e data relative to crimes without v.ir.!tims wcis cill 
t;ut rion-f;xist~ent in both time periods clnd all locational areas. Ar-
r~'Jts for crimes without victims dominated all three locational areas 
duriug the riot and were proportionately large in all three areas dur-
ing 'th'~ normal period. There was not one traffic offense reported to 
crnd recorded by the police in any locationa 1 area during both time per-
iods, whi. le traffic citations clearly dominated the normal time period 
,.rnd represented 1.1. 5 per cent of all the riot period arrests for which 
l.01;atiomil information was available. Obviously, the general public 
did not report vie timless crime or traffic violations during either time 
period, while the police response to traffic dominated the normal time 
r,eriod with a substantial number of arrests for crimes without victims, 
and arr es ts for the former and ?the latter exchanged positions during the 
riot period in all three locational areas. Put another way, while traf-
fic and crL~es without victims accounted for proportionately 0.5 per cen t , 
u.u per cent, and 0 . 7 per cent of the offenses reported during the normal 
time period in the riot, corridor, and non-riot areas, respectively, 
these two general categories during the normal time period accounted 
proportionately for 82.6 per cent, 89.8 per cent, and 86.7 per cent of 
cJll the arrests made in the riot, corridor, and non-riot areas, re-
S[Jec tively . While these two categories proportionately represented 
0 ,7 per cent, 6.0 per cent, and 0.5 per cent of all offenses reported 
during the riot period in the riot, corridor, and non-riot areas, re-
::;r;ectively, they represented proportionately 79.8 per cent, 79.1 per 
r..:ent , and 79. l per cerit of all the arreats made during the riot time 
98 
(Jeriod in the riot, corridor, and non-riot areas, respectively. Of-
fern,e:s related to crimes against property remained constant proportion-
cl tely in the corridor and non-riot area s between the normal and riot 
Gime periods, while they decreased in the riot areas. Arrests for 
r..:.cimes against property increased in all locational areas proportiona te-
ly during the riot period. 
Finally, the data clearly indicate that both the offenses and ar-
rests f or crimes against persons proportionately decreased during the 
riot period. While the proportions for crimes against persons from the 
a rrest data were very small during the normal time period in all three 
locational areas, they proportionately decrea sed to almost minute per-
centages during the riot period. In the normal time period, the of-
fe nses reporting crimes against persons accounted for 1 out of every 4 
in the riot and corridor areas, with the non-riot areas accounting for 
b11t 15 .1 per cent of the total. During the riot period, these offenses 
not only decreased in proportion to less than 1 out of every 10 offenses 
but also decreased in absolute number. 
Tables 4, 10, and 16 report the offense data relative to the s ub-
r..:ategories of crimes against persons for the riot, corridor, and non-
riot areas, respectively. Tables 23, 30, and 37 report the arrest data 
for the same variables. Since the percentages of these tables were de-
rived from absolute nwnbers which were very small, perhaps the better 
interpretative procedure would be to comment on the numbers themselves . 
In all three locational areas the total offenses reported diminished 
during the riot period, while the arrests remained fairly constant, if 
minimal , in number. Within the subcategories, the offense data indi-
~ate that the general public reported fewer crimes involving actual 
99 
r~8rs onal injury--i. e., homicides, r apes, aggravated assaults , etc. --and 
f8we r c rimes involving a threat of bodily harm--i.e,, robberies --duri ng 
rJr11;; riot period in all three locational areas . Further, the arrest s 
nwde for these two subcategories either remained constant or decreased 
i n a ll thre e areas during the riot. On the other hand, both the of-
fenses and arres ts for weapons violations--crimes which involved a fu-
tur e potential for personal violence--increased in number in all three 
lu~otional areas during the riot period, Along with the proportionate 
r educt i on in arrests and offenses during the riot in crimes against 
persons, as noted in the preceding paragraph, we may add that on the 
Gont inuum of potential to actualized personal violence, the riot time 
per iod was accompanied by a reductlon in actual harm and an increase in 
r1otential bodily harm. 
Tables 5, 11, and 17 report offenses related to property for the 
r iot , corridor, and non-riot areas, respectively, while Tables 24, Jl, 
and J8 report arrests for the same variables. For the most part, the 
offenses and arrests between the normal and riot periods for vandalism, 
stolen property, and auto theft, were reasonably constant and propor-
tionately minimal when compared to the other types of crimes against 
property, although there were small increases in absolute numbers for 
t he latter two during the riot period within the riot and non-riot 
:.J r ea s within the arrest data. With the exception of a minimal in-
~rease in arrests within the corridor areas, the riot time period wa s 
a~r!ompanied by a rather large reduction of offenses reported and ar-
r 8st s made for larcenies. Just as dramatic were the riot time period 
l nc r eases in offenses and arrests for burglary in all three l ocational 
~rea3 . Finally, the data reveal that there was a large number of 
2.00 
: - <:..n::;e;.; rC:portect and arrests made outside the corridor and riot areas 
,.:. '..; '..lTlr;,:ontrated riot period destruction. 
T.:iLle,3 6, 12, and 18 report the offenses for crimes without victims 
..JI.' the riot , corridor, and non-riot areas, respectively. Tables 26, 
/J, and 40 report the arrests made for crimes without victims for the 
,urne vciriables . That the notation "crimes without victims" is appropri-
J V.c L; obvious from the offense data which indicate that tne general 
r.,ul:.i.Lic did not report these crimes in any meaningful numbers during 
e .ither time period, Further, there was a tremendous increase in number 
arid in proportion of disorderly conduct (also curfew) arrests during 
the riot in all three locational areas. 
Tables 7, lJ, and 19 report fire-related offenses for the riot, 
;orridor, and non-riot areas, respectively. Tables 27, 34, and 41 re-
port arrests for fire-related crimes by the same variables. Arrests 
for reporting false fire alarms and arson were negligible during both 
time periods in all three locational areas. The offense data contains 
a 11fire" classification which incorporated the calls reporting the in-
~idence of fires to the fire department within the offense data. Ob-
viously, a call for assistance does not reflect a crime committed . 
But these calls not only daninated this s ubcategory but also all the 
offense data during the riot period, and when compared against reports 
of false fire alarms and arson, which were minimal, indicate that the 
public was more concerned with the incidence of fire than with the 
cause of the fire, assuming that a large nwnber of riot fires were 
deliberately set--a not unreasonable assumption. 
The reader will recall that there was not one offense reported to 
the police relating to a traffic violation during either tl.!Il? e peri?o d 
101 
lri a ny of t.he three locational areas. Obviously, crimes relating to 
t r af l'ir::, if reported, were not rer1orded by the police. Tables 25, 32, 
,Jn d 39 report tne l:lrrests made for movinr; and equipmen L violations for 
u ')th time periods and tne three locational areas, res pee tively. These 
t~ bl?s have been analyzed elsewhere and need not be reiterated here. 
:3uffice it to say that compared with the offense data, traffic viola-
t ions dominated the police I s response during the normal time ryeriod, 
while the riot reduced their numbers and proportions greatly. However ., 
the r::i.ot did not change the proportional distribution of the subcate-
guries and considering the riot situation, traffic citations were re-
markably high. 
The following summary, descriptive statements present the major 
f indings of this sec Lion. 
1. Traffic vi.ofations and crimes without victims are infrequently 
reported during both time periods in all three locational 
areas, while these categories dominate the arrest data in all 
three locational areas during the normal time period. (Tables 
3, 9, 15, 22, 29, and J6) 
2. During the riot time period, arrests for crimes without victims 
clearly dominate the police's response, with arrests for 
crimes against property ranking a substantial second. (Tables 
22, 29, and 36) 
J. In the general category of crimes against persons, the offens es 
reported decreased in all three locational areas during the 
riot period, while the arrests m&de remained reasonably con-
stant, although minL~al in number and proportion. (Tables 
3, 9, 15, 22, 29, and J6) 
102 
4 . Both offenses reported and arrests made for crimes resulting 
ln actual bodily i.njury and for threatening bodily injury de-
creased during the riot ti.rue period with one exception--the 
corridor area arrests, which were constant and minimal in 
number--while both the offenses reported and arrests made for 
crimes with a future potential for bodily injury increased 
somewhat . (Tables 4, 10, 16, 23, 30, and J7) 
5. In the cr~nes against property category, larceny offenses and 
arrests decreased and burglary-house breaking offenses and 
arrests increased during the riot time period, while auto 
theft, stolen property, and vandalism remained comparatively 
constant. (Tables 5, 11, 17, 24, 31, and 38) 
6. In the crimes without victims subcategory, the offenses reported 
in all three locational areas were minimal during both time 
periods, while arrests for drunkenness dominated the normal 
time period and disorderly conduct (curfew) arrests increased 
in number and proportion tremendously during the riot period, 
dominating the arrest data in all three locational areas. 
(Tables 6, 12, 18, 26, 33, and 40) 
7. Reports of fir es dominated the offense data in the crimes re-
lated to fires category during both time periods in all three 
locational areas, while false fire alarm reports decreased 
during the riot period and arrests made for cri.rnes related 
to fires were minimal in number and proportion during both 
time periods in all locational areas . (Tables 7, 13, 19, 27 , 
34, and 41) 
8. Rather large numbers of offenses reported and arrests made 
10.3 
during the riot time period occurred in the non-riot area 
locations. (Tables 3, 9, 15, 22, 29, and J6) 
Evaluation of the comparative hypothesis. It was hypothesized that 
t,.lH:: degree of association betw-sen the police's response and the selected 
riot-normal time period is less than the degree of association between 
the offenses reported and the selected riot-normal time period, In 
other words, when the degree of association between the arrest data in 
tbe two time periods is compared to the degree of association between 
the offense data in the two time periods, it is expected that the former 
would evidence smaller lambda values than the latter. 
Again , for the reasons cited earlier in the evaluations of the of-
fense data hypothesis and tne arrest data hypothesis, certain lambda 
values have been disregarded. But an additional problem occurs in the 
analysis of the final hypothesis. Some general categories in the of-
fem,E: data contained minima l numbers, while the corresponding categor-
ies in the arrest data contained substantial numbers of cases. For 
example , the offense data for crimes without victims contained but J8 
reports (Tables J, 9, and 15 ), while the arrest data for th:is same 
r.!ategory contained 4366 cases (Tables 22, 29, and J6). A quite similar 
pattern may be noted in the traffic data where the offenses reported 
and recorded total not one single item, and the arrest data totaled 
2760 cases (Tables 22, 29., and J6), Further, one general category in 
U 8 offense data contained a great number of items, while the corre-
sponling arrest data category contained minimal numbers of items, The 
offenses reported and the incidences of fires called into the fire de-
partment totaled 2360 items (Tables J, 9, and 15), while the arrests 
made for fire-related crimes were but J (Tables 22, 29, and J6). Since 
104 
n ,;O!nparison between two lambda values, where one of them has been cal-
r;ula ted on a minimal number of cases, would not be meaningful, compari-
s ons between the arrests and offense lambdas for traffic violations, 
crimes without victims, and fire-related items are not made. Similarly, 
a;~ noted previously , the general category of miscellaneous crlmes con-
t ained minimal numbers in both the offense and arrest data, and compari-
s ons are not made. In other words, for the reasons cited above and for 
reasons of comparability, the evaluation of the comparative hypothesis 
will rest solely upon the lambda values for the general classification 
ond the collapsed subcategories of crDnes against persons and crimes 
against property, disregarding the corridor area lambdas values in 
these two subcategories. 
Of the seven sets of lambda values, only one--the non-riot area 
c rimes against property subcategory--minimally supported the stated hy-
pothesis. In the general category in all three locational areas, the 
arrest lambda values were greater than the offense l ambda values, Fur-
ther the riot and non-riot area crimes against persons subcategory evi-
denced greater arrest lambda values than offense lambda values ., which 
also characterized the riot area crDnes against property subcategory. 
Therefore, with one exception , where comparability is possible, 
it may be concluded that the police's response was less consistent be-
tween the riot period of organized response and a representative normal 
tDne period than the general public's reportage of criminal activity 
between these two tDne periods. Put another way, the offense lambdas 
displayed three min i mal (0 .0000 - 0,1000) and four weak (0.1001 - 0.2000) 
ass ociational values , while the arrest lambdas displayed one mir1imal, 
one weak, and five moderate (0.2001 - 0.5000) associational values. 
105 
However, the lambda values iri both Tables 21 and 4.3 indicate that 
,?.er?.: were differences in both the arrest and offense data between the 
,,10 ~ime periods selected for study. Rather, the degree of difference 
,ias gr2a ter amongst the arrest data than amongst the offense data, 
Swmnary 
The final objective in this chapter is to simplify the official 
;..ioJ.L: e stat is tics by eliminating those categories and types of crimes 
,1hl-:!h seemed unaffected by the riot itself. Further , those violations 
or categories which increased or decreased substantially during the 
riot period are noted. While it is certainly conceded that the offense 
cind orrest data presented in this chapter do not reflect, respectively, 
the ~rue crime rate and the complete response of the police to the of-
fense data, we can make two points. First, the offense data must be 
t.:.hat which is reported to the police which the latter filtered through 
their perceptual screen and recorded on the official offense record. 
AS such, these data do represent inputs into the police system, grant-
ing that these inputs are manipulated by the collecting agency. Second, 
the arrest data represent those formally filed apprehensions which the 
police officially made in response to the riot situation. If one i s 
i nterested, as we are, with the police response to one time period dur-
ing the total span of the riot, then such data are important, when com-
pared to a non-riot time period, in indicating the nature of the police 
response. 
The entire general category of miscellaneous crimes may be elim-
inated from further consideration, since both the offenses reported 
a nd arrests made during either time period totaled less than 5 per cent 
106 
0f the total items. 
In the crimes against persons category, the riot time period was 
uccompanied by increases in both offenses reported and arrests made for 
weapons violations. While the proportions are minimal because of the 
tremendous increases in the numbers of violations in other categories, 
the weapons offenses reported increased by 70 items and the arrests 
made by 50 items . Further, robbery offenses dropped substantially by 
167, while the arrests made remained constant and minimal in number . 
Finally, crimes involving actual bodily .harm--homicides, rapes, assaults, 
~ex crL~es , and crimes against family and children--dropped substan-
tially on both the offense and arrest records by 59 and 41, respective-
ly . These data indicate that the degree of violence involved in 
repor ted offenses and arrests made for crimes against persons decreased 
during the riot time period. 
Focusing on the crimes against property category, burglary-house 
breaking offenses and arrests experienced substantial increases of 251 
violations and 690 violations, respectively, during the riot time 
period. Larceny offenses and arrests during the riot decreased, while 
the remaining violations--auto theft, stolen property, and vandalism--
remained reasonably and comparatively constant. 
Since there were very few offenses reported duririg both time per-
5.ods for crimes without victims and traffic, we will focus only upon 
the arrest data for these categories . Traffic arrests decreased tre-
mendously during the riot from a normal total number of 2125 to a riot 
period total number of 635. Just as dramatic were the increases in 
rli.sorderly conduct {and curfew) arrests from a normal time period num-
ber of 161 to a riot period number of 3318. Arrests for drunkenness 
107 
remained comparatively consLant., uccounting for 443 normal period ar-
r~sts and J92 riot period arrests . 
The arson offenses reported to the police increased durine the 
r.Lot by only 28 t.o a t.otal of JO, while the false fire alarm offenses 
decreased. by 38 from a normal time period total of lJJ. The incidenr.~es 
of fires dominated the offens? data accounting for 759 reports during 
the normal t .. i.me period and lJl+l reports during the riot time period. 
There were very few arson arrests and no arrests for reporting a false 
fire alarm . 
In surrunary, the offense data indicate that public repor?tage during 
the riot period was dominated by reports of fires with increases in 
crimes involving a future potential for personal bodily injury and 
burglary-house breaking. Substantial decreases in reported offenses 
for crimes involving actual personal harm and actual threat of harm, 
larceny, and reported false fire alarms also occurred. Interestingly, 
arscn reports were minimal in number. Obviously, the official input 
to the police stressed fires, weapons violations ., and burglary offenses, 
with no apparent concern relative to the causes of the fires. 
The arrest data indicate that the police responded with massive 
us':? of disorderly conduct violations (curfew) and somewhat less use 
of burglary-house breaking. Further, even though there was a consider-
able decrease in traffic citations, the police still issued 635 tickets 
i.n this general category. Finally, assuming an increase in arson dur-
ing the riot--a not unreasonable assumption--, the police were unable 
to respond to this activity with similar charges. 
CHAPTER V 
THE CONCLUSIONS 
Introduction 
This final cha pter is divided into three sections. First , a brief 
::rnmrnary of the major results of the f oregoing is presented. Second , 
f1rrther resec1rch objectives are suggested, dealing with a more refined 
analysis of the ecological dimensions of time and space. Using some 
of the descriptive findings in the previous chapter, a nwnber of in-
ferred propositions are also presented as possible explanations of 
i,he.se results. In addition, some of the seemingly logical contradic-
tions in the data will be mentioned for further analysis. Third, the 
methodologicc1l and theoretical implications of the foregoing are pre -
Gented . 
Major Results of this Study 
The substantive problematic investigated her ein is an ecological 
analysis of a portion of a riot utilizing official statistics as indi-
~~tors of the inputs to and responses made by the police as a formal 
agency of social control. The time period investigated was conceptual-
ized as the period of organized response, as suggested by the disaster 
literature, and the data during this time period was compared to a 
representative normal time period, the latter providing the benchmark 
against which to measure the nature and degree of the chang? in police 
dE[.;artment activity during the riot time period selected for study. 
108 
109 
The spatial dimension was on eptualized as a trichotomy--the riot, 
corridor, and non-riot areas--, as also suggested by the disaster lit-
<::rature . Two b,::isic questions were posited: (1) what degree of dif-
ference existed between the defined riot and normal time periods and 
(2) what kinds of differences were evident? The three hypotheses 
posl ted were applicable to the first question., while the indicators 
of riot3 focus attention on the nature of the differences , as well as 
.simplifying the rather voluminous amount of official data available 
for analysis . 
Relative to the hypotheses investigated, the basic assumption wa s 
that the closer one gets spatially to an area of maximum riot activity, 
the greater the degree of association in both the offense and fire data 
between the two time periods, on the one hand, and the arrest data be-
tween the two time periods, on the other hand, More specifically, it 
wa;3 hypothesized thc1t the degree of association in the offenses c1nd 
fires reported b etween the riot and normal time periods would decrease 
as one moved from the non-riot areas of minimal destruction to the 
corridor areas of sporadic des truction to the riot areas of major 
destruc tion. Whi le the general categorization and the subcategorization 
of crimes against property indicate increasing association, the degree 
of change was relatively smc111. 1 Further, the additional collapsed 
::rnbcategories of fire-related items and crimes against persons evi-
denced comparatively constant degrees of association. Since the data 
dfaplayed no clear pattern , it was adjudged that these data did not 
r:learly s uggest that the offense a nd fire data hypothesis should be r e-
ject8d or not be rejected. It was concluded, however, that on the 
bf:rnJs of the associational values themselves, the offense and fire data 
110 
,Jid m...inifest rather small degrees of differences between the two time 
Derlods in all three locational areas. 
Second, t he arrest data hypothesi s posits that the degree of asso-
ciation between t he normal time period arrests and tue riot time period 
8rrests would increase as one moved f rom the riot to t he cor ridor to 
the non-riot areas, After disre garding the associational values which 
were calculated on minimal numbers of arrests , the remaining values did 
nut .support this hypothesis with the exception of t he subcategory of 
srimes against property. 
Third, i t was hypothes ized that the ass ociation between the normal 
and riot time periods within the arrest data would be less than the 
association between the normal and riot time periods within the off ense 
and fire data. After eliminating lambda values calculated from minimal 
numbers of items and maintaining comparability wherever possible, of 
the seven sets of spatial categories investigated, six s ets of arrest 
and of offense and fire lambda values did not support the stc1ted hypo-
thesis with contradictory evidence only in the non-riot area crimes 
agains t property s ubcategory. 
However, if one looks solely at the mc1gnitude of the lambda values 
in Tables 21 and 43, it is obvious that there were differences in both 
the arrest and the offense and fire data between the normal c1nd riot 
time periods s tudied, although the degree of difference wa s greater 
for the former than the latter. This indicates that the riot was ac-
sompanied by a different type of response by the police, and the na-
ture of that response became the second basic problema tic . 
The police data were simplified by s uggesting that only some types 
of violations dramatically increased or decreased or proportionately 
111 
sontributed grec1tly to the input to the police or their response during 
the riot.. First, the offense data during the riot time period studied 
wc.18 dornina ted by reports of the incidence of fires, although little 
r:oncern was apparent relative to the nature of the causal agent in 
these fire reports. Burglary-house breaking--the police department 
,_: l assification of looting--was also reported in substantial numbers. 
The offense record showed substantial decreases in larceny and false 
f ire alarm reports during the riot period. Within the collapsed sub-
category of crimes against persons, those involving actual personal 
harm (i.e., homicides) and those which actually threatened harm (rob-
beries) decreased substantially during the riot, while weapons viola-
tions increased substantially . It must be concluded that the reported 
level of personal violence decreased during the riot. Strangely, while 
the absolute numbers of reported stolen property violations decreased 
and the absolute numbers of reported vandalism increased during the riot 
period studied , these changes were minimal, hence relatively constant. 
Second, the arrest data was dominated during the riot period by 
arrests for disorderly conduct and curfew violations with burglary-
house breaking clearly ranking second. While traffic citations de-
creased substantially, they still accounted for a rather large number 
of arrests during the riot. Substantial decreases in arrests also oc-
curred during the riot for larcenies., while neither stolen property nor 
vandalism arrests changed substantially when compared to burglary ar-
rests . Further, the police did not respond to the burning during the 
riot with any substantial amount of arson c1rrests. As with the offense 
data, arrests for crimes against persons showed a decrease in the level 
of personal violence . 
112 
F'lnall y , when one compares the offense and fire data with the ar-
r?,:.;:_; 1, 'ii.l La d ur ing either time period, it becomes obvious tha t public r e-
, ,CJ"'tage of c rimes without vic tims and traffic violations were quite 
r~1i nimally r epresented, while the polic e responded principally with these 
;a t<-1gories of crime. 
Sugge s t ed Further Res earch 
The spati.al variable. we have noted earlier that it seems logica l 
to expec t the greatest number of offenses reported and arrests made 
durin g the riot period of organized response to have occurred within 
the ma.jar areas of riot damage. Conversely, one would suppose thc1t the 
l8a s t number of offenses reported and arr ests made would have occurred 
".1i?thin the non-riot areas of least destruction, with the corridor area s 
of oo me destruction falling somewhere between the extremes of the defined 
riot and non-riot areas. 
Tsble 44 reports the total offenses by locational area and by time 
period for all thooe data for which locational information wa s avail-
a b l e. The riot areas accounted for 16.6 per cent proportionately of 
all the normal period offenses and for an increase of 11.7 per cent 
duri ng the riot time period to a percentage of 28.J. While the offenses 
r e ported during both t i me periods in the corridor areas remained pro-
p or tionately constant, the non-riot areas proportionately accounted for 
77.4 per cent and 65 . 5 per cent of all the offenses during the normal 
a nd riot periods, respectively . The same pattern is fairly well. repli-
~ate d in Table 45, which reports the total arrests made by locational 
a r ea and by time period, again for all those items specifying locational 
i nf ormation. The riot areas proportionately accounted for 14. 3 per 
113 
'~ent of the normal period arrests, which increased by 16 .8 per cent dur-
Lng the riot to Jl.l per cent. Further, the non-riot areas accounted 
pruportionutely for 82.9 per cent of the normal period arrests , which 
decreased by 22. 4 per cent during the riot to 60. 5 per cent. The cor-
ridor areas experienced small proportionate values in arrests during 
the riot period with a normal percentage of 2. 7 and a riot percentage 
uf 8 .4. 
Three points need to be made about these data. First, as stated 
before but shown quite clearly in these tabular presentations, the riot 
activity, while proportionately increasing in the riot locational areas 
and proportionately decreasing in the non-riot areas, was nonetheless 
s ubstantial in the latter. Second, while the proportionate changes be-
t ween the normal and riot time periods are in the directions deemed 
logical, they are quite minimal in degree. In other words, the non-
rlot areas accounted for well over 50 per cent of all the offenses and 
fires reported and arrests made during both time periods. Third, it 
seems rather strange that the proportions within the corridor areas 
were as small as they were, 
There are two possible implications resulting f rom the locational 
~nalys is. The first is the substantive one which declares that the riot 
activity was more dispersed than one might logically imagine. The 
second, which will concern us here, is a methodological one and is most 
sur ely interrelated to the first. Perhaps, the riot and corridor areas 
were too rigidly defined within the present analysis. The assumption 
was to enclose the areas of major and sporadic destruction, as defined 
in a report by the District of Columbia Redevelopment I~nd Agency and 
the National Capital Planning Commission, in the center of a two- block 
114 
'".i,Jius of geographic space, thereby generating the riot and corridor 
'..l r?,~:..i,c.;. This is to say that the defined riot and corridor arec1s geo-
r~r:.1 phi~a 11.y encompass a very small area when compared to the non-riot 
=:Jr~~s (see Appendix D). The absolute numbers in Tables 44 and 45 sug-
g<j<JI:. this possibility. But with so great a proportion of offenses re-
~orted and arrests made falling outside these locational areas, it is 
::;,.1ggested that the locational specification be subjected to a more de-
i.,cdled future analysis. For example, it would be interesting to know 
WDtther or not the majority of the non-riot area offenses and arrests 
fE:11 within , say, five blocks of the defined riot and corridor spatial 
~reas er whether or not they were randomly distributed throughout the 
non-riot areas . What would be necessary here is a computerized mapping 
?.;rogram which could sort the data items into very small geographic 
un i ts--i . e. , two-block square areas--, into the normal and riot time 
peri od, into the offense or arrest category, and into the specific 
~riminal category . 
In defense of the defined locational areas, it would be difficult 
to refute the assertion that the major destruction to property occurred 
in areas other than those surveyed by the two above mentioned District 
of Columbia agencies. Further, while fires, for example, were reported 
in the non-riot areas, ?these fires might represent burning wa ste cans 
:Jnd brush fires rather than the major conflagrations which were evident 
within the riot areas . Unfortunately, as with burglaries and other 
crimes , the seriousness of any specific fire cannot be determined from 
the available data sources . In s um, there seems no logical reason to 
discount the defined locational area s analyzed above, although a word 
of caution and a suggestion for future investigations seem appropriate. 
115 
Given that a more detailed investigation of the locational variable 
would be basically supportive of the locational assumptions made, there-
by justifying the operational definitions of space utilized in this study, 
~ number of i nteresting questions appear. For example , as indicated by 
the official records, does there appear to be qualitatively different 
~inds of police responses amongst the four major areas of riot destruc-
tion? Sine the southern portion of the 14th Street riot area was very 
~lose to the central business district and several major federal govern-
mental installatimis and since the H Street riot area was primarily ci 
neighborhood commercial center for the surrounding ghetto population, 
the police might have pursued qualitatively different kinds of responses. 
Further , while it was not a focus of the present resecirch, the arrest 
data alone contains, where known, the home address of the individual 
arrestee . It would be interesting to compare the distance between the 
home address and the place of apprehension for each arrestee under the 
hypothesis that the people arrested during the riot were residents of 
the neighborhood. The results of this kind of analysis would be partic-
ularly interesting for crimes against property and crimes without victims 
and relatively meaningless for traffic citations, so some selectiv:i.ty 
would hqve to be exercised in deciding which crimes to include in such 
an ;,rnalysis. 
The temporal variable. In the original formulation of this re-
S"Jure;h , the idea of a day by day analysis in which t he first riot day 
would be compared with its representative normal day and which would 
also i.n~orporate the general and collapsed subcategories of the offense 
~nd fire data and the arrest data by the three locational areas was 
sntertained. It doesn't take a plethora of imagination to realize that 
116 
J. :11 an '-rnalysis would generate a very large and cumbersome number of 
tJ l>1-E:,:i und for this reason it was discounted, However, as was mentioned 
i-r, an earlier chapter , too often studies of riots tend to dichotomize 
-Jr ot,herwise oversimplify the data. For example, it has been argued 
-t:tiut studies which characterize the general public as falling into one 
vt' the other category of "rioters" and "non-rioters 11 approaches the sub-
j ec..: t matter in a naive manner. Therefore, the author would be remiss if 
c;ome suggestion of a more detailed analysis of the time variable was 
not included. 
With the data simplifications suggested, a more detailed investi-
gation of the temporal dimension is possible. It is suggested that the 
next topic for specific analysis be the nature of the distribution of 
the offense and fire data and of the arrest data by hours of the day 
?throughout the defined eight day organized response period of the riot. 
One would expect variable quantities of offenses reported and arrests 
rnad8 from day to day and from hour to hour throughout any day. Assum-
ing that these data partially reflect the inputs to and reactions made 
by the police, a tentative analysis indicates that the riot in Washing-
ton during April of 1968 did not follow the pattern of riots in other 
2 
major metropolitan communities during the decade of the 1960's. The 
activity reported to the police and the response made by the police 
seem to have occurred primarily during the daylight hours. 
By pursuing these units in the temporal dimension, we may be able 
to determine not only the fluctuations in police activity but also de-
velop a feeling for the peak of the riot activity. That these fluctu-
ations occur is indisputable, but movement toward a descriptive riot 
curve of r esponse activity by the police would perhaps be possible. 
117 
Further , it would be interest ing to subdivide the riot areas into their 
,~,::imponent parts of lltth Street, 7th Street, H Street, and SE Washington. 
While these areas are generally recognized as the areas of major riot-
related destruction, we do not know if these areas simultaneously ex-
p<,rrien~ed riot a~tivity or if they were more or less four independent 
riot zones which coincidentally happened to be contained within the 
same political jurisdiction. Also of interest would be the distribu-
tion of specific offenses and arrests by days and by hours of the day. 
For example, on e might hypothesize that the traffic citations issued by 
the poli ce occurred primarily during the non-curfew hours of t he day 
and after the riot activity peak toward the end of the defined organized 
response time period. Also, if the majority of disorderly conduct classi-
ficat i ons appeared during the curfew hours, we might hypothesize that 
these arrests were primarily responses to the newly enacted curfew re-
strictions against personal movement. 
Inferred propositions for further testing. If one looks closely 
at the numbers and proportions of some criminal categories, a number of 
seemingly logical contradictions appear in the data , It is to some of 
these that we now direct attention. 
As noted in the preceding chapter , a number of factor s make the 
interpretation of the burglary statistics difficult, We have argued 
that the upper echelons within the District of Columbia Metropolitan 
Police Department in con junction with the United States Attorney's 
Office decided to charge all looters with burglary- -a felony charge and 
a aerious one in terms of penalty- -with the idea of r educing the charge 
if clear evidence was not available to anticipate the probability of 
~onviction. However, a rather poorly performed ex post factum study 
118 
by the District of Columbia Cormnittee on the Administration of Justice 
under Emergency Conditions i ndicated that some police officers self-
n,ported that their instructions for handling looters were unclear and 
that they used their own discretion relative to looting activity or 
that some police officers defined their social role during the riot as 
one of being needed on the streets rather than within the court build-
ing3 filing formal felony chclrges. Rather, the latter group of officers 
minimized their off the street time by charging looters with dis orderly 
conduct or curfew violations , which are misdemeanors and necessitate 
much less paperwork than felony charges, Further, data collected by 
the Washington Civil Disorder Survey indicated a total of 1352 business 
establishments and 1199 real properties3 were damaged during the riot. 4 
Let us further assume that all the arrests made by the police during 
the riot period of organized response for burglary (937), larceny (42), 
stolen property (85), vandalism (51), and arson (7)--a total of 1122 
arrests--occurred in response to real properties and businesses damagect. 5 
If we make the further rather unrealistic assumption that there was only 
one person performing only one of the legally defined criminal activi-
ties per real property damaged--the smaller of the figures for damaged 
properties and businesses - -, then we must conclude that the police did 
not respond to crimes against property with a maximum utilization of 
charges for these crimes . Put quite simply, the official statistics 
indicate that the rather stringent policy suggested by the United State" 
Attorney 's Office and the upper echelons of the police department was 
not pursued by the police officers on the streets. 
Further ., if one looks at the disorderly conduct arrests, whfoh 
also include curfew violations, the police charged 4501 persons with 
119 
these crimes during the riot period of organized response . v/hen one 
compares the crimes against property subcategory arrests and arsor1 ar-
rests with the disorderly conduct arrests, it becomes even more obvious 
that the police officer on the street was not adhering to the officially 
defined policy. 
A number of quali.f'i.cations must be placed on this interpretation, 
however, The reader will remember that in the event that some one per-
son was arrested for multiple violations, he was classified once and 
only once for the most serious crime from all the alleged infractions. 
Hence, if a person was arrested for burglary and assault with a deadly 
weapon, that person might very well have been lost in the above statis-
tics to the more serious criminal category of assaults during the clas-
sifica tion process within the police department I s data processing 
division. Second, the reader will also recall that the arrest record 
does not conta in juvenile offenders, who might have been quite active 
in the looting and burning of the c ommercfal property damaged during 
the riot . Third, the police might have made a number of informal and 
unofficial responses to crimirn~l activities during the riot . 1md 
fourth, there could very well have been too much riot activity for the 
poli ce to make any kind of response to the majority of these activities. 
While it is necessary to probe more deeply i nto the above men-
tioned factors for more conclusive evidence, the discrepancies in the 
data noted above do suggest the tentative hypothesis tha t ?r:;he police 
officers on the streets during the Washington riot of April, 1968 pur-
sued a response policy of r emoval of the causal riot agents rather 
than of attempting to punish such offenders. In other words, the data 
suggest a massive use of restrained force designed to remove rioters 
from the scene of the riot by the simplest and most efficient legal 
me,rns pos sible--dis orderly conduct and curfew violations. 
Given that this was the orientation of most of the police of f i~ Er~, 
t hen a more difficult question becomes the reason for this arrest pos -
ture. Did the police perceive a majority of the rioters as basically 
law abiding individuals, who happened to be caught up in the "spirit" 
,:;r contagion of a r iot situation? Or did t hey really feel that the 
priority should be placed on their function in the streets which gener-
ated a reluctance to file more time consuming felony charges? Or, per-
haps, they felt it easier to gain a conviction for curfew violations 
t han for some more serious charge. The reader will recall from the pre-
f ace that the military personnel aided the police by holding offenders 
until the police arrived, while the latter filed any forma l charges 
deemed appropriate. Since the rules for evidence leading to conviction 
for criminal charges in the court system are quite rigid, then perhaps 
policemen charged the misdemeanor of curfew, because one would only 
have to prove that the defendant was out ori the street at a forbidden 
time of the day. 
Let us now focus upon the police arrests for disorderly conduct 
(curfew), burglary-hous e breaking, traffic related violations (which 
include moving violations., non-moving violations, and intoxicated driv-
ing), and drunkenness, irrespective of whether or not locational in-
formation was available (Tables 48 and 49), since these crimina l 
sategories ranked from one to four in number of arrests made during 
the riot period of organized response. v/hen compar ed to a representa-
tive normcll time period, burglc1ry and disorderly conduct arrests in-
~reased substantially, and traffic citations decreased substantially, 
121 
while a rrests for drunkenness remained c ompara"tively constant in number , 
i:llth ough they halved proportionately. Interestingly, if we add the 
tuLc.il number of arres ts made wit hin these four categories , the riot 
[Jeri. od accounted for 91.8 per c ent i:lDd the normal period accounted for 
8/t .6 per cent of all the arrests made in these respective time periods. 
',rJe mus t c onclude that the police made quite consistent r esponses between 
t.he t wo time periods studied for these violi:ltions as a whole. Granting 
Lhat s ome categories increased and some decreased during the riot, the 
puint r emains that the police response during either time period c on-
se nt ra ted upon these four categories. 
A number of subpoints are interesting also. Even though traffic 
:; i t ations decreased from a normal time period number of 2317 to a riot 
r,ime period number of 712, the latter figure was still large enough to 
r arik third during the riot period behind disorderly conduct (curfew) 
c1 nd burglary arrests, In a time when public attention was directed 
toward the more visible riot indicators of looting and burning, the 
pol.:i.c e still continued to make substantial nwnbers of traffic r e lated 
a rrests . To a lesser degree the same may be said of the arrests for 
drunkenness. Second, the reader will note that public reportage of 
c rimes for disorderly conduct and curfew, traffic violations, and 
drunkenness were obviously minimal in number during both time periods 
(Tables 46 und 47), while these categories accounted for 3124 normal 
t ime period arrests and 5770 riot time period arrests (Tables 48 c1nd 
49). Not only did the police response seem to be reasonably con s istent 
during both time periods, but also it seemed directed toward thos e 
cat e gories of criminality which presumc:1bly failed to excite the public 
t o the degree that they were reported to the police. These data lend 
122 
f' ur't.ber .:;upport to Dynes contention, as noted earlier, that the police 
:Jr e most r~omfor Luble in their traditional role behaviors. 
It would be interesting tc, know if the majority of tho traffic; 
,:ita tions occurred at times when t.he daily fluctuations in other riot 
ci r res ts were low. 'vie might hypothesize that traffic citations were 
issued when the burglary and disorderly conduct arrests were minimal. 
Or, perhaps, they were made toward the end of the period of organized 
r ea ponoe, cifter the peak time of riot-relcited activity. Does the volume 
c,f' trcifflc cita Lions indicate ci police policy related to the generation 
of revenue for the municipality? Further, are arrests for drunkenness 
s o prevalent because they represent a relatively easy mechanism by which 
the police officer attains his monthly arrest quota? These particular 
criminal violations are interesting, because they do not seem to elicit 
much concern from the general public. 6 
As noted in the review of the literature chapter, many sociological 
studies posit some variant of the basic frustration-c:iggression theme as 
the sause of riots. If, as some argue, the fire and police departments 
c1re indicative of the white power structure's quasi-military occupation 
of the black ghetto areas, it would seem as if false fire alarms would 
increase if for no other reason than the pure harassment of these de-
J:.,a rtments, Since the vast majority of fire box alarms c1re false, the 
[JOlice monitor the fire department I s alarm communication center and 
automcitically respond to the scene of a fire box alarm on the assumption 
thcit it will be false. 7 It would appear logically that false fire alcirm 
reports would increase during a riot, but, as noted before, they de-
crea::.;ed in all three of the locationcil areas studied. 
Did the general public curtail the reporting of their normal rates 
123 
POof?  t f a.1  se fire alarms, because they perceived the incidence off.ir e as 
Or, perhaps, a normal number of false fire alarms 
entially harmful? 
,,,ere reported but because of the tremendous number of fires during the I 
:riot ' th department was almost assured of finding a real fire e fire 
close which the call originated that the enough to the f1?re box from 
can .,,,as not recorded as false by the police. ivfaybe, because of the 
incre ased vo .. ume of reports made to the police plus the incrE.:ased street 1 
<1Ctivit not able to respond to fire box alarms in Y' the pol:i.ce were 
their  normal fashion. 
Meth odological and Theoretical Implications of this study 
The preceding has established a number of major descriptive points 
con ce.rning First, riots are complex and variable phenomena. riots. 
'trh.' ner; y ar(.?)  c c,nplex in that many different types of behaviors are performed 
Not only are there the causal fac-
oughout the duration of a riot . 
'to:r 8 lead to riots, but also there are the responses made by per-\v'hich 
son s ana agencies within and without the affected community. Riots are 
bV ~tl'.'iabl e i. n the sense that the actions performed are not evenly distri-
,.U tea t hroughout the time span encompassed by the riot. In other words, 
l.ots e V.i dence peak times of activity followed by depressions in activ-
iltiyt.  Second ? it has been argued that the majority of the soci? o1  og?i ca1.  
c er-ature dealing with riots has been concerned with the independent 
ausa1 variables which lead to tbe dependent variable of the incidence 
Of t'l.. ots or s0111e aspect of riots neglecting other aspects of riots 
Wltich h Passed vir?tually unstu, died. Third, it was noted that many 
ave 
conceptualization of the 
OPbr evt ous research studies oversimplified the 
Jects of the research in these studies. 
--
12L+ 
It is maintained spec i fica lly in this thesis that the ecological 
dimens ions of space and time , suggested by the disaster literature, pro-
vi,JE:o a m~ciningful and useful interpretative schema for the analysis of 
off i~ial statistics as partial indicators of the degree of difference 
and of the nature of the differences in the respons e behavior of one 
ag1:;ncy of social c ontrol to the activity of rioters thems e lves . The 
disaster literature s uggests that qmilitatively different b ehaviors 
are to be expected a s one moves through the various sequential time 
periods and the various spatial zones, Specifically, we have focused 
upon one time period--the organized response period--and upon three lo-
sational areas--the riot, corridor, and non-riot areas. W'nile the des-
cript ive findings reported in the r esults chapter are relevant to the 
nature and degree of the police response, we have argued further in 
this final chapter that the analysis of these findings s uggest a number 
of se,;mingly interesting cJnd relevant inferred propositions which are 
hypothesized as indicative of other aspects of police behavior and the 
beha vior of persons wh o are not police officers. In other words , two 
major s ubstantive contributions are made herein. First, we have par-
tially described the nature and degre e of difference in the police's 
response b e tween the period of organized response and a representative 
normal time period, and second, we have maintained that the official 
statistics, while not providing conclusive evidence, do s uggest re le-
vant subs tan ti ve directioris for future analysis. 
The reader might well ask why the police r esponse to the riot has 
been the focus of attention. It is argued that the police are the one 
major social control agency most likely to become involved in any riot 
and are therefore one of the principal components in the community' s 
'" "'o~_ _ ,,_. .... ~,-, ... 
~ . 
I 
125 
response to the conditions created by riots. Further, while the analy-
sis of police statistics is accompanied by a number of confounding fac-
tora, as noted in the third chapter, these data are important data for 
a number of reasons. First, they are quantified data. vfnile riots 
bi:lve been extensively studied, most of the available data has been 
qualitative, and the more comprehensive studies have been essentially 
attempts to recreate the chronology of riots in the style of "f'fm Blor:ks 
from the 1~rnite House or Rivers of Blood, Years of )2arkness. We know 
that riots are accompanied by increases in arrests, but the utilization 
of official police statistics makes it possible to incorporate addition-
al variables--such as time and location--into the analysis of this as-
per..:t of a riot situation. Second, police data are continuously collected, 
which allows the researcher to compare the riot data against non-riot 
dci ta as done in this thesis. While other sources of riot data are ex-
tremely interesting, they often lack continuity of collection which 
makes the establishment of a benchmark impossible. Third, and perhaps 
most important, police statistics are available for any riot anywhere 
in the na t:i.on. The uniform crime reporting program used by police 
Jurisdictions allows the researcher to compare the responses of the 
Washington police force to the responses of other police forces in 
other political jurisdictions. It is argued here that with the data 
simplifications and indicators suggested in this thesis and with the 
spatial and temporal organizing conceptualizations from the disaster 
rnoclel, comparative studies in police response to riots can be done 
between metropolitan areas. Although the interpretations must be 
~arefully considered because of potential official biases, the availa-
bility, continuity, and comparability of these quantitative data 
126 
:~ -__,,_1r,:.: 0c:0 should not oe underestimated--a point which requires major ern-
:,'ct,_:i ::, is. In the urea of collective behavior, where data rarely possesses 
',t,'c: r~J:-iuract(~rlf;tics deemed most desiraole by researchers, police sta-
'~ ::_::_: Lir~s are an in1portant source of systematically presented information. 
In sum, it may be argued that the analysis of police statistics pro-
?;ir18 us with information which we did not know previously. This is to 
:_;uy thut we can state when and where all different types of officially 
rer_:: orded criminality occurred. While many of the descriptive findings 
rriight c1 ppear as an exercise in the painful elaboration of the obvious, 
3 sm1:, findings appear unusual, for example the quantity of arrests made 
outside the areas of major and minor riot destruction, and therefore 
wc:.1rrant further investigation. 
1vlethodologica lly, there are a nwnber of problems which require a 
?~crtain amount of difficult rethinking. Since the disaster model di-
m1:;nsions have not been applied to the analysis of riots, the operation-
al definitions of the time periods and the spatial zones are made more 
difficult. One can not utilize the operational definitions in other 
disaster studies because the nature of the "disaster"--in this case a 
riot--has not been so studied. Further, while the data are quantified, 
v1hen one compares one criminal category to another criminal category, 
0ne can only asswne a nominal level of measurement. With the substan-
tive problem of association, the number of statistical measures avail-
able which cari incorporate nomirial data are minimal. Lambda was selected 
as the best, for reasons cited earlier, but one must realize that this 
statistic is a weak measure of association. This means that the lambda 
measure will underestimate the amount of true association in the data, 
J\lso it is argued that small numbers in any table will generate a lambda 
127 
F,J. 1-'--1s v1hid1, if interpreted, is without substantial meaning. Yet, more 
~,r__,c,itiv<c:ly, we cun suy the data Jo permit the utilization of some sta-
',i.stieal procedurea which generate an additional level of information 
';1rilC:h has been hitherto unavailable to many students of riots. 
In the review of the literature chapter, allusions were made to 
:~ '..JI!te of the other dimensions of the sociologica 1 disaster literature? 
.-lhil<:, these dimensions have not been the specific focus of this thesis, 
W?:: would like to argue that they appear to be relevant to the analysis 
c.;?.' :r?iots, Most definitions of disaster emphasize a cessation in the 
ric.;rmDl functioning of a social system--a disruption which is infrequent 
?.::nough that this disruption and the reaction it engenders have not be-
'-! ome incorporated into the normative structure of that system and which 
'--1sua lly requires some measure of extra-community assistance before the 
social system can return to a state of normalcy. Students of disasters 
,sse these phenomena as events which happen and which then necessitate 
::Jome kind of action on the part of the conununity. Therefore, the dis-
8 ster literature concentrates attention upon the social system's re-
sponse to the effects created by the disaster agent. More specifically, 
it directs attention toward the variety of actors and agencies which 
fYc:rforrn some kind of response to the disaster situation. 
It is argued here that the disaster conceptualization is poten-
tially valuable to the sociological student of riots, because it focuses 
attention on a variety oi' structures in the community which have hither-
to been neglected, such as the agencies which provide "relief" functions 
for the victims of the riot, and which are operative during a riot. 
This means that the disast.er approach incorporates into one systematic 
analysis seemingly disparate elements of the social system and moves 
128 
cJ.oser ?t o a 
.Si tu<:it?.?: 
more sophisticated multivariate conceptualization of the 
r.on. r 
,',l .l ct ? n Other words, the disaster literature is not in contra-
:ton t-lith th
ltat12et- e Previously mentioned causal interpretations of riots. 
' it :i_c, c,U l 
Ol'Je I,., "' '' PP emental to these conceptions in that it organizes 
,, Perception 
f\Jct of :l:'iots toward the nature of the reaction of the af'-
ect social 
system. This is a point which requires major emphasis, 
In su 
coin m, bec<l Use of the neglected aspects of riots as well as the 
PJ.e:ici.ty 
.?1 .zcit~ cUJd Variability of riots, a movement toward tlie conceptual-,on 
Of riot 
:fo:t'e s as disasters would be useful and meaningful. There-
, a lllo1:?e , 
th e d. comprehensive view of riots may be attained by combining 
:toast . 
J. l?s l1ed er h. terature conceptualizations with the more well estab-
caus.,~1 . . t 
J: t it :.t.nterpretations already in the present riot htera ure. n.,... 
q.,_&'Ueq b 
cl:, trect . ere that this approach to riots is a potentially fruitful 
.:ton 
t.s for further research. icl:l) It is further argued that this thesis 
initi 1 
a exploratory effort in such a direction. 
APPENDIX A 
NOTES 
.. r.,? 
,/  .c:i/ ": 
Notes to the Preface 
1. See Office of Civil Defense, Government of the District of 
Columbia, Operation Bandaid One: April 4 - April 12, 1968, P? 1. 
2. See .!.!?.!..2.?, P? 3. 
3. See Ben W. Gilbert, Ten Blocks from the White House : Ana tom_x 
of the Washington Riots of 1968 (New York, New York: Frederick A. 
Praeger, Publishers, 1968), p. 19. 
4. See .!J2..!E_., P? ?7. 
5. See Office of Civil Defense, Operation Bandaid One, P? 9. 
6. See Gilbert, Ten Blocks from the White House, P? 29. 
7. See .!J2.!..g_., P? 41. 
9. See~., pp. 46f. 
10. See Government of the District of Columbia, Report on Civil 
Disturbances in Washington, D. c. April, 1968, Section on schools. 
11. See Gilbert, Ten Blocks from the White House, P? 55. 
13. l.!u.??, P? 81. 
14. Ibid., pp. 7lf. 
15. See Government of the District of Columbia, Report on Civil 
Disturbances, Section on sanitation. 
16. See Office of Civil Defense, Operation Bandaid One, P? 17. 
17. See !.!.?.!.?. ' pp. 18f. 
18. See Gilbert, Ten Blocks from the White House, P? 88. 
19. See Office of Civil Defense, Operation Bandaid One, P? 20. 
20. See Gilbert, Ten Blocks from the 'White House, P? 85. 
21. See !.!.?.!.?. , P? 86. 
130 
131 
n. See R. D. Heinl, Jr?' "Washington ?s Three Days of Burning, 11 
Fire Engineering (July, 1968), P? 40. 
23. Ibid., P? 39. 
24. See Office of Civil Defense, Operation Bandaid One, P? 26. 
25. See Gilbert, Ten Blocks from the White House, P? 89. 
26. See ibid., p. 104. 
27. See ibid., P? 96. 
28. See Office of Civil Defense, Operation Bandaid One, PP? 28f. 
29. See ibid., P? 30. 
30. See ibid., P? 33. 
31. See ibid., PP? 58f. 
32. See ibid., P? 83. 
33. See Gilbert, Ten Blocks from the White House, PP? 83f. 
34. See ibid., P? 84. 
35. See u. S. Senate, Hearings before the Permanent Subcommittee 
on Investigation of the Committee on Government Operations, Riots, Civil 
and Criminal Disorders--Part 17 (Washington, D. C.: U.S. Government 
Printing Office, 1969), PP? 3174-3205. 
36. See footnote number three in Chapter V for the distinction be-
tween real property and business establishment. 
37. See u. S. Senate, Riots, Civil and Criminal Disorders, P? 3176. 
38. See ibid., p. 3178e 
39. See ibid., PP? 317 Sf? 
40. See ibid., P? 3177. 
41. See ibid. 
42. See !!!.!&?' P? 3191. 
43. See ibid. 
44. See !!!.!&?, P? 3183. 
45. .!!?.!.?. , P? 3207 ? 
132 
46. See i!2.!..Q.?, p. 3126. 
47. See Gilbert, Ten Blocks from the White House, P? 212. 
48. See Office of Civil Defense, Qeeration Bandaid One, p. 102. 
49. See Government of the District of Columbia, Report on Civil 
Disturbances, Section on the Department of Public Welfare. 
50. See Joel T. Broyhill, "Praise of the Salvation Army in Wash-
ington, D. C. During the Recent Rioting Disorders, 11 Congressional 
Record (Vol. 114, No. 87), Item E4471. 
51. See National Red Cross, Red Cross News, Special issue on the 
activities of the first week of the riot. 
52. See Interracial Council for Business Opportunity of Greater 
Washington, Report dated May 10, 1968 and an accompanying letter dated 
August 12, 1968. 
53. See personal communique to Howard Gabennesch from Bertram L. 
Keys, Jr., Executive Secretary of the Montgomery County Commission on 
Human Relations dated July 23, 1968. 
54. See Government of the District of Columbia, Report on Civil 
Disturbances, Section on Citizen Information Answering Service. 
55. Ibid., Section on Department of Public Welfare. 
56. See Office of Civil Defense, Operation Bandaid One, P? 79. 
57. See lE..!.g_., P? 81. 
58. See lE..!.g_., P? 85. 
59. See personal communique to Howard Gabennesch from the Chesa-
peake and Potomac Telephone Company dated August 7, 1968. 
60. Letter of April 1, 1968, from Keven P. Charles of the Young 
Lawyers Section of the Bar Association of the District of Columbia to 
Joseph F. Hennessey, Chairman of the Young Lawyers Section of the Bar 
Association of the District of Columbia. 
61. See William A. Dobrovir, Justice in Time of Crisis s A Staff 
Report to the District of Columbia Committee on the Administration of 
Justice Under Emergency Conditions (Washington, D. c.: U.S. Govern-
ment Printing Office, 1969), p. 6. 
62. See lE..!.g_., P? 7. 
63 ? .!J?.!..2.?, PP? !Of. 
64. See ill.??, P? 15. 
133 
65. See !!ti??' P? 14. 
66. See .!.2.!&?' P? 16. 
67. See ibid., P? 19. 
68. See ibid., P? 21 ?..t passim. 
69. ill!!?, P? 30. 
70. See ~-' PP? 13f. 
71. See ibid., P? 87. 
72. See Gilbert, Ten Blocks from the White House, P? 11. 
Notes to Chapter I 
1. See Clark McPhail, "Civil Disorder Participation: A Critical 
Examination of Recent Research, 11 American Sociological Review, 3 6 
(December, 1971), PP? 1058-73. 
2. u. s. Department of Justice, Federal Bureau of Investigation, 
Uniform Crime Reports--1970 (Washington, D. c.: U.S. Government 
Printing Office, 1970). 
Notes to Chapter II 
1. See John Dollard, Leonard w. Doob, Neal E. Miller, o. H. Mowrer, 
and Robert R. Sears, Frustration and Aggression (New Haven, Connecti-
cut: Yale University Press, 1939). 
2 ? See Leonard Berkowitz, "The Study of Urban Violence: Some Im-
plications of Laboratory Studies of Frustration and Aggression,"~ 
and Rebellion: Civil Violence in the Urban Community, ed. Louis H. 
Masotti and Don R. Bowen (Beverly Hills, California: Sage Publications, 
Inc., 1968), PP? 39-49. See also Ted Gurr, "Urban Disorder: Per-
spectives from the Comparative Study of Civil Strife," Riots and Re-
bellion: Civil Violence in the Urban Community, PP? 51-67. 
3. See Seymour Martin Lipset, Political Man: The Social Bases of 
Politics (Garden City, New York: Doubleday & Company, Inc., 1960), 
pp .. 61-72. 
4. See James A. Geschwender and Benjamin D. Singer, 11Deprivation 
and the Detroit Riot," Social Problems, 17 (Spring, 1970), pp. 457-63. 
5. See Claude Brown, Manchild in the Promised Land (New York, New 
York: The Macmillan Company, 1965), PP? 7f. 
134 
6. See Kurt Lang and Gladys Engel Lang, "Racial Disturbances as 
Collective Behavior," Riots and Rebellion: Civil Violence in the Urban 
Community, PP? l2l-30. See also Jeffery M. Paige, "Political Orien-
t:n t: ion and Riot Participation, 11 American Sociological Review, 3 6 (Oc-
tober, 1971), pp. 810-20. -
7 ? See Irving A. Spergel, "Youth Gangs and Urban Riots, 11 Riots 
Q.nd Rebellion: Civil Violence in the Urban Community, PP? 143 - 56 ? 
8. See Gustave LeBon, The Crowd: A Study of the Popular Mind 
(New York, New York: The Viking Press, Inc., 1960). See also Jules 
J ? Wanderer, 111967 Riots: A Test of the Congruity of Events," Social 
f..roblems, 16 (Fall, 1968), PP? 193-98. 
9. See Robert M. Fogelson, "White on Black: A Critique of the 
McCone Commission Report on the Los Angeles Riots," Political Science 
lliiarterly, LXXXII (September, 1967), 337-67. See also Tom Parmenter, 
"Breakdown of Law and Order: A Firsthand Account in the Words of De-
t:roi t: 's Ghetto Dwellers," Trans-action, 4 (September, 1967), PP? 13-22. 
For critiques of this position, see also E. L. Quarantelli and Russell 
R ? Dynes, "Property Norms and Looting: Their Patterns in Community 
Crises," Phylon, 31 (Summer, 1970), PP? 168-72. And Anthony Oberschall, 
"The Los Angeles Riot of August 1965, 11 Social Problems, 15 (Winter, 
l967), pp. 322-41. And David o. Sears and John B. Mcconahay, "Parti-
cipation in the Los Angeles Riot," Social Problems, 17 (Summer, 1969), 
PP? 3-20. And Robert Blauner, "Whitewash Over Watts," Trans-action, 
3 (March-April, 1966), pp. 3-9, 54. 
10 ? See Lee Rainwater, "Open Letter on White Justice and the 
Riots," Trans-action, 4 (September, 1967), pp. 22-32. And Irving Louis 
Horowitz, "Comment: Black Sociology," Trans-action, 4 (September, 
1967), pp. 7f. 
11. See Marilyn Gittell and Sherman Krupp, 11A Model of Discrimin-
ations and Tensions," Riots and Rebellion: Civil Violence in the Urban 
Community, pp. 69-83. See also James H. Laue, "Power, Conflict, and 
Social Change," Riots and Rebellion: Civil Violence in the Urban Com-
munit~, pp. 85-96. 
12. See Seymour Spilerman, 11The Causes of Racial Disturbances: A 
Comparison of Alternative Explanations," American Sociological Review, 
35 (August, 1970), PP? 627-49. 
13. 12.!.?., P? 645. 
14 ? .!2.i2.? 
15 ?. See Clark McPhail, "Civil Disorder Participation: A Critical 
Examination of Recent Research," American Sociological Review, 36 
(December, 1971), pp. 1058-73. 
16 ? .!!?..!..??, P? 1064 0 
-~...---.-,. -
135 
17 ? Albert K. Cohen, "The Sociology of the Deviant Act: Anomie 
Theory and Beyond," American Sociological Review, 30 (February, 1965 ) 
~_.J> ~ 9 $ 
18 .. Allen D. Grimshaw, "Three Views o f Ur  b an Vi olence: Civil 
Disturbance, Racial Revolt, Class Assault, tt ~R:.!i.!::o~t:.:s:.....:a::,;n::.:.;:d;...:;.R;.;;e_b_e_ l_ l_i_o_n: 
Civil Violence in the Urban Community, P? 117. 
19. Lee Rainwater, 110pen Letter on \lhlte Justice a nd the RiotS," 
Trans-action, P? 26. 
20. E. L. Quarantelli and Russell R. Dynes, "Property Norms a nd 
Looting: Their Patterns in Community Crises," Phylon, P? 182 ? 
21. See Clark McPhail, "Civil Disorder Participation: A Critical 
Examination of Recent Research, 11 PP? t059f ? 
22,. See l!u..??, P? 1068. 
23. l.!u.?.?, PP? t068f. 
24. See ~-, PP? 1070f,. 
25 .. .l!u&?, P? 1071 ? 
26. ~- , P? 1070. 
27. See Nicholas J. Demerath and Anthony F. C. Wallace, "Intro? 
duct ion to Human Adaptation to Disaster, 11 Human Organization, 16 (Sum-
mer, 1957), PP? lf; Allen H. Barton, Communities in Disaster: A 
Sociological Analysis of Collective Stress Situations (Garden City, 
New York: Doubleday & Company, Inc., 1969), P? 39; Hoyt Lemons, 
"Physical Characteristics of Disasters: Historica 1 and Statistical 
Review," The Annals of the American Academy of Poli ti cal and Social 
Science, 309 (January, 1957), P? 2; Charles E. Fritz, "Disaster," f.Qn.-
,1:emporarv Social Problems, ed. Robert K. Merton and Robert A. Nisbet 
(New York, New York: Harcourt, Brace & World, Inc., 1961), P? 655; 
Charles E. Fritz, "Disasters," International Encyclopedia of the Social 
Sciences, Vol. 4, ed. David L. Sills (New York, New York: The Mac-
millan Company and The Free Press, 1968), 202-207; Lewis M. Killian, ml 
Introduction to Methodological Problems of Field Studies in Disasters, 
Committee on Disaster Studies Report, No. 8 (Washington, D. C.: 
National Academy of Sciences--National Research Council, 1956), pp. lf; 
James D., Thompson and Robert vl. Hawkes, "Disaster, Community Organ-
ization, and Administrative Process," Man and Society in Disaster, ed., 
George W. Baker and Dwight W. Chapman (New York, New York: Basic 
Books, Inc., 1962), p. 268; Gideon Sjoberg, 11Disasters and Social 
Change," Man and Society in Disaster, P? 357; Dwight W. Chapman, "Di-
mensions of Models in Disaster Behavior," Man and Society in Disaster 
p,. 319; Ira H. Cisin with Walter B. Clark, "The Methodological Chat- ' 
lenge of Disaster Research, 11 Man and Society in Disaster, p. 30; 
and Aleta Brownlee, "Disaster and Disaster Relief, 11 Encyclopaedia of the 
Social Sciences, Vol. 5, ed., Edwin R. A. Seligman (New York N y k? 
The Macmillan Company, 1931), 161., ' ew or ? 
.,.--~---'--~-: ... .,_ --? 
136 
28 
~, ? See Charle , 
lo lea P. 655; and As E. Fritz, 'Disasters," Contemporary Social Prob-
I Ana1 sf llen H. Barton, Communities in Disaster: A Socio-
29 s of Collective Stre~ Situations ' pp, 40f. &laz ? See Al I 
''Dis Sis of c011 en H. Barton, Communities in Disaster: A Sociological 
astel",11 C0 ective Stsr ess Situations, P? 67; and Charles E. Fritz ntem00 
3o ~ rary .ocial Problems, pp. 677f. 
' 
'l> 
"-hi:! ,L ? See e 
E:, p!Ynet-tc80 J ? L. Quarantelli, ''The Nature and Conditions of Panic," 
afli l:'itz , ''Di ournat O f S i I ;ast ? oc o o , LX (November, 1954 
) , 26 7 - 75 ; Charles 
a. l\lf1<'ltfon, 1 ers Compared in Six American Commun! ties," Human Ore 
'l'he Itarns, "riSummer, 1957)., pp. 6.9; and Charles E. Fritz and Harry 
309 t1na1s of the Human Being in Disasters: A Research Perspective.," 
January 19 e American Academ of Poli ti cal and Social Science, 
3 ' 
5 7) I Po 45 ? 
tn Ot 1. See Ch 
Stl!<f.t Sa.stet's: artes E. Fritz and J. H ? .M athewson, Convergence Behavio,r 
Natt es Report A Problem of Social Control, Committee on Disaster 
ona1 Resea;chNo. 9 {Washington, D. c.: National Academy of Sciences--
32 Council, 1957) ,t- ? 
:...lQa ! 8 ? See A.l 
is of len H. Barton ,?9mmunfties in Disaster: A Sociolo~!.E.!:l 
Collective Stres~ Situations, pp.' 163-79. .. 
l 8 33 111s ? See L 
liti!llcln Of Field elolfs M. Killian, An Introduction to Methodol_ogical Prob-
Stt e Behavior tS udfes in Disastei="s, PP? 5f; and Anthony F ? c. Wallace., 
3
Port tions f in Extreme Si tua,tions: A Surve of the Literature and 
~es!:!' No, 1 (or Further Research. Committee on Disaster Studies Re-
<ll"ch Cou Washington, D. c.: National Academy of Sciences--National 
nci!, 1956). 
to 34. S 
n, ?A?-,Lq ssa ehe  Russ e 11 
usetts: R. Dynes, Organized Behavior in Disaster (Lexing-c D. c. Heath and Company, 1970). 
35 
F'? ? Lelli 
letd Stud: M. Killian, An Introduction to Methodological Problems 
es in Disasters, P? 2. 
36 
? See 'b 
~., p. 3., 
l 37 
~ ? Pp? . See Ch 658 art es E. Fritz, "Disasters," Contemporarl S oc f a 1 Pr o b .. -63. 
3a s 
? ee ib 
--.!,Q., p. 657. 
39 
8 is?  OSfe eC A ll en H. Barton, Communities in Disaster: A S oc i o I og i ca I 
01 Iectfve Stress Situations, P? 49? 
01.>_  to 4o ? See B 
tlta e lv'ashi en W. Gilbert, Ten Blocks from the White House: Anatomi 
eger, ~Ubn ton Riots of !968 (New York, New York: Frederick A. 
lishers, 1968), P? 11. 
...... ......,r--.........._...,_,,~_.. ............. .,__ 
,.,.._.,>c, .. ?~"'- ... ...,..;c.,_,.,. __ __,,._,_. __ ? ........ ._ ................_ __, __. .._._...._._. .......... -_,_ "<-~4. _,,.,_ 
- __ -::__::.--:-.~--?--?- ---- -- ---?. -?-?- ._. -- -,:_-
137 
41. See unpublished letter of April 1, 1968, from Keven P. Charles 
of the Young Lawyers Section of the Bar Association of the District of 
Columbia to Joseph F. Hennessey, Chairman of the Young Lawyers Section 
of the Bar Association of the District of Columbia. 
42. See, for example, Neil J. Smelser, Theory of Collective Be-
havior (New York, New York: The Free Press, 1962). 
43. See Report of the National Advisory Commission on Civil Dis-
orders (Washington, D. c.: u. s. Government Printing Office, 1968), 
P? 47. 
44. See !.!2.!.?.., P? 33. 
45. See Robert Conot, Rivers of Blood, Years of Darkness (New 
York, New York: Bantam Books, Inc., 1967), PP? 7ff. 
46. See Lewis M. Killian, An Introduction to Methodological 
Problems of Field Studies in Disasters, p. 3. 
47. See Anthony F. c. Wallace, Human Behavior in Extreme Situ-
ations: A Survey of the Literature and Suggestions for Further 
Research, pp. lBf. 
48. See Russell R. Dynes, Organized Behavior in Disaster (Lexing-
ton, Massachusetts: D. c. Heath and Company, 1970), PP? 136-49. 
49 ? .!2.!.?.?, P? 141. 
Notes to Chapter III 
1. Each of these three records contains additional variables 
which are not mentioned here as they have not been utilized in this 
analysis. 
2. Very infrequently, one finds an "unfounded" classification on 
this record. For example, in fiscal year 1968, 1383 of the 59,380 
offenses or 2.3 per cent were false or baseless reports. See Annual 
Report of the Metropolitan Police Department, Washington, D. C., Fiscal 
Year: 1968, P? 20. Even in these cases, the police respond until 
they find the report false. 
3. This information was presented by Sgt. Fralin of the Data 
Processing Division of the District of Columbia Metropolitan Police 
Department, who was the chief programmer working on the offense record 
at the time the data were secured by the author. 
. ?,-"'- ---~ ...... .-- --- -- --
138 
4. See Donald R. Cressey, "Crime, 11 Contemporary Social Problems, 
2nd edition, ed. Robert K. Merton and Robert A. Nisbet {New York, New 
York: Harcourt, Brace &.World, Inc., 1966), PP? 141-45. This section 
relies heavily upon this article. 
5. See John I. Kitsuse and Aaron v. Cicourel, "A Note on the Uses 
of Official Statistics," Social Problems, 11 (Fall, 1963), PP? 131-39. 
7. See ibid. 
8. 1El..9.? 
9. !!tl.??, P? 135. 
10. See Report on Civil Disturbances in Washington, D. C. April, 
1968, Section on health. 
11. See District of Columbia Code: Annotated, 1967 Edition {Wash-
ington, D. c. u. S. Government Printing Office, 1967). 
12. This information was presented by Sgt. Fralin and Mrs. Taylor 
of the Data Processing Division of the District of Columbia Metro-
politan Police Department, who, respectively, were the chief program-
mers of the offense and arrest records at the time these records were 
secured by the author. 
13. See Codes of Offenses prepared by the Office of Crime Analysis. 
14. In the previous paragraph, only 30 general categories were 
indicated. The discrepancy is a minor point and explained on page 39. 
15. The reason for placing robbery in this category will be dis-
cussed in the succeeding pages of this chapter. 
16. See District of Columbia Code: Annotated, 1967 Edition {Wash-
ington, D. c.: u. S. Government Printing Office, 1967) Vol. II, Title 
22, Chap. 29, Section 22-2901, 1189. 
17. The latter was a single typewritten sheet of paper with no 
title and some of the typed figures were scratched out and replaced by 
hand written data. While the summary statistics reported for the nor-
mal and riot periods wash out this discrepancy, it is herein noted for 
the reader's evaluation. 
18. J. Edgar Hoover, Director, Federal Bureau of Investigation, 
"Crime in the United States," Uniform Crime Reports--1970 (Washington, 
D. c.: U. s. Government Printing Office, 1970), P? 21. 
19. See Office of Civil Defense, Government of the District of 
Columbia, Operation Bandaid One: April 4 - April 12, 1968, p. 1. 
. -??--~------ -~-- _,_ -- -
?. . .,,- --...... ,....-.- ... --
139 
20. See ibid., p. 3. 
21. See Ben W. Gilbert, Ten Blocks from the White House: Anatomy 
of the Washington Riots of 1968, p. 19. 
22. See 12..i.??, P? 27. 
23. See Office of Civil Defense, Operation Bandaid One, p. 9. 
24. See Ben W. Gilbert, Ten Blocks from the White House: Anatom~ 
of the Washington Riots of 1968, P? 29. 
25. See J. Edgar Hoover, 11Crime in the United States," Uniform 
Crime Reports--1970, pp. 26f. 
26. See ibid., PP? 48f. 
27. See ibid., P? 50. 
28. See Annual Report of the Metropolitan Police Department, Wash~ 
ington1 D. c. 2 1968, P? 25. 
29. See.!.!!.!!!?, P? 27. 
30. See ibid., PP? 28-31. 
31. See ibid., PP? 42-5. 
32. Elmer Hubert Johnson, Crime, Correction, and Society, Revised 
Edition (Homewood, Illinois: The Dorsey Press, 1968), P? 45. 
33. See National Capital Planning Commission, Civil Disturbances 
in Washington, D. C. April 4-8, 1968: A Preliminary Damage Report, 
May, 1968, P? 5. 
34. See Report on Civil Disturbances in Washington, D. C. April, 
12.2.?.. 
35. See the Street Address Map of Washington, D. c. by Robert E. 
LaRue which is for sale from the above mentioned at P.O. Box 182, Sil-
ver Spring, Maryland 20907. These maps are used by the Washington 
Metropolitan real estate brokers and similar maps are available for 
Montgomery County and Prince George's County. 
'--? _,.,..,,._,. __ ..... 
140 
36. We must digress to indicate and explain the major difficulties 
with the District of Columbia Metropolitan Police Department official 
statistics with respect to the locational data. Very few of the items 
on the offense record are without locational specification. But this 
is not true of the arrest data. While some of the locational omissions 
are due to the arrest addresses simply not being reported, a certain 
number of "locations" had to be classified as 11miscoded," since the 
addresses reported could not possibly be where they were listed. 
For example, the arrest data employs a ten digit numerical code 
for the location of the arrest. The first digit indicates the quad-
rant of Washington: NW, NE, SE, or SW. The last digit indicates loca-
tion at an intersection, location within the block, location unknown, 
or out of town location. If the final digit indicates an intersection, 
the remaining eight digits refer to two streets. If the final digit 
indicates a locality within the block, the first four digits of the 
remaining eight identify the street and the second four digits identify 
the house number on that street. One of the possible miscoding errors 
would be a locality designation indicating an intersection and two 
streets which run parallel to one another. Since this researcher can 
not presume which part of the ten digit code was in error, any error, 
even the slightest difficulty in interpretation, was classified as mis-
coded and subsequently dropped. 
37. William L. Hays, Statistics for Psychologists (New York, New 
York: Holt, Rinehart and Winston, 1963), P? 610. 
38. See i!tl.!t?, PP? 596f. 
39 ? .!!tl!!?, P? 613. 
40. Hubert M. Blalock, Jr., Social Statistics (New York, New York: 
McGraw-Hill Book Company, Inc., 1960), P? 232. 
41. William L. Hays, Statistics for Psychologists, P? 603. 
42. Ibid., P? 610. 
43 ? .!!tl..2.?, P? 614. 
44. See Leo A. Goodman and William H. Kruskal, "Measures of Associ-
ation for Cross Classifications, 11 Journal of the American Statistical 
Association, 49 (December, 1954), PP? 735-45. 
Notes to Chapter IV 
1. While a systematic analysis of the items possessing no loca-
tional information is still to be performed, a hurried spot check in-
dicated no serious pattern in the data. In other words, some lost items 
were serious crimes, while other lost items were less heinous viola-
tions. 
141 
2. Remember that the total number of offenses reported for crimes 
against persons within the corridor areas was very small. Hence, the 
58.4 per cent proportional reduction must be interpreted with caution. 
3. See appendix D. 
4. See William A. Dobrovir, Justice in Time of Crisis: A Staff Re-
port to the District of Columbia Committee on the Administration of Jus-
tice Under Emergency Conditions (Washington, D. c.: U.S. Government 
Printing Office, 1969), PP? 9 and 14. 
5. See the earlier section of this chapter which discusses the 
offense and fire data hypothesis. 
Notes to Chapter V 
1. The reader is reminded that for the reasons cited in the results 
chapter, some lambda values were disregarded in the analysis of this and 
the two subsequent hypotheses. 
2. This analysis was done by the author but is not reported herein 
in detail, because, first, it is not part of the defined problematic, 
and second, it was done with less methodological rigor and less detail. 
3. "Real property" is legally defined as "? ?? land, and generally 
whatever is erected or growing upon or affixed to land." See Henry Camp-
bel I Black, M.A., ~lack's Law Dictionary: Definitions of the Terms and 
Phrases of American and English Jurisprudence, Ancient and Modern, Re-
vised Fourth Edition (St. Paul, Minnesota: West Publishing Co., 1968), 
P? 1383. See also definition of 11fixture, 11 ibid., P? 766. According to 
Black, the term "business" has no definite o~ecise legal meaning (see 
ibid., P? 248). Therefore, a certain degree of assumption is necessary 
here. For the purposes of the survey, it would appear that real proper-
ty referred to the land, building, parking lot, sidewalk, and similar 
kinds of materials which are the property of the lessor in a lessor-
lessee relationship. Business establishment referred to the inventories 
(merchandise) and movable appurtenances (cash registers, display counters, 
for example) which the businessman (lessee) uses for his economic liveli-
hood apart from the building itself. Such a distinction, however, would 
not be legally as distinct as assumed above. 
4. See "Summary of Purposes, Methodology and Limitations, and Major 
Findings of the Washington Civil Disorder Survey," Riots, Civil and Crim-
inal Disorders--Part 17, Hearings before the Permanent Subcommittee on 
Government Operations United States Senate, May 27 and 28, 1969 (Wash-
ington, D. c.: u. s. Government Printing Office, 1969), PP? 3174-95. 
S. See Table 49. 
6. Some of these ideas were suggested to this researcher by a former 
member of the Baltimore City police force. 
7s From personal communication with Capt. Miller of the Operations 
Research and Planning Division of the District of Columbia Metropolitan 
Police Department. 
. ,?~------.- ?--
APPENDIX B 
TABLES 
-?- ---?--- --?-----------
?-???- -??? - -~--" "---?~~-?---": ~ _. ._  ~---------??- . 
143 
Table 1. The General Code1 and Description of that Code for Arrests 
and Offenses Recorded by the District of Columbia Metro-
politan Police Department. 
Code Category Code Category 
0100 Homicide 1800 Narcotics and Drugs 
0200 Rape 1900 Gambling 
0300 Robbery 2000 Family and Children 
0400 Aggravated Assault 2100 Intoxicated Driving 
0500 Burglary-House Breaking 2200 Liquor Laws 
0600 Larceny2 2300 Drunkenness 
0700 Auto Theft 2400 Disorderly Conduct3 
0800 Other Assaults 2500 Vagrancy 
0900 Arson 2600 Other 
1000 Forgery and Counterfeiting 2700 Suspicion 
1100 Fraud 3300 Traffic4 
1200 Embezzlement 3500 Other Trafflc4 
1300 Stolen Property 3800 Equipment Violations5 
1400 Vandal ism 3000 Unknown6 
1500 \.leapons 3100 Unknown6 
1600 Prostitution and Vice 3400 Unknown6 
1700 Sex Offenses 2649 False Fire Alarm 7 
1. While it is irrelevant to this research, the reader should note that 
these codes represent the adoption of a standardized identifica-
tion system of crime classification for the United States as a 
whole. This system of reporting crimes will greatly facilitate 
comparative criminological research among different law enforce-
ment jurisdictions. See J. Edgar Hoover, Director, "Crime in the 
United States," Uniform Crime Reports-1970 (Washington, D. C.: 
U.S. Government Printing Office, 1970), 6lf. 
2. This category also includes thefts from a motor vehicle. 
3. During the riot, curfew violations were categorized as disorderly 
conduct violations by the police. 
4. Two series of codes are necessary here for the large number of 
moving violations. Note also that no non-moving violations appear 
in this information. 
5. These are violations of malfunctioning appurtenances relative to 
motor vehicles; l. e., break failure, no turn signal. 
6. The arrest record indicates the codes 3000,3100, and 3400. Since 
there are only three cases of codes 3000 and 3100, these might 
very well be coding errors. However, there were 73 arrests for 
code 3400, which has been classified as unknown, since no such 
code appears in the police system of classification. 
7. The only specific code used in tho riot analysis herein is 2649. 
The number of false fire alarms has been subtracted from the 
remaining violations in the 2600 series. 
144 
Table 2. Number of Disorderly Conduct1 and Curfew Violations 2 During 
the April, 1968, Riot in Washington, D. c. by Date. 
Date Disorderly Conduct Curfew 
April 5 483 253 
6 1232 1116 
7 1132 1024 
8 712 781 
9 468 470 
10 177 165 
11 209 164 
123 88 76 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
2. Data from an unpublished informal handout of the District of Colum-
bia Metropolitan Police Department, no date. 
3. Curfew against free movement of civilian personnel was lifted on 
April 12th, 1968. Regulations governing the sale of alcoholic 
beverages and firearms, however, were still in effect. 
145 
Table 3. Percentage of Offenses by General Category and by Time 
Period for Riot Area Locations.l 
General Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Persons 26.4 7.2 -19.2 
(97) (55) 152 
Property 44.7 32.9 -11.8 
(164) (250) 414 
Traffic 
(0) (0) ---
Crimes Without Victims o.5 0.1 0.2 
(2) (5) 7 
Fires2 25.l 58.B 33.7 
(92) (447) 539 
Miscellaneous 3.3 o.4 -2.9 
(12) (3) 15 
Total 100.0 100.0 
(367) (760) 1127 
1 .. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Data from the District of Columbia Fire Department daily alarm log. 
Lambda ? 0.1288 
146 
Table 4. Percentage of Offenses by Collapsed Category of Crimes 
Against Persons and by Time Period for Riot Area Locations.I 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Potential Threat of 3.1 30.9 27.8 
Bodily Harm (3) (17) 20 
Threat of Bodily Harm 62.9 25.5 .37 .4 
(61) (14) 75 
Actual Bodily Harm 34.0 43.6 9.6 
(33) (24) 57 
Total 100.0 100.0 
(97) (55) 152 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda .. 0.1818 
~"---.,-:...;_,,..,. . ~.;._, ........... ~""-, _...., 
' - ~""1-........,_#,..IJ.~.,."_ ......... ;;...,_ ...... ?' ' '-"~-,--?~~~-,,.,~,,,,. ............. ""'_... ?~,,___~-~~-
147 
Table 5. Percentage of Offenses by Category of Crimes Against Prop-
erty and by Time Period for Riot Area Locations .1 
Category /Time Per Cent Per Cent % Riot ? Total 
Normal Riot % Normal Number 
Burglary-House Breaking 40.9 73.6 32.7 
(67) (184) 251 
Larceny 38.,4 6.0 -32.4 
(63) (15) 78 
Auto Theft 16.5 12.s -3.7 
(27) (32) 59 
Stolen Property 0.6 5.2 4.6 
(1) (13) 14 
Vandalism 3.7 2.4 -1.3 
(6) (6) 12 
Total 100.1 100.0 
(164) (250) 414 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda "' 0. 146 8 
148 
Table 6. Percentage of Offenses by Collapsed Category of Crimes 
Without Victims and by Time Period for Riot Area Locations.1 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Non-Riot Related 100.0 100.0 ----
(2) (S) 7 
Drunkenness ----
(0) (0) 
Disorderly Conduct ----
and Curfew2 (0) (0) 
Total 100.0 100.0 
(2) (S) 7 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Riot time period will contain curfew offenses classified as dis-
orderly conduct violations by the police. 
Lambda .. 0.0000 
149 
Table 7. Percentage of Offenses by Category of Crimes Related to 
Fire and by Time Period for Riot Area Locations.I 
Category /Time Per Cent Per Cent % Riot ? Total 
Normal Riot % Normal Number 
False Fire Alarm 13.0 2.2 .10.s 
(12) (10) 22 
Fires2 85.9 96.2 10.3 
(79) (430) 509 
Arson 1.1 1.6 o.s 
(1) (7) 8 
Total 100.0 100.0 
(92) (447) 539 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Data from the District of Columbia Fire Department daily alarm log, 
1968. 
Lambda .. o.0164 
150 
Tables. Percentage of Offenses by Collapsed Category of Miscella-
neous Crimes and by Time Period for Riot Area Locations.l 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot: % Normal Number 
Fraud ---- ----
(0) (0) ---
Liquor Laws so.o ---- -so.o 
(6) (0) 6 
Varied so.o 100.0 so.o 
(6) (3) 9 
Unknown ---- ----
(0) (0) ---
Total 100.0 100.0 
(12) (3) 15 
le Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda? 0.0000 
151 
Table 9. Percentage of Offenses by General Category and by Time 
Period for Corridor Area Locations.I 
General Category/Time Per Cent Per Cent % Riot? Total 
Normal Riot % Normal Number 
Persons 23.1 9.6 -13.5 
(31) (16) 47 
Property 28.4 31.3 2.9 
(38) (52) 90 
Traffic ----
(0) (0) ---
Crimes Without Victims 6.0 6.0 
(O) (10) 10 
Fires 2 48.5 52.4 3.9 
(65) (87) 152 
Miscellaneous 0.6 o.6 
(0) (1) 1 
Total 100.0 99.9 
(134) (166) 300 
1.. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Data from the District of Columbia Fire Department dally alarm log, 
1968. 
Lambda .. 0.0532 
152 
Table 10. Percentage of Offenses by Collapsed Category of Crimes 
Against Persons and by Time Period for Corridor Area 
Locations.l 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Potential Threat of 6.5 so.o 43.,5 
Bodily Harm (2) (8) 10 
Threat of Bodily Harm 64.S 6.2 -58.3 
(20) (1) 21 
Actual Bodily Harm 29.0 43.,8 14.8 
(9) (7) 16 
Total 100.0 100.0 
(31) (16) 47 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda? 0.3095 
I "' ~/,~6--~ _,_ 
i---------------<?? 
! 
153 
Table 11. Percentage of Offenses by Category of Crimes Against Prop-
erty and by Time Period for Corridor Area Locations.1 
Category/Time Per Cent Per Cent % Riot ... Total 
Normal Riot % Normal Number 
Burglary-House Breaking 42.1 75.0 32.9 
(16) (39) 55 
Larceny 28.9 11.s -17.4 
(11) (6) 17 
Auto Theft 15.8 1.1 .s.1 
(6) (4) 10 
Stolen Property ---- ----
(0) (0) ---
Vandalism 13.2 s.s -7.4 
(5) (3) 8 
Total 100.0 100.0 
(38) (52) 90 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda ? 0 .1233 
}' 's"-/.~;-F~.:'...f,"~Ii,::=:;.?~~;.., :..1,"'?::~:_:,. ., ? .. ~ .. <~ .. ; 
l.-.--..... ......s~~:;:_~~&":'+.t:~~~S!G--t~<:~-::i..~,,..::.._.: - . 
-~=?~~~~--.... ,~;t..,.,'<-...,,.,.........,..,_~.~-~-' ?.-,,.~ ......... ,.?~ ...~ ~~,.??,.,.""-'.,.,.. ?.,' ,,....._,J/4:...~? ..... .:..~.-... 
I ~""--?---?~-......-.......; ..... ..__~.~--... ..,__...,._. - -? - - --
154 
Table 12. Percentage of Offenses by Collapsed Category of Crimes 
Without Victims and by Time Period for Corridor Area 
Locations.l 
Collapsed Category/Time Per Cent Per Cent % Riot .. Total 
Normal Riot % Normal Number 
Non-Riot Related 100.0 100.0 
(0) (10) 10 
Drunkenness ----
(0) (0) 
Disorderly Conduct 
and Curfew2 ---- ---- ----(0) (0) 
Total ---- 100.0 
(0) (10) 10 
1 ? Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2 ? Riot time period will contain curfew offenses classified as dis-
orderly conduct violations by the police. 
Lambda? 0.0000 
155 
Table 13. Percentage of Offenses by Category of Crimes Related to 
Fire and by Time Period for Corridor Area Locations.I 
Category /Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
False Fire Alarm 29.2 3.4 -25.8 
(19) (3) 22 
Fires2 70.8 87.4 16.6 
(46) (76) 122 
Arson ---- 9.2 9.2 
(0) (8) 8 
Total 100.0 100.0 
(65) (87) 152 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Data from the District of Columbia Fire Department dally alarm log, 
1968. 
Lambda? 0.1684 
156 
Table 14. Percentage of Offenses by Collapsed Category of Miscella-
neous Crimes and by Time Period for Corridor Area Locations.1 
Collapsed Category/Time Per Cent Per Cent % Riot? Total 
Normal Riot: % Normal Number 
Fraud 100.0 100.0 
(0) (1) 1 
Liquor Laws ---- ----
(0) (0) ---
Varied ---- ----
(0) (0) ---
Unknown ---- ----
(0) (0) ---
Total 100.0 
(0) (1) 1 
1 ? Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda., 0.0000 
157 
Table 15. Percentage of Offenses by General Category and by Time 
Period for Non-Riot Area Locations.I 
General Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Persons 15.1 9.0 -6.1 
(258) (159) 417 
Property 38.6 36.2 -2.4 
(661) (63 7) 1298 
Traffic ---- ----
(0) (O) ---
Crimes Without Victims 0.1 o.5 -0.2 
(12) (9) 21 
Fires2 43.1 53.0 9.9 
(737) (932) 1669 
Miscellaneous 2.5 1.3 -1.2 
(43) (22) 65 
Total 100.0 100.0 
(1711) (1759) 3470 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Data from the District of Columbia Fire Department daily alarm log, 
1968. 
Lambda? o.0419 
158 
Table 16. Percentage of Offenses by Collapsed Category of Crimes 
Against Persons and by Time Period for Non-Riot Area 
Locations. I 
Collapsed Category/Time Per Cent Per Cent % Riot? Total 
Normal Riot % Normal Number 
Potential Threat of 2.7 35.8 33.l 
Bodily Harm (7) (57) 64 
Threat of Bodily Harm 57.4 29.6 -27.8 
(148) (47) 195 
Actual Bodily Harm 39.9 34.6 .. 5.3 
(103) (55) 158 
Total 
100.0 100.0 
(258) (159) 417 
1
? Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda .. 0.1575 
' )'. ;l;?,l":,-~#,if;.;:-;.,.;?,;..,f :Ci. ......: :;.;J. . , :..~' . .~ .::?~ ?.. 
-~_...::--~it,.q~,~i>_::.,>. ?. ??: ?"'"'~--~~?-'---"---~??~~--:..--..c..~ . -? .... ___ ,. . ,. 
-- -_..,.....~,_:::"~_......-...... _~-
159 
Table 17 ? Percentage of Offenses by Category of Crimes Against Prop-
erty and by Time Period for Non-Riot Area Locations.l 
Category /Time Per Cent Per Cent % Riot? Total 
Normal Riot % Normal Number 
Burglary-House Breaking 37.4 56.2 18.8 
(247) (358) 605 
Larceny 35.1 15.2 -19.9 
(232) (97) 329 
Auto Theft 21.3 21.0 -0.3 
(141) (134) 275 
Stolen Property 0.2 3.0 2.8 
(1) (19) 20 
Vandalism 6.1 4.6 -1.s 
(40) (29) 69 
Total 100.1 100.0 
(661) (637) 1298 
1 ? Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda ? 0.0970 
,~-
r 
160 
Table 18. Percentage of Offenses by Collapsed Category of Crimes 
Without Victims and by Time Period for Non-Riot Area 
Locations.I 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Non-Riot Related 100.0 100.0 ...... 
(12) (9) 21 
Drunkenness ---- ----
(0) (0) 
Disorderly Conduct ----
and Curfew2 ---- ----(0) (0) ---
Total 100.0 100.0 
(12) (9) 21 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Riot time period will contain curfew offenses classified as dis-
orderly conduct violations by the police. 
Lambda? 0.0000 
161 
Table 19. Percentage of Offenses by Category of Crimes Related to 
Fire and by Time Period for Non-Riot Area Locations.I 
Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
False Fire Alarm 13 .8 a.a -5.0 
(102) (82) 184 
Fires2 86.0 89.6 3.6 
(634) (835) 1469 
Arson 0.1 1.6 1.s 
(1) (15) 16 
Total 99.9 100.0 
(737) (932) 1669 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Data from the District of Columbia Fire Department daily alarm log, 
1968,. 
Lambda .. 0.0213 
_--------:--:---:--.::- -- -
162 
Table 20. Percentage of Offenses by Collapsed Category of Miscella-
neous Crimes and by Time Period for Non-Riot Area Locations.I 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Fraud 27.9 54.5 26.6 
(12) (12) 24 
Liquor Laws 23.3 18.2 -5.1 
(10) (4) 14 
Varied 48.8 27.3 -21.5 
(21) (6) 27 
Unknown ----
(0) (0) 
Total 100.0 100.0 
(43) (22) 65 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
Lambda ? 0.1000 
163 
Table 21. Lambda Values for the Offense Data by General Category and 
Collapsed Subcategory and by Spatial Area. 
Category/Spatial Area Riot Corridor Non-Riot 
General 0.1288 o.os32 0.0419 
Crimes Against Persons 0.1818 o.3o95 0.1575 
Crimes Against Property 0.1468 0.1233 0.0970 
Traffic Violationsl ----.-- ------ ------
Crimes Without Victims2 0.0000 0.0000 0.0000 
Fire-Related Items 0.0164 0.1684 0.0213 
Miscellaneous Crimes 2 0.0000 0.0000 0.1000 
1. No violations reported in this category in any spatial area during 
either time period. 
2. Small numbers of violations reported in this category. 
164 
Table 22. Percentage of Arrests by General Category of Violation and 
by Time Period for Riot Area Locations.I 
General Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Persons 5.4 1.9 -3.5 
(25) (33) 58 
Property 8.3 17.6 9.3 
(38) (301) 339 
Traffic 47.1 5.1 -42.0 
(216) (88) 304 
Crimes Without Victims 35.5 74.7 39.2 
(163) (1278) 1441 
Fires ----
(0) (0) 
Miscellaneous 3.7 0.6 -3 ? 1 
(17) (10) 27 
Total 100.0 99.9 
(459) (1710) 2169 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda ? 0.1584 
165 
Table 23. Percentage of Arrests by Collapsed Category of Crimes 
Against Persons and by Time Period for Riot Area Locations.l 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot '? Normal Number 
Potent ia 1 Threat of 16.0 54.5 38.5 
Bodily Harm (4) (18) 22 
Threat of Bodily Harm 32.0 18.2 -13.8 
(8) (6) 14 
Actual Bodily Harm 52.0 27.3 -24.7 
(13) (9) 22 
Total 100.0 100.0 
(25) (33) 58 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda? 0.2459 
s, '~'ff~;?f'.~:<,r.:;;,::~:,,.:.;);J::.:<:..,;t,?C~ ~.:, .. , 
-?- --~~.f.~J;':/,:;c.;;;,,:.,...,..C..:..._::Ca~?-~.~~--------~~~,~~.-....;....  .:...,_.. ??;  __ ,. . ,,.,.,, 
166 
Table 24. Percentage of Arrests by Category of Crimes Against Prop-
erty and by Time Period for Riot Area Locations. 1 
Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Burglary-House Breaking 28.9 85.0 56.l 
(11) (256) 267 
Larceny 60.5 1.0 -59.5 
(23) (3) 26 
Auto Theft 2.6 0.1 -1.9 
(l) (2) 3 
Stolen Property 1.0 1.0 
(0) (21) 21 
Vandalism 7.9 6.3 -1.6 
(3) (19) 22 
Total 99.9 100.0 
(38) (301) 339 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda? 0.2909 
, .......,..._.. _ ~- ?----? -- ~?----_____, __ ,_ _ 
167 
.Percentage of Arrests by Collapsed Category of Traffic 
Violations and by Time Period for Riot Area Locations.I 
Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
M-o"-ing- ----------------------
Violations 93.5 94.3 O.B 
(202) (83) 285 
tqu1Prnent 
Violations 6.5 5.7 -0.8 ----- (14) (5) 19 Tota1 -------------------------
100.0 100.0 
(216) (88) 304 
1~. .:lla:: .:::.::.:.:::..::-::::=====-==========-===-=-=-= 
ta fr 
A om the District of Columbia Metropolitan Police Department 
rrest Record 1968. 
tamb , 
da .. 0.0000 
168 
Table 26. Percentage of Arrests by Collapsed Category of Crimes 
Without Victims and by Time Period for Riot Area Loca-
tions. I 
Collapsed Category/Time Per Cent Per Cent r. Riot - Total 
Normal Riot r. Normal Number 
Non-Riot Related 6.7 0.5 -6.2 
(11) (6) 17 
Drunkenness 66.9 10.0 -56.9 
(109) (128) 237 
Disorderly Conduct 26.4 89.5 63. 1 
and Curfew2 (43) (1144) 1187 
Total 
100.0 100.0 
(163) (1278) 1441 
1 ? Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
2
? Riot time period will contain curfew arrests classified as disorder-
ly conduct violations by the police. 
Lambda .. 0.1703 
169 
Table 27. Percentage of Arrests by Category of Crimes Related to 
Fire and by Time Period for Riot Area Locations.I 
Category /Time Per Cent Per Cent % Riot? Total 
Normal Riot % Normal Number 
False Fire Alarm ---- ---- ----
(0) (0) ---
Arson ---- ----
(0) (0) 
Total ---- ----
(0) (O) ---
1 ? Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda. 0.0000 
170 
Table 28. Percentage of Arrests by Collapsed Category of Miscella-
neous Crimes and by Time Period for Riot Area Locations.I 
Collapsed Category/Time Per Cent Per Cent % Riot? Total 
Normal Riot 7. Normal Number 
Fraud 5o9 ---- -5.9 
(1) (0) 1 
Liquor Laws 29.4 10.0 -19.4 
(5) (1) 6 
Varied 35.3 70.0 34. 7 
(6) (7) 13 
Unknown 29.4 20.0 -9.4 
(5) (2) 7 
Total 100.0 100.0 
(17) (10) 27 
1 .. Data from the District of Columbia Metropolitan Police Department 
Arrest Record. 1968. 
Lambda - 0.0417 
171 
Table 29. Percentage of Arrests by General Category of Violation and 
by Time Period for Corridor Area Locations. 1 
General Category/Time Per Cent Per Cent % Riot? Total 
Normal Riot % Normal Number 
Persons 3.4 2.2 -1.2 
(3) (10) 13 
Property 1.1 17.6 16.5 
(1) (82) 83 
Traffic 71.6 9.0 -62.6 
(63) (42) 105 
Crimes Without Victims 18.2 70.l 51.9 
(16) (326) 342 
Fires ---- ?--(- ----(0) 0) ---
Miscellaneous S.7 1.1 -4.6 
(S) (S) 10 
Total 100.0 100.0 
(88) (465) 553 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda ? 0.2274 
172 
Table 30. Percentage of Arrests by Collapsed Category of Crimes 
Against Persons and by Time Period for Corridor Area 
Locations.l 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Potentia 1 Threat of 10.0 10.0 
Bodily Harm (0) (7) 7 
Threat of Bodily Harm ---- ----
(0) (0) 
Actual Bodily Harm 100.0 30.0 -70.0 
(3) (3) 6 
Total 100.0 100.0 
(3) (10) 13 
1 ? Data from the District of Columbia Metropolitan Police Department 
Arrest Record~ 1968. 
Lambda ... 0.3333 
. . . 
-~~~----.,,,. .. .,.cv.,.....;.w, ..~ ,.__..i:.......__.......,_-"'=--.,,....:..., ... ,.__,_ .,,..,~~-?,., _..~_.,_, __ .,, ______ ,._,_~_ _ __,.,.. ... _,.., "'--,,...,_...,._, __ _ 
173 
Table 31. Percentage of Arrests by Category of Crimes Against Prop-
erty and by Time Period for Corridor Area Locations.l 
Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Burglary-House Breaking 100.0 87.8 -12.2 
(1) (72) 73 
Larceny 6.1 6.1 
(0) (5) 5 
Auto Theft 3.7 3.7 
(0) (3) 3 
5t0len Property ---- ---- ----
(0) (0) 
Vandalism 2.4 2.4 
(0) (2) 2 
Total 100.0 100.0 
(1) (82) 83 
1 ? Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda? 0.0000 
174 
Table 32. Percentage of Arrests by Collapsed Category of Traffic 
Violations and by Time Period for Corridor Area Locations.I 
Collapsed Category/Time Per Cent Per Cent ? Riot - Total 
Normal Riot % Normal Number 
Movtng Violations 96.8 97.6 0.8 
(61) (41) 102 
Equipment Violations 3.2 2.4 -0.8 
(2) (l) 3 
Total 
100.0 100.0 
(63) (42) 105 
1 
? Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda .. 0.0000 
175 
Table 33. Percentage of Arrests by Collapsed Category of Crimes 
Without Victims and by Time Period for Corridor Area 
Locations.I 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Non-Riot Related ----
(0) (0) 
Drunkenness 56.2 s.a -50.4 
(9) (19) 28 
Disorderly Conduct 43.8 94.2 so.4 
and Curfew2 (7) (307) 314 
Total 100.0 100.0 
(16) (326) 342 
1 ? Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
2 ? Riot time period will contain curfew arrests classified as disor-
derly conduct violations by the police. 
Lambda? 0.0455 
"??-- --??-,-----------...? ------?---~---?-____.__,-cc__ - :=_ ----??-?------, -
176 
Table 34. Percentage of Arrests by Category of Crimes Related to 
Fires and by Time Period for Corridor Area Locations. 1 
Category /Time Per Cent Per Cent % Riot? Total 
Normal Riot % Normal Number 
False Fire Alarm ----
(0) (0) ---
Arson 
(0) (0) 
Total ----
(0) (0) 
1
? Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda? 0.0000 
177 
Table 35 ? Percentage of Arrests by Collapsed Category of Miscella-
neous Crimes and by Time Period for Corridor Area Loca-
tions.l 
Collapsed Category/Time Per Cent Per Cent % Riot? Total 
Normal Riot % Normal Number 
Fraud ---- ----
(0) (0) ---
Liquor Laws ---- ---- ----
(0) (O) ---
Varied 100.0 60.0 -40.0 
(5) (3) 8 
Unknown ---- 40.0 40.0 
(0) (2) 2 
Total 100.0 100.0 
(5) (5) 10 
le Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda ? 0.2857 
----?---------- - ?-
- -----~, 
178 
Table 36., Percentage of Arrests by General Category of Violation and 
by Time Period for Non-Riot Area Locations.I 
General Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Persons 4.6 3.4 -1.2 
(123) (113) 236 
Property 4.2 15.1 10.9 
(112) (503) 615 
Traffic 69.5 15.2 -54.3 
(1846) (505) 2351 
Crimes Without Victims 17.2 63 .9 46.7 
(456) (2127) 2583 
Fires 0.1 0.1 
(0) (3) 3 
Miscellaneous 4.5 2.3 -2.2 
(119) (77) 196 
Total 100.0 100.0 
(2656) (3328) 5984 
le Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda,.. 0.4595 
' 
179 
Tnb le 3 7 ? Percentage of Arrests by Collapsed Category of Crimes 
Against Persons and by Time Period for Non-Riot Area 
Locations.I 
Coll apsed Category/Time Per Cent Per Cent ?Riot? Total 
Normal Riot % Normal Number 
Potential Threat of 26.0 54.0 28.0 
Bodily Harm (32) (61) 93 
Threat of Bodily Harm 13.0 12.4 -0.6 
(16) (14) 30 
Actual Bodily Harm 61.0 33.6 -27.4 
(75) (38) 113 
Total 
100.0 100.0 
(123) (113) 236 
1
? Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda ,.. 0.2203 
... ????-----?--??-??? ?-------c. __ cc______ -?-? ??--- - --, ?-
180 
Table 38. Percentage of Arrests by Category of Crimes Against 1 
Property and by Time Period for Non-Riot Area Locations. 
Ca t:egory /Time 
Per Cent Per Cent % Riot:? Total 
Normal Riot % Normal Number 
Burglary-House Breaking 36.6 82.5 45.9 
(41) (415) 456 
Larceny 
34.8 3.2 -31.6 
(39) (16) 55 
Auto Theft 
10.7 5.0 -5.7 
(12) (25) 37 
S t olen Property 
5.8 5.8 
(0) (29) 29 
Vandalism 
17.9 3.6 -14.3 
(20) (18) 38 
Total 
100.0 100.1 
(112) (503) 615 
1. Data from the District 
Arrest Record, of Columbia Metropolitan Police Department 1968. 
Lambda ? o .0923 
181 
Table 39. Percentage of Arrests by Collapsed Category of Traffic 
Violations and by Time Period for Non-Riot Area Locations.I 
Collapsed Category/Time Per Cent Per Cent ? Riot - Total 
Normal Riot % Normal Number 
Moving Violations 95.6 93. 7 -1.9 
(1765) (473) 2238 
Equipment Violations 4.4 6.3 1.9 
(81) {32) 113 
Total 100.0 100.0 
{1846) {505) 2351 
I. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda? 0.0000 
182 
Table 40. Percentage of Arrests by Collapsed Category of Crimes 
Without Victims and by Time Period for Non-Riot Area 
Locations.I 
Collapsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot ? Normal Number 
Non-Riot Related 4.4 o. 7 -3. 7 
(20) (15) 35 
Drunkenness 71.3 11.5 -59.8 
(325) (245) 570 
Disorderly Conduct 
and Curfew2 24.3 87.8 63.5 
(111) (186 7) 1978 
Total 100.0 100.0 
(456) (2127) 2583 
1 ? Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
2 ? Riot time period will contain curfew arrests classified as dis-
orderly conduct violations by the police. 
Lambda .. 0.2818 
183 
Table 41. Percentage of Arrests by Category of Crimes Related to 
Fires and by Time Period for Non-Riot Area Locations.! 
Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
False Fire Alarm ---- ----
(0) (0) 
Arson ---- 100.0 100.0 
(0) (3) 3 
Total ---- 100.0 
(0) (3) 3 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
Lambda? 0.0000 
184 
Percentage of Arrests by Collapsed Category of Miscella-
neous Crimes and by Time Period for Non-Riot Area Loca-
tions.I 
Co11 apsed Category/Time Per Cent Per Cent % Riot - Total 
Normal Riot % Normal Number 
Fraud 
6.7 1.3 -5.4 
(8) (1) 9 
Liquor Laws 
3.4 3.9 o.5 
(4) {3) 7 
Varied 
65.5 61.0 -4.5 
(78) (47) 125 
Unknown 
24.4 33.8 9.4 
(29) (26) 55 
Total 
100.0 100.0 
(119) (77) 196 
1. 
Data from the District 
Arrest Record. 1968. of Columbia Metropolitan Police Department 
Lambda? 0.0000 
185 
Table 43. Lambda Values for the Arrest Data by General Category and 
Collapsed Subcategory and by Spatial Area. 
Category/Spatial Area Riot Corridor Non-Riot 
General 
0.1584 0.2274 0.4595 
Crimes Against Persons 0.2459 o.3333 0.2203 
Crimes Against Property 0.2909 0.0000 0.0923 
Traff le Violations 
0.0000 0.0000 0.0000 
Crimes Without Victims 0.1703 0.0455 o.2a1s 
Fire-Related Crimes 1 ---.. -- ------ ----- ... 
Miscellaneous Crimes 
0.0417 o.2ss1 0.0000 
ls The only three items in this entire category were arrests made in 
the non-riot areas during the riot period. 
186 
Table 44 ? Percentage of Offenses Reported to the Police! by Time 
Period and by Locational Area.2 
Area/Time Per Cent Per Cent % Riot - Total 
Normal Riot 1. Normal Number 
Riot Areas 16.6 28.3 11.7 
(367) (760) 1127 
Corridor Areas 6.1 6.2 Owl 
(134) (166) 300 
Non-Riot Areas 77.4 65.5 -11.9 
(1711) (1759) 3470 
Total 100.1 100.0 
(2212) (2685) 4897 
1
? Part of these counts were taken from the District of Columbia Fire 
Department daily alarm log, 1968. 
29 
Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968 and contains only those items for which 
locational information was available. 
Lambda? 0.0000 
--,~-?-?. ...... ~ .... -"""_;..,_.,_, --?-??""'~""*"? '---<.,~,,__ 
-"---~-~---. -.......... ,_~_,_..., 
--- ---~~"'-"i... .. _"""""'-? ---~---
187 
Table 45. Percentage of Arrests Made by the Police by Time Period 
and by Locational Area.l 
Area/Time Per Cent Per Cent 7. Riot? Total 
Normal Riot % Normal Number 
Riot Areas 14.3 31.1 16.8 
(459) (1710) 2169 
Corridor Areas 2.7 8.4 5.7 
(88) (465) 553 
Non-Riot Areas 82.9 60.5 -22.4 
(2656) (3328) 5984 
Total 99.9 100.0 
(3203) (5503) 8706 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968 and contains only those items for which lo-
cational information was available. 
Lambda,.,, 0.0000 
188 
Table 46. of Violation and by Total Number of Offenses by Category 
Date for the Normal Time Period? 1 
Total 
Category/Date 3-29 3-30 3-31 4-1 4-2 4-3 4-4 3-29 
Homicide 1 1 1 
1 5 
1 8 
Rape 1 3 2 1 l 270 
Robbery 32 32 48 23 33 29 32 41 14 94 
Aggravated Assault 14 15 15 12 3 13 8 
Burglary-House Breaking 47 44 40 53 40 58 68 
47 397 
384 
Larceny 59 59 61 31 54 33 40 47 201 
Auto Theft 25 34 25 14 18 24 36 
25 
Other Assaults 6 14 7 7 3 6 10 
6 59 
Arson 1 2 l 2 
Forgery et al 1 1 
Fraud 1 2 1 7 1 1 1 
3 3 Embezzlement 
3 
Stolen Property 1 1 1 
Vandalism 64 5 8 16 6 9 3 12 5 
Weapons 3 2 2 4 2 3 
16 
Prostitution and Vice 
Sex Offenses 1 2 1 3 10 3 
Narcotics and Drugs 4 1 4 9 
Gambling l l 4 6 1 13 
Family and Children 
Intoxicated Driving 1 1 
Liquor Laws 2 10 2 2 16 
Drunkenness 
Disorderly Conduct 
Vagrancy 2 1 3 
Other2 1 4 1 10 3 6 3 1 29 
Suspicion 5 5 
Traffic 
Other Traffic 
Equipment Violations 
False Fire Alarm 26 33 26 22 17 14 27 26 191 
Fires3 106 138 100 79 92 59 79 106 759 
Total (without fires) 226 270 201 223 163 205 278 226 1792 
Total (with fires) 332 408 301 302 255 264 357 332 2551 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2e This category contains all offenses classified by the police as 
"other" with the exception of "false fire alarm" which has been 
treated separately. 
3. Data from the District of Columbia Fire Department daily alarm log, 
1968. 
------~==1._u.-=u:i~ ?--
189 
Table 47. of Violation and by Total Number of Offenses by Category 
Date for the Riot Time Period.l 
Total 
Category /Date 4-5 4-6 4-7 4-8 4-9 4-10 4-11 4-12 
1 10 Homicide S 2 2 5 
Rape 1 1 
3 
76 
Robbery 32 13 3 3 2 3 9 
11 
9 5 7 s 49 Aggravated Assault S 12 6 33 679 
Burglary-House Breaking 367 21 138 40 24 21 35 
Larceny 36 7 4 7 10 21 15 
46 146 
236 
Auto Theft 26 74 32 22 11 19 24 28 
3 9 3 47 Other Assaults 14 6 4 S 3 
Arson 4 1 2 32 12 10 3 4 
Forgery et a 1 1 2 1 
Fraud 2 1 8 
l 12 
Embezzlement l 1 45 
Stolen Property 20 14 4 3 4 
Vandalism so 24 1 7 4 4 3 2 5 
Weapons 99 16 30 21 10 5 9 1 7 
Prostitution and Vice 1 1 
Sex Offenses 2 l 3 
Narcotics and Drugs l 3 3 4 10 2 l 24 
Gambling 1 l 
Family and Children 
Intoxicated Driving 
Liquor Laws 1 l 2 4 
Drunkenness 
Disorderly Conduct2 
Vagrancy 
Other3 3 1 l 2 l 1 3 12 
Suspicion 7 6 1 14 
Traffic 
Other Traffic 
Equipment Violations 
False Fire Alarm 9 2 5 25 21 21 27 29 139 
Fires4 194 352 197 121 ll7 123 114 123 1341 
Total (without fires) 629 273 126 108 105 134 138 176 1689 
Total (with fires) 823 625 323 229 222 257 252 299 3030 
1. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
2. Also includes curfew violations classified as disorderly conducts. 
3e This category contains all offenses classified by the police as 
"Other" with the exception of "false fire alarms" which have 
been treated separately. 
4 .. Data from the District of Columbia Fire Department daily alarm 
log, 1968., 
---::---~?~ _-:_..:. .. -;,1 _______ ., ___ ..J.: ____ _. L~= ===-----?-
190 
of Violation and by 
Table 48. Total Number of Arrests by Category 
Date for the Normal Time Period.l 
Category /Date 3-31 4 1 4 2 4-3 4-4 3-29 
Total 
3-29 3-30 ? -
3 
Homicide l 1 l 9 
Rape l 2 2 4 2 28 
Robbery 5 4 2 7 3 2 3 5 3 60 
Aggravated Assault 3 5 6 10 10 18 17 3 76 
~urglary-House 17 Breaking 3 9 8 12 7 
Larceny 8 6 
88 
6 14 3 13 9 29 17 
Auto Theft 1 4 1 1 3 
7 
6 42 
Other Assaults 6 6 3 5 9 1 
6 
Arson 4 
Forgery et al 1 1 1 1 
Fraud 1 l 6 l 1 
l 11 
2 
Embezzlement l l 5 
Stolen Property l l 2 
1 
Vandalism 2 5 5 4 6 3 5 2 
32 
Weapons 3 4 9 4 8 7 6 
47 
6 
Prostitution and Vice 
Sex Offenses 4 1 l 1 l 
Narcotics and Drugs 9 l 1 6 9 
26 
Gambling 4 2 3 7 4 20 
Family and Children 1 l 
Intoxicated Driving 5 2 1 1 1 1 11 
Liquor Laws 3 9 7 2 3 3 27 
Drunkenness 64 122 72 64 77 61 73 64 597 
Disorderly Conduct 31 38 20 19 16 8 47 31 210 
Vagrancy 2 1 3 
Other2 12 5 10 16 11 10 25 12 101 
Suspicion l 4 5 6 16 
Traffic 205 222 189 234 305 219 191 205 1770 
Other Traffic 26 23 28 86 92 70 74 26 425 
Equipment Violations 15 15 12 13 12 10 19 15 111 
False Fire Alarm 
Unknown 4 4 5 2 4 14 4 37 
Total 396 505 385 504 581 485 531 396 3783 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
2. This category contains all offenses classified by the police as 
"Other" with the exception of "false fire alarm" which has been 
treated separately. 
191 
'.l'otat N 
~ te f Umber of Arrests b 
or the Riot :Y Category of Violation and by 
1'ime Period.l 
1{%1 4 .. 5 4-6 
fti.lp Ct <le 4-7 4-8 4-9 4-10 4-11 4-12 Total 
ltoo: 
a4&  e.t-y &l'e~ l l l llt <ltec1 1 3 
L ~lql:'y~Hi 1 4ssauzt 9 2 6 G.t-c <>us 5 4 8 2 1 
<ii,t:o eny e Breaking 5 4 3 
444 3 5 
26 
3 4 2 
Ot.ha ':theft 282 2 85 23 28 
4t- ? 14 26 l:' 26 37 937 
~ 14 .!loll S.EJlflt 7 1 3 2 
O.t- S l 5 7 3 42 
li' 6 ?'e.t'Jr 2 12 4 3 3 8 14 41 
l',al.lct et 01 l 2 2 2 6 
3 5 30 ~baa 1 2 
Stoz tle41 l 7 
lfc1 en ?1 ent 
llci :t'op 
~a <llts 8 t'ty 3 3 6 
i.lpon lll 
.E> s 35 l 1 
t-o.st 32 6 1 4 2 l 4 
.Sa~ 1 tut1 22 3 85 6 3 8 
~ 2 Otfe oll alld 16 1 6 51 
G l'cotj .nses 32 'Vf ce 25 11 11 12 8 4 119 
"'- elllb l l cs tt.nct 
<lll)t l llg- Drugs l l 
lttt Y '1nc1 2 5 2 l 8 
ltq 3 o~.icot Ch11c1ren l 22 
Ot- llol:' OCf /)r l l 
b "nk I ennl. ett.rs 1 Vlng 
l 
1 2 
i, ~ol:'<1 ess l 1 2 
ct.g,l:'ct 8.t-ly C 2 1 1 
sO tfi llcJr onduct.2 
46 4 83 84 54 
8,t,J 41 74 74 483 101 557 
t.rsPi 1232 1132 712 468 177 209 88 4501 
!':t-i.l.f Cf 011 3 2 
ffc s Otti 
t . 13 8J:- '.l' 10 3 8 6 12 13 8 73 
%1p l:'(lft1 2 1 
.i,,e l lltent c 1 l l 3 9 40 52 45 88 34 
i--1 V101at1 57 trllk 99 153 568 Se 4 
llo~ l:'e '41 ons 15 13 23 11 14 8 13 101 
c'l.t"tn 2 9 1 5 2 1 8 13 41 
:ii 
Ot;:,~1 
l 4 5 8 6 13 37 
i. ~ 1151 1782 1415 964 624 406 488 478 7308 
<. l' '1t~t- t-fe ;om the I.J.:? :.::.::.::=================================--
'.hts c t: ltecord t.s t:rict of Columbia Met:ropolit:an Pollce Department 
3, :ff <lteg0 r , 1968. 
l'1t,1, ~ ecd~ at s c1;1,;o r~d1e1r11  contain arrests for curfew v1olia t:inns classi-
Ot0 egot-y Y conduct: v1olat:1ons by the pol ce. 
t'.t-8<'l e.t-,, ~i th contains all offenses classified by the police as 
? ted Sep< !Irtahtee lye,.x ception of "false fire alarm" which has been 
I "V~':./ ??'?"' -
L_,._.,.~ ,~&--~- ........ :: ..~  .. ~--- ..,,..,;_ -... ~' -?--~'""--.....__~~.. ................ -.....,;.. ____. .._"""'?-4?--?-.... ,. ...... .,..., ?. """..-..-...... 
192 
1 
Ta b le SO. Total Number of Offenses by Date and by Location. 
Location Total 
Date Non- Corridor Riot Riot Riot 
Riot 
Area Unknown 
Riot Area Area Area Area 
Area 14th 7th H 
SE 
3 48 226 3-29 143 11 7 8 6 
49 270 
3-30 156 17 21 14 12 
1 
201 
3-31 41 117 8 14 13 8 
57 223 4-1 124 10 16 9 5 
2 
1 26 163 4-2 111 5 10 3 7 
19 205 4-3 133 15 15 12 10 1 
4-4 150 11 36 15 12 3 
51 278 
Total 1566 934 77 119 74 60 11 
291 
4-5 310 41 58 49 45 13 113 629 
4-6 152 21 13 12 20 5 50 273 
4-7 66 7 6 4 4 1 38 126 
4-8 74 7 5 5 5 1 11 108 
4-9 58 3 9 4 3 28 105 
4-10 75 5 11 5 3 1 34 134 
4-11 85 4 8 9 7 2 23 138 
4-12 104 2 11 6 5 48 176 
Total 924 90 121 94 92 23 345 1689 
l. Data from the District of Columbia Metropolitan Police Department 
Offense Record, 1968. 
r1r?,-?,;..~ 
I 3 ~.Lei~-
:~--::.--?--?--?-??.::.:....:. 
193 
1 
T bl Date and by Location. 
a e St. Total Number of Arrests by 
Mis- Loca- Total 
Date Non- Cor- Riot Riot Riot 
Riot 
coded tion 
Riot ridor Area Area Area 
Area 
7th H SE 
Un-
Area Area 14th known 
2 19 
39 396 
3-29 272 9 21 23 11 
32 69 505 
3-30 320 11 31 24 18 
26 42 385 
3-31 233 10 33 22 19 --
41 504 
4-1 383 13 17 15 6 
1 28 
29 44 581 
4-2 447 18 28 7 1 7 
17 51 485 
4-3 354 14 21 22 6 --
74 531 
4-4 11 375 15 23 24 9 
360 3387 
Total 162 2384 79 164 158 76 4 
60 3 21 254 
1151 
4-5 486 120 133 74 
4-6 412 
1782 
838 160 144 107 105 3 13 
4-7 614 60 116 146 62 1 20 
396 1415 
4-8 48 1 16 191 964 424 45 142 97 
4-9 292 34 95 33 34 9 127 
624 
4-10 174 12 43 18 17 1 6 135 406 
4-11 222 18 31 67 36 16 98 488 
4-12 278 16 41 26 26 7 84 478 
Total 3328 465 745 568 388 9 108 1697 7308 
1. Data from the District of Columbia Metropolitan Police Department 
Arrest Record, 1968. 
194 
Table 52. Total Number of Fires by Date and by Location.I 
Date Non-Riot Corridor Riot Riot Riot Riot Total 
Area Area Area Area Area Area 
14th 7th H SE 
3-29 93 5 5 I 2 106 
3-30 108 14 3 8 5 138 
3-3 l 88 2 3 3 3 l 100 
4 .. 1 65 5 5 2 2 79 
4-2 84 5 2 I 92 
4-3 43 9 4 l 2 59 
4-4 60 1 10 4 3 I 79 
Total 541 41 32 19 18 2 653 
4-5 89 8 47 30 18 2 194 
4-6 187 30 49 52 30 4 352 
4-7 120 9 18 34 15 1 197 
4-8 84 3 13 18 2 1 121 
4 .. 9 91 5 6 8 5 2 117 
4-10 92 8 6 9 7 1 123 
4-11 BO 8 1 17 6 2 114 
4-12 92 5 8 14 3 1 123 
Total 835 76 148 182 86 14 1341 
1. Data from the District of Columbia Fire Department daily alarm log~ 
1968. 
APPENDIX C 
FIGURES 
/ 
f 
L 
196 
Figure 1. First General Classification of Arrests and Offenses, 
Indicating the Code and the Description of the Code. 
I. Crimes Against Persons: 
1. Homicide (0100) 5. Other Assaults (0800) 
2. Rape (0200) 6. Weapons (1500) 
3. Robbery (0300) 7. Sex Offenses (1700) 
4. Aggravated Assault (0400) B. Family and Children (2000) 
II. Crimes Against Property: 
1. Burglary-House Breaking (0500) 4. Stolen Property (1300) 
2. Larceny (0600) s. Vandalism (1400) 
3. Auto Theft (0700) 
III. Traffic Violations: 
1. Intoxicated Driving (2100) 3. Other Traffic (3500) 
2. Traffic (3300) 4. Equipment Violations (3800) 
IV. Crimes Without Victims:l 
1. Prostitution and Vice (1600) 4. Drunkenness (2300) 
2. Narcotics and Drugs (1800) s. Disorderly Conduct (2400) 
3. Gambling (1900) 6. Curfew Violations (no code) 
v. Crimes Related to Fires: 
1. Arson (0900) 3. Number of Fires2 
2. False Fire Alarm (2649) 
VI. Miscellaneous Crimes: 
1. Forgery & Counterfeiting (1000) 6. Other (2600)
3 
2. Fraud (1100) 7. Suspicion (2700) 
3. Embezzlement (1200) s. Unknown (3000) 
4. Liquor Laws (2200) 9. Unknown (3100) 
5. Vagrancy (2500) 10. Unknown (3400) 
1. Curfew violations obviously only occurred during the riot period 
and then they were treated as disorderly conducts for the pur-
poses of classification by the police. It is not possible to 
separate the curfew violations from the more general disorderly 
conduct category. 
2. The incidences of fire were only used in the offense categoriza-
tion and reflect the District of Columbia Fire Department data. 
3. False fire alarms, which are treated above, have been eliminated 
from this category. 
.- .. ?~---??.?? 
197 
Figure 2. First General Classification of Arrests and Offenses 
Indicating the Subcategories and the Codes and Their 
Description. 
I. Crimes Against Persons : 
1. Potential Threat of Bodily Harm: Weapons (1500) 
2. Threat of Bodily Harm: Robbery (0300) 
3. Actual Bodily Harm: Homicide (0100), Rape (0200), Aggravated 
Assault (0400), Other Assaults (0800), Sex Offenses (1700), 
and Family and Children (2000) 
II. Crimes Against Property (no subcategorization): Burglary-House 
Breaking (0500), Larceny (0600), Auto Theft (0700), Stolen 
Property (1300), and Vandalism (1400) 
III. Traffic Violations: 
1. Moving Violations: Intoxicated Driving (2100), Traffic (3300), 
and Other Traffic (3500) 
2. Equipment Violations: Equipment Violations (3800) 
IV. Crimes Without Victims: 
1. Non-Riot Related: Prostitution and Vice (1600), Narcotics and 
Drugs (1800), and Gambling (1900) 
2. Drunkenness: Drunkenness (2300) 
3. Disorderly Conduct and Curfew: Disorderly Conduct (2400) and 
Curfew (no code) 
v. Crimes Related to Fires (no subcategorization): Arson (0900), 
False Fire Alarm (2649), and number of fires reported to 
the District of Columbia Fire Department (no code) 
VI. Miscellaneous Crimes: 
1. Fraud: Forgery and Counterfeiting (1000), Fraud (1100), and 
Embezzlement (1200) 
2. Liquor Laws: Liquor Laws (2200) 
3. Varied: Vagrancy (2500), Other (2600), and Suspicion (2700) 
4. Unknown: Unknown (3000), Unknown (3100), and Unknown (3400) 
APPENDIX D 
MAPS 
/a! v
,~  / 
'.I 
199 
1 
Map l Indicating the Riot, Corridor, and Non-Riot Areas. 
1. 
SELECTED BIBLIOGRAPHY 
Books 
Barton, Allen H. Communities in Disaster: A Sociological Anallsis of 
Collective Stress Situations. Garden City, New York: Doubleday 
& Company, Inc., 1969. 
Berkovitz, Leonard. 11The Study of Urban Violence: Some Implications 
of Laboratory Studies of Frustration and Aggression." Riots and 
Rebellion: Civil Violence in the Urban Communitr? Louis H. 
Masotti and Don R. Bowen, ed. Beverly Hills, California: Sage 
Publications, Inc., 1968. Pp. 39-49. 
Blalock, Jr., Hubert M. Social Statistics. New York, New York: 
? McGraw- Hi 11 Book Company, Inc., 1960. 
Blauner, Robert. "Internal Colonialism and Ghetto Revolt. 11 Social 
Problems. Vol. 16, No. 4 (Spring, 1969). Pp. 393-408. 
Bloombaum, Milton. "The Conditions Underlying Race Riots as Portrayed 
by Multidimensional Scalogram Analysis: A Reanalysis of Lieberson 
and Silverman ?s Data. 11 American Sociological Review. Vol. 33, 
No. l (February, 1968). Pp. 76-91. 
Baskin, Joseph. "The Revolt of the Urban Ghettos, 1964-1967. 11 The 
Annals of the American Academy of Political and Social Sele;;;;;. 
Vol. 382 (March, 1969). Pp. 1-14. 
Baskin, Joseph, ed. Urban Social Violence in the Twentieth Centurl? 
Beverly Hills, California: Glencoe Press, 1969. 
Bowen, Don R. and Louis H. Masotti? 11Civi 1 Violence: A Theoretical 
Overview." Riots and Rebellion: Civil Violence in the Urban 
Communi tv. Louis H. Masotti and Don R. Bowen, ed. Beverly Hills, 
California: Sage Publications, Inc., 1968. Pp. 11?31. 
Brown, Claude. Mane hi ld in the Promised Land. New York, New York: 
The Macmillan Company, 1965. 
Brownlee, Aleta. "Disaster and Disaster Relief." Encyclopaedia of 
the Social Sciences. Vol. 5. Edwin R. A. Seligman, ed. New 
York, New York: The .Macmillan Company, 1931. Pp. 161-66. 
Broyhill, Joel T. "Praise of the Salvation Army in Washington, D. c., 
During the Recent Rioting Disorders." Congressional Record. 
Vol. 114, No. 87. Item E4471. 
200 
201 
Carmichael, Stokely and Charles V. Hamil ton. Black Power: The Poli tics 
of Liberation in America. New York, New York: Random House, 1967. 
Chapman, Dwight W. "Di mens ions of Models in Disaster Behavior. 11 tlfil1_ 
and Society in Disaster. George W. Baker and Dwight W. Chapman, 
ed. New York, New York: Basic Books, Inc., 1962. Pp. 305-36. 
Cisin, Ira H. with Walter B. Clark. "The Methodological Challenge of 
Disaster Research. 11 Man and Society in Disaster. George W. Baker 
and Dwight W. Chapman, ed. New York, New York; Basic Books, Inc., 
1962. Pp. 23-49. 
Cohen, Albert K. "The Sociology of the Deviant Act: Anomie Theory and 
Beyond. 11 American Sociological Review. Vol. 30, No. 1 (February, 
l 96 5) ? Pp. 5-14 ? 
Cressey, Donald R. "Crime. 11 Contemporary Social Problems. 2nd ed. 
Robert K. Merton and Robert A. Nisbet, ed. New York, New York: 
Harcourt, Brace & World, Inc., 1966. Pp. 141-45. 
Demerath, Nicholas J. "Some General Propositions: An Interpretative 
Summary." Human Organization. Vol. 16, No. 2 (Summer, 1957). 
Pp. 28f. 
Demerath, Nicholas J. and Anthony F. c. Wallace. "Introduction to 
Human Adaptation to Disaster." Human Organization. Vol. 16, 
No. 2 (Summer, 1957). Pp. lf. 
Dobrovir, William A ?.J ustice in Time of Crisis: A Staff Report to the 
District of Columbia Committee on the Administration of Justice 
Under Emergency Conditions. illashington, D. c.: U.S. Government 
Printing Office, 1969. 
Dollard, John, Leonard W. Doob, Neal E. Miller, O. H. Mowrer, and 
Robert R. Sears. Frustration and Aggression. New Haven, Connecti-
cut: Yale University Press, 1939. Pp. 1-54. 
Dynes, Russell R. Organized Behavior in Disaster. Lexington, ~~ssa-
chusetts: D. c. Heath and Company, 1970. 
Fanon, Frantz. The Wretched of the Earth. New York, New York: Grove 
Press, Inc., 1963 0 
Fritz, Charles E. "Disaster.'' Contemporary Social Problems. Robert 
K. Merton and Robert A. Nisbet, ed. New York, New York: Harcourt, 
Brace & World, Inc., 1961. Pp. 651-94. 
Fritz, Charles E. "Disasters." International Encyclopedia of the So-
cial Sciences. Vol. 4. David L. Sills, ed. New York, New York: 
The Macmillan Company and the Free Press, 1968. Pp. 202-207. 
Fritz, Charles E. "Disasters Compared in Six American Communities 11 O 
Human Organization. Vol. 16, No. 2 (Summer, 1957). Pp. 6-9. 
--... -..... ,'.....,;.-.....,_,' 
_... ......... ,....._. .. _,._ ____ ~_.,.,_,.,.,,,, _______, _ ' 
. f.~i.:~-?~-
202 
Fritz, Charles E. and Hc1rry B. Williams. "The Human Being in Disasters: 
A Research Perspective." The Annals of the American Acad(!mY o.f 
Political and Social Science. Vol. 309 {January, 1957). Pp. 42-
51. 
Fritz, Charles E. and J. H. Mathewson. Convergence Behavior in Disas-
ters: A Problem of Social Contr<:?!.? (Committee on Disaster Stud-
ies Report, No. 9). Washington, D. c.: National Academy of 
Sciences--National Research Council, 1957. 
Geschwender, James A. and Benjamin D. Singer. "Deprivation and the 
Detroit Riot." Social Problems. Vol. 17, No. 4 {Spring, 1970). 
Pp. 457-63. 
Gilbert, Ben W. Ten Blocks from the White House: Anatomy of the 
Washington Riots of 1968. New York, New York: Frederick A." Praeg-
er, Publishers, 1968. 
Gittell, Marilyn and Sherman Krupp. "A Model of Discriminations and 
Tens ions." Riots and Rebell ion: Civil Violence in the Urban 
?ommunity. Louis H. Masotti and Don R. Bowen, ed. Beverly Hills, 
California: Sage Publications, 1968. Pp. 69-83. 
Goodman, Leo A. and William H. Kruskal. "Measures of Association for 
Cross Classifications." Journal of the American Statistical As-
~ociatio!!.!_ Vol. 49 (December, 1954). Pp. 735-45. 
Grimshaw, Allen D. "Three Views of Urban Violence: Civil Disturbance, 
Racial Revolt, Class Assault." Riots and Rebellion: Civil Vio? 
lence in the Urban Community. Louis H. Masotti and Don R. Bowen, 
ed. Beverly Hills, California: Sage Publications, Inc., 1968. 
Pp. 103-19. 
Gurr, Ted. "Urban Disorder: Perspectives from the Comparative Study 
of Civil Strife." Riots and Rebellion: Civil Violence in the 
Urban Community. Louis H. Masotti and Don R. Bowen, ed. Bever-
ly Hills, California: Sage Publications, Inc., 1968. Pp. 51-67 ? 
Harris, C. D. and Edward L. Ullman. "The Nature of Cities." The 
Annals of the American Academy of Political and Social Science. 
Vol. 242 (November, 1945). Pp. 7-17. 
Hays, William L. Statistics for Psychologists. New York, New York: 
Holt, Rinehart and Winston, 1963. 
Heinl, Jr., R. D. "Washington's Three Ulys of Burning." Fire Engin-
eering. (July, 1968). Pp. 38ff. 
Horouitz, Irving Louis. "Comment: Black Sociology." Trans-acti<2!!,? 
Vol. 4, No. 9 (September, 1967). Pp. 7f. 
Johnson, Elmer Hubert. Crime, Correction, and Society. Revised ed. 
Homewood, Illinois: The Dorsey Press, 1968. 
/ 
??------~---??-----?--?-------?- -? ?????--.:. 
________ ._ _______ ?? ~- .... -~~".., ..... ~ . ..._._---~- . __. ,.,__,.? ... .._ ......., ,...,?..,,~~~ 
203 
Killian, Lewis M. An Introduction to Methodological Problems of Field 
Studies in Disasters. (Committee on Disaster Studies Report, No. 
8). Washington, D. C.: National Academy of Sciences--National 
Research Council, 1956. 
Kilpatrick, F. P. "Problems of Perception in Extreme Situations." 
Human Organization. Vol. 16, No. 2 (Summer, 1957). Ppe 20ff. 
Kitsuse, John I. and Aaron V. Cicourel. "A Note on the Uses of Official 
Statistics." Social Problems. Vol. 11 (Fall, 1963). Pp. 131-39. 
Lang, Kurt and Gladys Engel Lang. "Racial Disturbances as Collective 
Protest." Riots and Rebel lion: Civil Violence in the Urban Com-
munity. Louis H. Masotti and Don R. Bowen, ed. Beverly Hills, 
California: Sage Publications, Inc., 1968. Pp. 121-30. 
Laue, James H. "Power, Conflict, and Social Change." Riots and Re-
bellion: Civil Violence in the Urban Community. Louis H. Masotti 
and Don R. Bowen, ed. Beverly Hills, California: Sage Publica-
tions, Inc., 1968. Pp. 85-96. 
LeBon, Gustave. The Crowd: A Study of the Popular Mind. New York, 
New York: The Viking Press, Inc., 1960. 
Lemons, Hoyt. "Physical Characteristics of Disasters: Historiclll and 
Statistical Review. 11 The Annals of the American Academy of Po-
li tical and Social Science. Vol. 309 (January, 1957). Pp. 1-14. 
Lieberson, Stanley and Arnold R. Silverman. "The Precipitants and 
Underlying Conditions of Race Riots." American Sociological Re-
.Y!..m!? Vol. 30, No. 6 (December, 1965). Pp. 887-98. 
Lipset, Seymour Martin. Political Man: The Social Bases of Politics. 
Garden City, New York: Doubleday & Company, Inc., 1960. Pp. 
61-72. 
McPhail, Clark. "Civil Disorder Participation: A Critical Examina-
tion of Recent Research. 11 American Sociological Review. Vol? 36 , 
No. 6 (December, 1971). Pp. 1058-73. 
Oberschall, Anthony. "The Los Angeles Riot of August 1965. 11 Social 
Problems. Vol. 15, No. 3 (Winter, 1967). Pp. 322-41. 
Paige, Jeffery M. "Poli ti cal Orientation and Riot Participation. 11 
American Sociological Review. Vol. 36, No. 5 (October, 1971). 
Pp. 810-20. 
Parmenter, Tom. "Breakdown of Law and Orders A Firsthand Account in 
the \lords of Detroit's Ghetto Dwellers." Trans-action. Vol. 4, 
No. 9 (September, 1967). Pp. 13-22. 
Quarantelli, E. L. 11The Nature and Conditions of Panic. 11 The Ameri-
can Journal of Sociolog~. Vol. LX, No. 3 (November, 1954). 
267-75. 
204 
Quarantelli, E. L. and Russell R. Dynes. "Looting in Civil Disorders: 
An Index of Social Change." Riots and Rebellion: Civil Violence 
in the Urban Community. Louis H. tvasotti and Don R. Bowen, ed. 
Beverly Hills, California: Sage Publications, Inc., 1968. 
Pp. 131-41. 
Quarantelli, E. L. and Russell R. Dynes. "Property Norms and Looting: 
Their Patterns in Community Crises." Phylon. Vol. 31 (Summer, 
1970). Pp. 168-72. 
Rainwater, Lee. "Open Letter on White Justice and the Riots." Trans-
action. Vol. 4, No. 9 (September, 1967). Pp. 22-32. 
Ransford, H. Edward. "Isolation, Powerlessness, and Violence: A Study 
of Attitudes and Participation in the Watts Riots." The American 
Journal of Sociology. Vol. 73, No. 5 (March, 1968). Pp. 581-91. 
Rayner, Jeannette F. "Studies of Disasters and Other Extreme Situa-
tions--An Annotated Selected Bibliography. 11 Human Organization. 
Vol. 16, No. 2 (Summer, 1957). Pp. 30-40. 
Sears, David o. and John B. NcConahay. "Participation in the Los An-
geles Riot. 11 Social Problems. Vol. 17, No. 1 (Summer, 1969). 
Pp. 3-20. 
Sjoberg, Gideon. "Disasters and Social Change." Man and Society in 
Disaster. George w. Baker and Dwight W. Chapman, ed. New York, 
New York: Basic Books, Inc., 1962. Pp. 356-84. 
Smelser, Neil J. Theory of Collective Behavior. New York, New York: 
The Free Press, 1962. Pp. 1-130. 
Sp erg el, Irving Ao "Youth Gangs and Urban Riots." Riots and Rebel-
lion: Civil Violence in the Urban Communitv. Louis H. lvasotti 
and Don R. Bowen, ed. Beverly Hills, California: Sage Publica-
tions, Inc., 1968. Pp. 143-56. 
Sp! lerman, Seymour. "The Causes of Racial Disturbances: A Comparison 
of Alternative Explanations." American Sociological Review. 
Vol. 35, No. 4 (August, 1970). Pp. 627-49. 
Spilerman, Seymour. "The Causes of Rae ia I Disturbances: Tes ts of an 
Explanation." American Sociological Review. Vol. 36, No. 3 
(June, 1971). Pp. 427-42. 
Strauss, Anselm L. Images of the American City. New York, New York: 
The Free Press of Glencoe, 1961. 
Thompson, James D. and Robert W. Hawkes. "Disaster, Community Organiz-
ation, and Administrative Process. 11 Man and Society in Disaster. 
George W. Baker and Dwight W. Chapman, ed. New York, New York: 
Basic Books, Inc., 1962. Pp. 268-300. 
205 
Tomlinson, T. M. "The Development of a Riot Ideology among Urban 
Negroes." American Behavioral Scientist. Vol. 12, No. 4 (March-
April, 1968). Pp. 27-31. 
Wallace, Anthony F. c. Human Behavior in Extreme Situations: A Sur-
ID of the Literature and Suggestions for Further Research. 
(Committee on Disaster Studies Report, No. I). Washington, D. c.: 
National Academy of Sciences--National Research Council, 1956. 
Wanderer, Jul es J. 11196 7 Riots : A Test of the Congruity of Events.,, 
Social Problems. Vol. 16, No. 2 (Fall, 1968). Pp. 193-98. 
Wi 11 lams, Harry B. "Some Functions of Communication in Crisis Behavior." 
Human Organization. Vol. 16, No. 2 (Summer, 1957). Pp. 15-19. 
Wirth, Louis. "Urbanism as a Way of Life." The American Journal of 
aocfologz. Vol. 44, No. 1 (July, 1938), Pp. 1-24. 
Government and Other Complete Publications 
Q,istrict of Columbia Code: Annotated, 1967 Edition. Washington, D. C.: 
U.S. Government Printing Office, 1967. 
filstrict of Columbia Code 2 1967 Edition: Cumulative Supplemen~ III. 
Washington, D. c., u. s .. Government Printing Office, 1970. 
Government of the District of Columbia. Annual Report of the Metro~ 
Q,olitan Police Department, Washington, D. c., Fiscal Year: 1968. 
Government of the District of Columbia. National Capita 1 Pla.nning Com-
mission. Civil Disturbances in Washington, D. c. April 4-8, 1968: 
:1..,Preliminary Damage Report. May, 1968. 
Government of the District of Columbia. Office of Civil Defense. Oper-
ation Bandaid One: Aeril 4-April 12, 1968. 
Government of the District of Columbia. Report on Civil Disturbances 
in Washington, D. c. April, 1968. 
National Red Cross. Red Cross News. Special issue on the activities 
of the first week of the Washington, D. c. riot in April, 1968. 
!,eport of the National Advisory Commission on Civil Disorders. Wash-
ington, D. c.: u. s. Government Printing Office, 1968, 
u. s. Department of Justice. Federal Bureau of Investigation. "Crime 
in the United States." Uniform Crime Reeorts--1970. Washington, 
D. c.: u. . s. Government Printing Office, 1970. 
u. S. House of Representatives. Hearings before the Committee on the 
District of Columbia. Civil Disturbances in Washington. 90th 
Congress, 2nd Session. May 15 and 16, 1968. Washington, D. C.: 
U.S. Government Printing Office, 1968. 
206 
U. s. House of Representatives. "Riot Damage in the District of 
Columbia." House of Representatives Report No. 1466. 90th 
Congress, 2nd Session. May 24, 1968. Washington, D. c.: 
U.S. Government Printing Office, 1968. 
U.S. Senate. Hearings before the Permanent Subcommittee on Investi-
gations of the Committee on Government Operations. Riotsr Civil 
and Criminal Disorders. Part 16. 91st Congress, 1st Session. 
May 9, 13, and 14, 1969. Washington, D. c.: u. s. Government 
Printing Office, 1969. 
u. S. Senate. Hearings before the Permanent Subcommittee on Investi-
gations of the Committee on Government Operations. Riots. Civil 
and Criminal Disorders. Part 17. 91st Congress, 1st Session. 
~~y 27 and 28, 1969. Washington, D. c.: u. S. Government 
Printing Office, 1969. 
U.S. Senate. Hearings before the Subcommittee on Business and Commerce 
of the Committee on the District of Columbia. Rehabilitation of 
District of Columbia Areas Damaged by Civil Disorders. Part I. 
90th Congress, 2nd Session. April 18 and 30, May 20, 28, and 29, 
1968. Washington, D. C.: u. s. Government Printing Office, 1968. 
U. s. Senate. Hearings before the Subcommittee on Business and Commerce 
of the Committee on the District of Columbia. Rehabilitation of 
District of Columbia Areas Damaged by Civil Disorders. Part II--
Appendix. 90th Congress, 2nd Session. Washington, D. c.: u. S. 
Government Printing Office, 1968. 
The Annals of the American Academy of Political and Social Science. 
Vol. 382 (March, 1969). 
Other Sources 
Letter to Howard Gabennesch of the University of Maryland from the 
Chesapeake and Potomac Telephone Company dated August 7, 1968. 
Letter to Joseph F. Hennessey, Chairman of the Young Lawyers Section 
of the Bar Association of the District of Columbia from Keven P. 
Charles of the Young Lawyers Section of the Bar Association of the 
District of Columbia dated April I, 1968. 
Letter to Howard Gabennesch of the University of Maryland from Bertram 
L. Keys, Jr., Executive Secretary of the Montgomery County Com-
mission on Human Relations dated July 23, 1968. 
Report of the Interracial Council for Business Opportunity of Greater 
Washington dated May IO, 1968.