The Conditions of Protest Behavior in American Cities

Author(s): Peter K. Eisinger

Source: The American Political Science Review, Vol. 67, No. 1 (Mar., 1973), pp. 11-28
Published by: American Political Science Association
Stable URL: http://www.jstor.org/stable/1958525
Accessed: 13-03-2017 10:11 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/1958525?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

American Political Science Association is collaborating with JSTOR to digitize, preserve and extend
access to The American Political Science Review

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
The Conditions of Protest Behavior in
American Cities*

PETER K. EISINGER
University of Wisconsin, Madison

Political Environment and Protest This type of analysis has depended on the
The complex interrelationship between politi- use of data on macro-level, or community,
cal environment variables on the one hand and characteristics for its independent variables.
political behavior on the other has been a per- The linkages between these diverse characteris-
sistent concern of analysts of urban politics. tics and patterns of political behavior and those
The purpose of this paper is to begin an explo- among the environmental variables themselves
ration of various environmental conditions as- have seldom been made explicit theoretically.
sociated with the incidence of political protest Such research efforts take on theoretical co-
activities directed toward urban institutions, herence, however, if it is understood in the
agencies, and officials in American cities. first instance that the environmental variables
Political environment is a generic term used are related to one another in the sense that they
variously in the literature of political science to establish a context within which politics takes
refer to, among other things, aspects of formal placed Furthermore, the possible linkages be-
political structure, the climate of governmental tween this context and the patterns of political
responsiveness, social structure, and social sta- behavior become evident if the elements of the
bility. Scholarly efforts have generally been di- context are conceived as components of the
rected toward examination of the extent to particular structure of political opportunities of
which specific configurations of environmental a community. That is to say, such factors as the
variables and distinctive patterns of local poli- nature of the chief executive, the mode of al-
tics occur together. Treating environmental ele- dermanic election, the distribution of social
ments as independent variables,' students have skills and status, and the degree of social disin-
shown relationships, for example, between re- tegration, taken individually or collectively,
formed municipal institutions and low voting serve in various ways to obstruct or facilitate
turnout,) reform government and high spend- citizen activity in pursuit of political goals.
ing and tax policies,3 centralization of local Other environmental factors, such as the cli-
power and urban renewal success,4 and less mate of governmental responsiveness and the
representative councilmanic institutions and the level of community resources, help to establish
incidence of race riots.5 the chances of success of citizen political activ-
ity. In short, elements in the environment im-
* This paper is a revised version of one prepared
pose certain constraints on political activity or
for a Seminar Panel on The Political Legacy of the
Urban Protests in the 1960s at the Annual Meeting
of the American Political Science Association, Sep- cal and Socio-Economic Characteristics of American
tember, 1971, Chicago, Illinois. The research was Cities," Municipal Yearbook 1965 (Chicago: Interna-
supported in part by funds granted to the Institute tional City Managers' Association, 1965), 82-97; and
for Research on Poverty at the University of Wis- Raymond Wolfinger and John Osgood Field, "Political
consin by the Office of Economic Opportunity pur- Ethos and the Structure of City Government," Ameri-
suant to the Economic Opportunity Act of 1964. can Political Science Review, 60 (June, 1966), 306-326.
I wish to thank a number of indefatigable assistants 2Robert Alford and Eugene Lee, "Voting Turnout
including William Corman, Mark Ginsberg, Eugene in American Cities," American Political Science Re-
Hahn, Freda Merritt, Robert Neis, and William view, 62 (September, 1968), 796-813.
Walker. I am also grateful to my colleagues Donald 3Edgar L. Sherbenou, "Class, Participation, and the
McCrone, Ira Sharkansky, Richard Merelman, and Council-Manager Plan," Public Administration Review,
David Seidman for their helpful comments and sug- 21 (Summer, 1961), 131-135.
gestions. The conclusions are, of course, my re- 4Amos H. Hawley, "Community Power and Urban
sponsibility alone. Renewal Success," American Journal of Sociology, 68
I It should be noted that a significant body of re- (January, 1963), 422-431.
search, based on similar intellectual assumptions, has 5Stanley Lieberson and Arnold R. Silverman, "The
also treated governmental structure as a dependent Precipitants and Underlying Conditions of Race
variable. These would include John H. Kessel, "Gov- Riots," American Sociological Review, 30 (December,
ernmental Structure and Political Environment," 1965), 887-898.
American Political Science Review, 56 (Sept., 1962), 6 Robert Lineberry makes a similar point in his
615-620; Leo Schnore and Robert Alford, "Forms of essay, "Approaches to the Study of Community Poli-
Government and Socio-Economic Characteristics of tics," in Community Politics, ed. Charles Bonjean,
Suburbs," Administrative Science Quarterly, 8 (June, Terry N. Clark, and Robert Lineberry (New York:
1963), 1-17; Robert Alford and Harry Scoble, "Politi- The Free Press, 1971), p. 20.

11

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 12 The American Political Science Review Vol. 67

open avenues for it. The manner in which indi- form systems, but their needs often cannot be
viduals and groups in the political system be- met using the criterion of efficient management.
have, then, is not simply a function of the re- Reform governments attempt to institutionalize
sources they command but of the openings, efficient management at the expense of recog-
weak spots, barriers, and resources of the polit- nizing particularistic and often costly interests.8
ical system itself. There is, in this sense, inter- This may lead to intense frustration on the part
action, or linkage, between the environment, of excluded groups. Lieberson and Silverman
understood in terms of the notion of a structure find evidence to support their hypothesis that
of political opportunities, and political behav- the less direct the relationship between the
ior. voter and his alderman (measured on a contin-
By measuring these environmental factors, uum which ranges from extensive at-large elec-
the analyst develops a means to judge the na- toral systems to small district ward systems),
ture of the biases which groups in a political the more likely it will be that race riots will oc-
system must confront. Such judgments lead to cur.9 Small districts provide access for more
conclusions about the ease with which people people and offer the possibility of minority rep-
can get what they want from the political sys- resentation, both of which serve to enlarge or
tem through collective action. Where the struc- open up the structure of opportunities. Sim-
ture of government is potentially more respon- ilarly, it would seem reasonable to suspect
sive to an electorate by providing opportunities that the incidence of protest, one manifestation
of formal representation for distinct segments of political frustration or impatience, is related
of the population (blacks, for example) or to the nature of the opportunity structure.
where the government is demonstrably respon- Several studies, notably those of Downes,10
sive to citizen needs and demands, the structure Spilerman,'1 and Palley and Palley,12 have at-
of opportunities is relatively open. There exist tempted to demonstrate linkages between as-
chances for diverse groups to exercise influence pects of what I have called the opportunity
through delegates on representative bodies and structure and the incidence of mass ghetto vio-
influence appears to elicit government action. lence. Their results have been uniformly nega-
Where formal or informal power appears to tive. Downes and Spilerman independently
be concentrated and where government is not found that the incidence of ghetto riots was es-
responsive, the opportunities for people to get sentially related to nothing more than the nu-
what they want or need through political action merical size of a city's black population, while
are limited. The opportunity structure is rela- Palley and Palley discovered that objective in-
tively closed. dicators of the degree of black social and eco-
One of the most explicit efforts to under- nomic deprivation were unreliable predictors of
stand elements of the political environment in ghetto violence.
terms of the structure of opportunities is found If one were to follow the notion of a number
in the work of Lineberry and Fowler. In their of scholars that ghetto violence is simply a
exploration of the differences in policy making form of protest,13 a more virulent extension of
between reformed and unreformed local gov- 8This view of reform government is most cogently
ernments, they conclude that put in Edward Banfield and James Q. Wilson, City
Politics (Cambridge, Massachusetts: Harvard and MIT
The translation of social conflicts into public Presses, 1963), pp. 40 ff.
policy and the responsiveness of political systems 9Lieberson and Silverman, p. 896.
to class, racial, and religious cleavages differs 10 Bryan T. Downes, "Social and Political Character-
markedly with the kind of political structure.... istics of Riot Cities: A Comparative Study," Social
Science Quarterly,
[Nion-partisan elections, at-large constituencies 49 (December, 1968), 504-520.
11 Seymour Spilerman, "The Causes of Racial Dis-
and manager governments are associated with a
turbances: Tests of a Theory," Institute for Research
lessened responsiveness of cities to the enduring
on Poverty Discussion Paper, University of Wisconsin,
conflicts of political life.7 Madison, Wisconsin, 1969.
12 Marian Lief Palley and Howard A. Palley, "Social
Reformed local government structure, whose Welfare Indicators as Predictors of Racial Disorders
rationale is to eliminate "politics" and partiality in Black Ghettos," a paper delivered at the Annual
from municipal administration, restricts the op- Meeting of the American Political Science Association,
September, 1969.
portunities which minority and outgroups have
13 See, for example, Robert Fogelson, Violence as
for success through political action. Not only is Protest (New York: Doubleday, 1971); Ralph Turner.
it likely that minorities will experience difficulty "The Public Perception of Protest," American Socio-
in gaining representation or even access in re- logical Review, 34 (December, 1969), 816-830; and
Edward Banfield, The Unheavenly City (Boston:
' Robert Lineberry and Edmund Fowler, "Reform- Little, Brown, and Co., 1970). Banfield speaks of
ism and Public Policies in American Cities," American "demonstrations" as a form of "rioting." See Chapter
Political Science Review, 61 (September, 1967), 715. 9, esp. p. 191.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
1973 The Conditions of Protest Behavior in American Cities 13

what is essentially the same political activity, command, in this case resources involving the
then this inquiry could stop here. There would ability of relatively unpredictable masses to dis-
be little reason to believe that "peaceful" pro- rupt and inspire fear. Only as a mass participa-
test and ghetto "revolts" are related to substan- tory action are such forces brought into play.
tially different sets of conditions. 2. Protest is a device by which actors mak-
I would contend, however, that protest ing demands in the political system attempt to
against local government targets and collective maximize the impact of their meager resources
ghetto violence are two forms of political ex- while at the same time they strive to minimize
pression which may be distinguished conceptu- the costs which they might incur by such de-
ally and empirically. The conceptual distinc- mand making.
tions-which I believe apply to protest directed
against any target and to most forms of collec- Efforts to Control Costs Distinguish Protest
tive political violence-simply suggest that pro- from Political Violence. Those who pursue
test and violence are so different in their basic violent political strategies are also attempting
dynamics that the likelihood that the conditions to maximize the impact of relatively insub-
associated with one will also be associated with stantial resources, but by their action they are
the other is not a foregone conclusion. Some of also exposing themselves to the possibility
the empirical distinctions suggest explicitly that that the costs of such behavior will be maxi-
the more narrow categories of behavior with mized, if death, serious physical injury, and
which we are concerned here are reactions to loss of freedom and legitimacy are taken as
different stimuli. That is, protest against local maximum costs. Protestors on the other hand
government targets is likely to be related to the seldom must make such expenditures.
nature of local politics, while ghetto violence is Protest is a product, then, of a cost-benefit
not. calculation; violent actors in contrast have es-
Let us examine some conceptual distinctions sentially thrown cost considerations to the
first. The term "protest" may be used either winds.
generically to refer to any form of verbal or 3. Political protest may also be distinguished
active objection or remonstrance or it may be from political violence by the fact that those
used technically to refer to a conceptually dis- who pursue the former rely for their impact
tinctive set of behaviors. I shall use it in the largely on the implicit threat of violence,17
latter sense. while those who pursue the latter are explicit in
Protest refers to a host of types of collective their intention to do physical harm in that they
manifestations, disruptive in nature, designed to either make open threats'8 or actually engage
provide "relatively powerless people"'14 with in violence. What is implicitly threatening in a
bargaining leverage in the political process.15 It protest is not only the socially unconventional
may be distinguished from other forms of ver- display by crowds of people, which offends and
bal objection'6 and politically inspired collec- frightens norm-abiding observers, but the vi-
tive violence by the following characteristics. sions which bystanders and targets conjure up
1. Protest is a collective act, carried out by 17 Donald Von Eschen and his colleagues argue that
those concerned with the issue and not by their the primary element leading to the success of a black
protest movement in Maryland which they studied
representatives. The vehement congressman
during the days of the civil rights turmoil was the ex-
who declares that he "protests" an action is not ploitation of elite fear of violence and civil disorder,
in fact relying on the forces, or resources, although the movement never intended violence. Don-
which protest, understood in the technical ald Von Eschen, Jerome Kirk, and Maurice Pinard,
"The Conditions of Direct Action in a Democratic
sense, brings into play. While the solitary con-
Society," Western Political Quarterly, 22 (June, 1969),
gressman relies on his status to lend his objec- 309. Ralph Turner also dwells on the theme of the
tion impact, protest is a mass action which re- manipulation of fear of violence as the motive force of
lies on the resources which groups of people can protest in "The Public Perception of Protest," p. 816.
See also H. L. Nieburg, Political Violence (New York:
14 Michael Lipsky uses this phrase in "Protest as St. Martin's Press, 1969), p. 129; and "The Threat of
Political Resource," American Political Science Re- Violence and Social Change," American Political Sci-
view, 62 (December, 1968), 1144-1158. ence Review, 56 (December, 1962), 872.
15 James Q. Wilson was among the first to see protest "8A theory of violence and protest must somehow
as a form of bargaining in "The Strategy of Protest: confront the problem of classifying verbal violence.
Problems of Negro Civic Action," Journal of Conflict i.e., where no physical harm is actually done. I would
Resolution, 5 (September, 1961), 291-303. contend that verbal violence changes the nature of a
"IWhen the solitary politician rises in the legislative political action and goes beyond the balance of threat
chamber to object to some official action, the press and and legitimate appeal struck by protestors. Thus it
public note that Congressman X protested on the floor may be classified with active violent behavior for the
of the House. This is to use "protest" generically to purposes of understanding the dynamics of categories
refer to vehement objection. of actions.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 14 The American Political Science Review Vol. 67

about what such obviously angry behavior data, is stimulated by specific grievances related
could lead to. Participants, targets, and third- to particular local agencies or officials. Whether
parties have come to understand that the pro- or not protest occurs in a city, then, is more
test confrontation provides a relatively fertile likely to be a function to some degree of the na-
matrix for the outbreak of violence. Indeed, ture of the local political system, while the like-
the society may react to protest by resorting to lihood of violence is not (witness the riots in
pre-emptive violence. That protest may lead to the "model" cities of New Haven and Detroit).
violence, then, is undeniable; that the latter is In addition, certain correlates of one form of
simply an extension of the former does not behavior are not present with the other. For ex-
follow, however. ample, while Downes, Spilerman, and the data
The implicit-explicit dividing line is crucial. collected for this present inquiry all show a
Protest harnesses aggressive impulses by con- strong relationship between the incidence of vi-
trolling and, to some extent, masking them, olence and the size of a city's black population,
while violence gives free reign to thse impulses. my data show no relationship between the fre-
The strategy of controlling aggression by basing quency of black protest and the size of the
behavior on the unacknowledged threat of vio- black population. Second, my data indicate that
lence follows from the cost-benefit calculation. violence, prompted either by the police or the
The advantages to the protesters of such a protesters, did not occur in the vast number of
strategy are manifold. Fear is an important mo- cases of protest, leading the observer to suspect
tive force, but the costs of acknowledging the that implicit threat manipulation and explicit
manipulation of threat to produce fear are threat manipulation are two distinct forms of
high. Once actors acknowledge the potential vi- political expression. And finally, a number of
olence which underlies their behavior, the goals scholars argue that recent ghetto violence,
sought in the action become secondary in the while politically motivated, was largely expres-
conflict. Target actors-those against whom the sive in nature, while some preliminary survey
protest is directed-may then refuse to debate evidence shows that most actors who engage in
the issues which gave rise to the protest. They protest activities, conventionally defined, do so
may turn instead to the issue of the legitimacy for instrumental purposes.22
of the means by which the demand is put. In short, the study of protest has an integrity
As long as protesters do not manipulate the all its own. The lessons of the studies of mass
threat of violence explicitly, they enjoy a slim violence in the nation's cities are not likely to
legality, even, occasionally, legitimacy. Once be fully applicable to this enterprise.
they employ their threat openly, however, they The concern of this study is to explore the
open the way for authorities to suppress their notion that the incidence of protest is in part a
movement or action. function of the nature of a city's political op-
Protest, then, is a device by which groups of portunity structure.
people manipulate fear of disorder and violence There appear to be two plausible hypotheses.
while at the same time they protect themselves One is that protest occurs most frequently in
from paying the potentially extreme costs of unresponsive and unrepresentative political
acknowledging such a strategy. systems-in other words, in cities in which the
Protest and violence may also be distin- opportunity structure is relatively closed. Pro-
guished empirically to some extent. One obvi- test may be viewed in this situation primarily as
ous point is that the ghetto violence of recent a frustrated response by groups unable to gain
years mobilized primarily blacks as demand- access to decision-making councils by conven-
makers, while urban protests have appealed to tional means.23 Groups which find or consider
both races.19 More important, several recent themselves deprived of political representation
empirical efforts have suggested that the ghetto or which cannot elicit favorable policy deci-
violence of the last decade can be viewed as a
21 See Peter Lupsha, "On Theories of Urban Vi
response to certain national forces which cut olence," Urban Affairs Quarterly, 4 (March, 1969),
across or transcend city lines: Spilerman cites 275; and Marian Lief Palley and Howard A. Palley,
vacillatory federal action, the development of "From Expressive Disorders to Issue-Oriented Poli-
tics," a paper delivered at the Annual Meeting of the
black consciousness, and the suggestive impact
American Political Science Association, September,
of the national news media, particularly televi- 1970, 5-6.
sion.20 Protest against local government targets 22 Peter K. Eisinger, "Protest Behavior and the Inte-
on the other hand, judging from the present gration of Urban Political Systems," Journal of Poli-
tics, 33 (November, 1971), 989-990.
9The analysis presented here makes an attempt 23toThis hypothesis finds support in the vast literature
control for the racial composition of the protest inci- on relative deprivation and psychological response to
dents under examination. frustration. For a summary of relevant materials, see
29 Spilerman, pp. 1-2. Ted Gurr, Why Men Rebel (Princeton, New Jersey:

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 1973 The Conditions of Protest Behavior in American Cities 15

sions from government will resort to the drama only from those cities with populations between
of protest to make it appear that some sort of 100,000 and one million. Of the 141 cities
felicitous response is imperative for the preser- which fell in this category in 1968,25 an origi-
vation of social peace. Protest, then, springs nal random sample of 56 was chosen. Newspa-
from the inability of certain groups to manipu- pers in 13 of these cities were not available on
late the political system to their advantage by microfilm. Hence, the final sample numbered
conventional means: the incidence of protest 43.
will vary negatively with indicators of an open Newspapers for a six-month period (May-
structure of political opportunities. This may be October) in 1968 were read. Where there was
designated the linear model of the conditions of more than one major newspaper, an increas-
protest. ingly rare circumstance in American cities, the
A second hypothesis conceives protest as afternoon journal was selected. This time pe-
much a signal of impatience as frustration. Ac- riod was deliberately chosen in order to control
cording to this model-the curvilinear model- for seasonal variations in the likelihood of pro-
protest occurs as a political system begins to test and to enhance the chances of obtaining a
open up.24 Or to put the hypothesis in such a large sample of protests. It was assumed that
way as to allow for testing in the absence of the warm spring through autumn months are
time-series data, protest will be most prevalent most conducive to the pursuit of outdoor politi-
in systems characterized by a mix of open and cal activities such as marches, picketing, or
closed factors. Protest is not likely to occur in mass gatherings. Despite the relatively short
extremely closed (repressive) systems or ex- time period, the effort of reading approximately
tremely open (responsive) systems. Hence the 180 daily editions for each of 43 cities consti-
relationships of system characteristics and the tuted an enormous task.
incidence of protest will be curvilinear. Protest The decision to use newspapers as the source
occurs in a mixed system because the pace of of the protest data was motivated by two con-
change does not keep up with expectations, siderations, one practical, the other theoretical.
even though change is occurring. As the politi- As for the former, there simply is no single offi-
cal opportunity structure becomes more open, cial or nonofficial tabulation of protest inci-
previously powerless groups begin to acquire dents in existence. Reliance on newspaper ac-
influence. The acquisition and development of counts, however, has substantial justification in
influence, however, is likely to come slowly. some theoretical writing on the dynamics of
Conventional strategies of political influence protest. According to Lipsky, protest activities
may appear too slow and unwieldy to effect sig- create resources for protestors by activating
nificant gratification. In a system which is sympathetic third-party interests, to whom the
opening up, the realization that the system may protest target is sensitive, to bring pressure to
be vulnerable or responsive to political efforts bear on behalf of the protestors.26 The critical
combined with the persistence of inequities be- link between the protestors and their targets'
comes intolerable for some groups. Hence, third-party reference publics is the news media.
these groups may resort to protest to express Lipsky writes:
their impatience, even when the system may be To the extent that successful protest activity de-
viewed in relative terms as a responsive one. pends upon appealing to and/or threatening other
In the following sections I shall examine groups in the community, the communications
some data gathered on protest in American cit- media set the limits of protest action. If protest
ies to determine which model best describes the tactics are not considered significant by the media,
conditions under which such activity occurs. or if newspapers and television reporters or edi-
tors decide to overlook protest tactics, protest
organizations will not succeed. Like the tree fall-
Data and Methods
ing unheard in the forest, there is no protest un-
Data on the incidence of protest activities less protest is perceived and projected.27
were gathered from the local newspapers of 43
widely scattered cities. To reduce the possible
While this conception of protest may be un-
distorting effects of size, the sample was drawn 25 The sample was drawn from the Municipal Year-
book 1968 (Chicago: International City Managers'
Princeton University Press, 1970), chap. 2, especially Association, 1968). Population and percentage of
pp. 24, 34-35, 38. blacks were taken from the 1970 census, 1970 Census
24 This hypothesis represents an amalgam of several of Population, Advance Report, U.S. Department of
classic explanations of the causes of revolution, namely Commerce, Bureau of the Census, Washington, D.C.
those of Soule and Brinton. As Soule has written, only 26Lipsky, p. 1146.
after the position of desperate people "is somewhat 2'Lipsky, p. 1151. Compare Turner's observation
improved and they have sensed the possibility of that protest cannot be projected as protest unless it
change, do they revolt effectively against oppression conforms to folk-theories of what protest is ("The
and injustice." Quoted in Gurr, p. 114. Public Perception of Protest," p. 818).

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 16 The American Political Science Review Vol. 67

necessarily narrow (it is possible to conceive Table 2. Range of Protest Frequency

protest incidents in which interaction occurs
privately between the protestors and their tar- Number of Protests Number of Cities
get), Lipsky's formulation offers a significant
rationale for the data source: Newspaper cov- 0 5
1 11
erage, however dubious the scholar might be
2 8
about its objectivity of comprehensiveness, is
3 6
vital in itself to the projection of protest. 4 4
In the selection of incidents, the research was 5 4
guided by conventional definitions of protest. 6 2
Microfilm readers were instructed to record the 7 1
details of all marches, sit-ins, demonstrations, 8
pickets, protest meetings or rallies, and any 9 1
10
other incidents which might possibly be con-
11 1
strued as collective protest. Much depended on
how the participants defined their actions.28 43
From this collection of incidents only those
which were carried out by two or more individ-
uals and directed at local governmental agen- shows the range of protest frequency. Protest
cies, officials, or institutions were finally coded.
against city government targets is clearly not a
The data do not include, then, campus protests frequent phenomenon. Five of the cities (three
or antiwar or antidraft protests. No instances of of them in the South: Fort Lauderdale, Win-
rioting, ambushes, shootings, assaults, looting ston-Salem, and Newport News) had no pro-
or threatened violence were recorded, even if test at all during this time period. Only five cit-
the perpetrators or their victims attributed such ies averaged one or more protests per month.29
behavior to political motivations. If, however, In general the cities averaged slightly fewer
violence broke out at the site of a peaceful pro- than three protests.
test and was a direct result of it, this was re- Larger cities had a greater tendency to expe-
corded. rience protest than smaller ones: while 28 per
A total of 120 protest incidents was finally cent of the cities in the sample have popula-
selected and coded. The characteristics of these tions over 500,000, 45 per cent of the protests
protests very widely, providing a multi-faceted occurred in those cities. City size and incidence
portrait of protest in American cities in the late
of protest are positively correlated (r = .46).3
1960s. Table 1 provides a summary of the dis- Estimates of the number of participants in
tribution of protests by region, and Table 2 each protest incident were reported in 96 of the
120 cases. The size of the protests ranged from
Table 1. Distribution of Protests by Region five to over 2000. Table 3 shows the range of
protest participation. A conservative estimate,
Number of Protests based on a computation of the median number
Region of protestors in each category times the number
0 1-2 3 or more of protests, is that over 17,000 people took part
in the 96 protests, an average of approximately
Northeast 1 3
180 per incident.3'
Midwest 1 5 5
South 3 8 2 Urban protest is a strategy used predomi-
West 1 5 9 nantly by blacks,32 a finding substantiated by
this body of aggregate data. In those instances
Totals 5 19 19 in which the race of the protestors was identi-
fied (95 out of 120 cases), 49 (52 per cent)
protests were composed entirely of blacks, 21
2' The case of an American court-martial, held in
Britain in the summer of 1971, comes to mind in
which the guilt of the defendant hung on the definition 29 These were San Francisco, Milwaukee, Sacra-
of an act he had admittedly committed. The defendant, mento, Boston, and Seattle.
an army lawyer, had been accused of organizing a 30 Unless otherwise noted, the correlations reported
demonstration against the war (which is illegal for here are significant at the .05 level.
Americans in uniform overseas) while the lawyer 31 Some participants may have taken part in more
argued that all he had done was to present a petition than one protest in a given city during the time span
at the American embassy, which is legal for military of this study, but there is no way of controlling for
personnel. In other words, the defendant refused to multiple participation.
define his action as protest. 32 Eisinger, p. 1005.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 1973 The Conditions of Protest Behavior in American Cities 17

Table 3. Size of Protest Incidents Table 5. Outcome of the Protest

Number of Participants Number of Protests Concessions made to the protesters 15% (18)
No concessions made 54% (65)
1-10 3 Action deferred, postponed 24% (29)
11-25 10 Outcome undetermined 7% (8)
26-50 18
51-75 5 100% (120)
76-100 18
101-200 15
201-300 10 In the next section the protest incidents are
301-500 9 treated as identical units of analysis in explor-
501-1000 3 ing the basic conditions associated with protest.
1001-2000 3
Since the intensity of individual protest inci-
2001 or more 2
dents differs, however, depending upon the
undetermined 24
number of participants, the duration of the pro-
120 test, and the number of sites at which a single
incident occurred, a later section will explore
the environmental conditions related to rela-
(22 per cent) were mixed black and white, and tively intense protest activity.
20 (21 per cent) were exclusively white.33
More of the protests were directed against Incidence of Protest and Environment
the city school system, particular schools, or An initial look at the zero-order correlations
school officials than other target (39 per cent; between incidence of protest and selected envi-
N = 47). The city council was the second ronmental variables suggests that what I have
most frequent target (12 per cent; N = 14), called the curvilinear model is more applicable
with the police (10 per cent; N = 12) and the to the American case than is the linear model.
welfare department (8 per cent; N = 9) next Protest occurs most frequently in cities in
in order. Other city agencies, the mayor, and which the political opportunity structure is
the city courts provided the bulk of the re- characterized by a mix of open and closed vari-
maining targets. bles.
All but seven (6 per cent) of the 120 protest One standard set of variables relates to the
incidents were carried out peacefully. Violence formal structure of local government. Scholars
was instigated in several of the instances by the have argued that mayor-council governments
protestors, according to the newspaper ac- are more accountable, and hence more avail-
counts. In the vast majority of cases, protestors able, to citizens than manager-council govern-
probably did not explicitly threaten violence; if ments. The mayor is an elected politician and
they had done so, the authorities would likely must please a constituency; the manager is a
have taken pre-emptive steps to halt the mani- professional, hired by the city council, who
festation. maintains his office at their sufferance. Simi-
As a tactic, protest appears effective largely larly, ward aldermanic elections afford residen-
as a means of breaching the political opportu- tially concentrated minorities greater opportu-
nity structure rather than manipulating it for nity for representation than at-large electoral
group ends. On the one hand, protestors were systems. In the latter the major groups in the
successful in a majority of cases in gaining ac- city dominate the electoral arena. Partisan sys-
cess to their targets to state their demand (Ta- tems seem to offer groups greater access to gov-
ble 4), but in only a fraction of instances were ernment than nonpartisan ones by virtue of the
concessions actually made (Table 5). fact that parties aggregate diverse interests as a
requisite of persistent viability and rely over
Table 4. Nature of Target Response at Site of Protest
time on identifiable blocs of voters for whom
they supply cues and to whom they must ac-
Met with or spoke to protestors 58% (69)
count.34
Refused to meet with or speak to protestors 35% (42)
In short, insofar as the particular institu-
Undetermined 7% (9)
tional arrangements which characterize reform
100% (120) governments make access and representation
more difficult for minorities to obtain, such
33 Mexican-Americans were responsible for the re-
mainder of those protests in which participants were 3 Lineberry and Fowler, "Reformism and Public
identified by race or ethnicity (4 per cent). Policies in American Cities," p. 715.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 18 The American Political Science Review Vol. 67

governments offer a relatively closed structure related to black protest, if the form of govern-
of opportunities. Some evidence indicates, in ment were simply an intervening variable.
fact, that the impact of reform institutions is However, black protest and black population
cumulative in this regard: that is, the more are not related, lending support to the finding
components of reform a government has-non- of the association of structure and protest.
partisanship, manager, and at-large elections- Formal institutional arrangements represent
the less responsive it is to minority group de- one aspect of the political opportunity structure
mands.35 in the sense that they seem to accord advan-
Protest, however, does not appear to be a re- tages in political competition to certain groups
sponse to a closed system of opportunities as and disadvantages to others. The nature of the
represented by formal governmental arrange- formal political structure, however, does not
ments, taking each institutional form individu- necessarily indicate where power lies or how it
ally. The incidence of protest is positively re- is distributed in the system.38 If power is con-
lated to mayor-council forms (r = .35).36 centrated, then groups without membership in
Black protest is even more strongly related to the circle of the select which attempt to enter
mayoral government (r = .42).3 Protest oc- political competition will neither have many
curs most freely, then, in political systems in points of access at which to make their case nor
which at least one aspect of formal structure- a great chance of acquiring significant allies.
the nature of the chief executive-has been Indeed, if power is concentrated, then it is
viewed as most conducive to groups seeking ac- likely that the resources necessary to wield
cess. power are also concentrated, indicating that
Taken as isolated variables, neither the ar- groups which seek to enter the political arena
rangements for electing aldermen nor the fact are likely to be poor in the necessary political
of partisanship bears any relation to the inci- currencies, or that there is, in Dahl's terms, lit-
dence of protest. However, combining these tle slack in the system.
two variables with the nature of the chief exec- In the terms of the present argument, one
utive to get a multiple correlation (.41) affords could claim support for the linear model of the
a 40 per cent increase in explained variance relationship between protest and the environ-
over the simple correlation of mayoral govern- ment if the incidence of protest rose in cities
ment with protest (.35). The three variables with a centralized distribution of power. Where
taken together explain 17 per cent of the vari- power is concentrated, one plausible response
ance, while mayoral government alone explains for outgroups, low in conventional resources
only 12 per cent. and cut off from regular access to the stations
It would be plausible to suspect that the rela- of power, would be to use protest.
tionships shown here are not in fact reflective On the other hand, if protest were found to
of the impact of formal structure but rather are increase with the dispersion of power, then the
a function of the type of population found pre- frustration model would be less appropriate.
dominantly in mayor-council cities. Since the With the possibilities manifest for coalition
number of blacks is related to mayor-council building, for acquiring allies with significant re-
government (although not at significant level), sources, groups which resort to protest do so
we should find that the number of blacks is also in a relatively open system. Such behavior
might be viewed in terms of Lipsky's formula-
I Lineberry and Fowler, pp. 713-714. The difference
tion of protest as a device to communicate the
between reform and traditional components is illus-
trated in the present body of data by the fact that need for allies with resources.39 This formula-
blacks are slightly more likely to have nearly propor- tion presupposes a dispersion of power: some
tional representation on the city council under a ward actors in the system might be favorable to the
system than under either a combined at-large/ward
system or an at-large system. The correlation coeffi- "Aiken has argued that reform governments and a
cient between the degree of proportional representation concentrated distribution of power are positively re-
and ward elections is quite small (r = .16) but it is lated, but his relationships are neither strong nor sig-
in the predicted direction. nificant. "The Distribution of Community Power:
"I This is significant at the .02 level. Form of govern- Structural Bases and Social Consequences" in The
ment was treated as a dichotomized dummy variable Structure of Community Power, ed. Michael Aiken
with the one commission government in the sample and Paul Mott (New York: Random House, 1970),
grouped with the manager governments. Robert Alford pp. 499-500. J. David Greenstone and Paul Peterson
and Eugene Lee found that grouping commission and however, have argued the exact opposite in "Reform-
manager governments for a procedure identical to the ers, Machines, and the War on Poverty" in City Poli-
one used here did not alter the correlation in a signif- tics and Public Policy, ed. James Q. Wilson (New
icant way. See their "Voting Turnout in American York: Wiley, 1968), p. 289. Their examination was
Cities," p. 803. based on a study of only four cities.
37 Significant at the .006 level. 39 Lipsky, pp. 1146-1147.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 1973 The Conditions of Protest Behavior in American Cities 19

protestors and will join the conflict against Table 6. Intercorrelations (Pearson's r) of Protest and
those who are not. If protestors have potential Selected Political Environment Variables
allies, then the opportunity structure is not
fully closed. Protest is a function of impatience All Black
in a system marked by some degree of flexibil- Protest Protest
Incidents Incidents
ity.
One problem in testing these alternative pos-
Population of city .46* .50*
sibilities is that summary measures of the dis-
Percentage black - .02
persion of power in a community are difficult Number of blacks .14
to come by. Nevertheless, there have been sev- Model cities grant . 26** .23
eral creditable attempts, the most influential Number of black elected officials .07
and controversial of which has been Hawley's Percentage of managers, propri-
MPO ratio, the proportion of managers, proprie- etors, and officials -. 30* -. 28**
Index crimes per 1000 population .34* . 27**
tors, and officials in the civilian labor force.40
Percentage of minority police .20 .03
Hawley argued that a high MPO ratio indicates
Mayor-council government .35* . 42*
that power is diffused among different commu- Ward aldermanic elections -.01 .08
nity subsystems, power which rests on manage- Partisanship .07 .11
rial skills. Hawley wrote: Incidence of riots .23 .22
Black representation ratio - . 30*
Proceeding from the notion that system power
resides in the subsystems or functional units of a
community, we can infer that it must be exercised a Data for eight cities were not availab
through the managerial functions of the sub- b This is a measure of proportional representation
systems. For it is those functions that co-ordinate of blacks in the city council, computed by dividing the
the several other functions in their respective proportion of black aldermen by the proportion of
subsystems and articulate the latter with the black population.
larger system." * p <.05.
** p <.10.
Hawley demonstrated successfully that the
greater the proportion of such white collar oc- incidence of all protests and of black protests
cupations in a city's work force, the less effec- is negatively related to a high MPO ratio (r =
tive a city was in implementing urban renewal -.30 and -.28, respectively). That is to say,
programs. Success, he argued, is a function of protest occurs more frequently in cities with a
the concentration of power. small white collar work force, an indication
The need to infer power from occupational that certain resources for the exercise of power
status, however, seems to me a serious flaw in are concentrated.
the MPO ratio as a measure of the dispersion The MPO ratio is positively related to
of power. Nevertheless, the ratio is not a mean- mayoral government (but not at a significant
ingless one. It affords a measure of the distribu- level), but the relationship between the MPO
tion of occupational status and skills, two criti- ratio and protest is not a function of form of
cal resources for the exercise of power. As such government: the MPO ratio is still negatively
it offers one means of estimating the disperison related to the incidence of protest when form
of potential power, or the degree of slack in the of government is controlled (partial r =
system.42 If the potential for wielding power is -.25) .43
limited in a system, then it is still a relatively
41It could be argued that the association of protest
closed system. with occupational makeup of a population is simply
In the cities selected for this study, the an indication that protest occurs where there are
greater numbers of people holding low status occupa-
40Hawley, "Community Power and Urbantions.
Renewal
That is, protest is not a response to the systemic
Success." Aiken has run the MPO ratio against his characteristic of resource concentration but rather is
own measures of power concentration and comes up a function of the existence of large numbers of poten-
with the exactly opposite conclusion from Hawley. tial protestors. (This assumes that, at least in the
Aiken is not entirely sure how to explain the difference aggregate, protest and lower-class status are related.
and ends up urging caution in the use of the measure. Survey data I have collected in Milwaukee, to be re-
See p. 503. ported elsewhere, show no relationship at the individ-
4Hawley, p. 424. ual level between low status and protest participation.)
42 The distinction between potential influence and Nevertheless, if this were the case using aggregate
manifest influence, or influence in repose and influence rather than survey data, we could explain Hawley's
in use, is germane here. See Robert Dahl, Modern findings by arguing that the larger the percentage of
Political Analysis (Englewood Cliffs, New Jersey: MPO's, the better-off the population, and the fewer
Prentice-Hall, 1970), pp. 28-29; and William Gamson, would be the policies geared to social welfare pro-
Power and Discontent (Homewood, Illinois: Dorsey, grams designed to aid low income groups. However,
1968), Chapters 4 and 5. the findings of Paulson and his colleagues suggest that

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 20 The American Political Science Review Vol. 6X

Table 7. Form of Government and Frequency of Protest

0 1-2 3 or more Totals

Mayor-council 6% (1) 29% (5) 65% (11) 100% (17)

Manager 15% (4) 54% (14) 31% (8) 100% (26)

Thus, protest appears to occur in cities in antee blacks real power in a city, but it does
which the formal political structure provides indicate that certain types of political opportu-
opportunities while the informal structure, as nities are available. Furthermore, it provides
measured by the distribution of occupational evidence that black spokesmen may speak from
and skill resources, is relatively restricted. For- a legitimate and official forum. This is not a
mal attributes of government may have little prerequisite for the exercise of power but it
bearing on the exercise of power. Structure makes the problem of wielding power a lesser
provides a framework within which certain one in that access has already been achieved.
groups are accorded advantages if they are able The incidence of black protest is not related
to seize them. The mayor is more vulnerable to the absolute number of black elected offi-
than the manager because the former must cials, including aldermen, school board mem-
please an electorate. But this vulnerability, a bers, and judges, but it is significantly related to
consequence of formal structure, assumes that the black representation ratio (r - .30). The
the mayor is in a position to please an elector- representation ratio was calculated by dividing
ate by being responsive and that organized seg- the proportion of aldermen who are black by
ments of the electorate can make demands on the proportion of the population which is
him. But the realities of the potential distribu- black. When the figure equals 1.0, then blacks
tion of power, symbolized by the MPO ratio, are, numerically speaking, perfectly propor-
may render the effects of structural biases im- tionally represented. When the figure is under
portant. 1.0, the typical case, blacks are underrepre-
While this is a plausible argument, evidence sented. In a few cases, the ratio exceeds 1.0,
indicates that blacks engage in protest espe- indicating that the black population has more
cially in those cities where they have been able than its proportional share of aldermanic repre-
to gain formal proportional representation. sentatives.
Representation in the form of elected officials As the representation ratio approaches zero,
-limited here to aldermen44 -does not guar- we may conclude that the structure of opportu-
nities is less open for blacks. In the case of the
the MPO ratio affords more than a simple indicator cities in our sample, however, protest increases
of the socioeconomic composition of the population
with the degree to which blacks are represented
and implies a distinctive configuration of power.
Paulson et al. discovered that North Carolina coun- in city government. Protest is associated here
ties with small percentages of MPO's-which would with an open structural characteristic. Accord-
indicate a relatively low level of socioeconomic well- ing to the linear model, it would have been rea-
being-had lower welfare expenditures than those
sonable to expect that black protest occurred
counties with high proportions of MPO's. They con-
clude that a high concentration of power can block most frequently in those cities in which blacks
what elites feel are undesirable programs, despite had been denied representation. Instead, pro-
demonstrable needs of the population. Wayne Paulson, portional representation may be viewed either
Edgar W. Butler, and Hallowell Pope, "Community
Power and Public Welfare," American Journal of
as an invitation to pose demands (through pro-
Economics and Sociology, 28 (January, 1969), 17-27. test, if that is the easiest way to mobilize sup-
4 Data on black elected officials was drawn from the port) or as a concomitant factor in a black
National Roster of Black Elected Officials, compiled
population's coming of political age, a matura-
by Metropolitan Applied Research Center, Washing-
ton, D.C. and Voter Education Project, Southern Re-
tion process marked by twin developments in
gional Council, Atlanta, Georgia, February, 1970. protest and electoral activity.

Table 8. Model Cities and Protest

Number of Protests

0 1-2 3 or more Totals

Model Cities grant 4% (1) 42% (10) 54% (13) 100% (24)
No Model Cities grant 21% (4) 47% (9) 32% (6) 100% (19)

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
1973 The Conditions of Protest Behavior in American Cities 21

Table 9. Minority Policemen and Protest

Percentage of Minority Number of Protests

Policemen 0 1-2 3 or more Totals

0-5% 18% (3) 59% (10) 24% (4) 101% (17)

6-10% 8% (1) 25% (3) 67O (8) 100%0 (12)
11% or more - 67%o (4) 33%0 (2) 100% (6)

Note: Data for eight cities were not available.

Formal political structure, the distribution of 1966-67.48 Of the sample of 43 cities in this
resources, and the nature of the incumbents in study 24 (56 per cent) had received planning
elected office all give indications of the poten- grants by the end of 1968, and 19 had not. Be-
tial various groups might have for wielding po- cause the Model cities program was thought to
litical influence. However, none of these indica- offer the means for a comprehensive attack on
tors supplies any sense of the extent to which a urban ills, the city which gained such funds
political system is in fact responsive to deeply could be said to have acted responsively to de-
felt needs. If a government does not demon- mands that were at the very least implicit in the
strate a willingness to respond to a diversity of urban condition.
demands, then the structure of opportunities Similarly, after the ghetto riots in the last de-
cannot be said to be a fully open one, regard- cade, one of the major prescriptions for restor-
less of the opportunities for gaining access and ing racial peace was to increase minority group
representation. representation on the uniformed police force.
To test the responsiveness of local govern- Most cities sought, unsuccessfully, to recruit
ments, two measures of policy outputs were blacks for their police deparments.49 The mea-
chosen. One was whether a city had received sure used here-the percentage of the force
Model Cities planning funds45 and the other composed of minority group members-is nec-
was the percentage of the police force com- essarily a static one. Data do not exist in reli-
posed of racial minority members.46 The for- able form prior to 1969. Thus, the measure does
mer may serve as some indication of respon- not take account of changes in the percentage
siveness to what is often the most frequently of blacks. Nevertheless, it may be assumed that
named problem facing the cities-the lack of simply having at a given point in time a sub-
decent housing.47 The latter is a response to the stantial proportion of blacks on the force is an
animosities which mark the relationship be- indication that the political system has at some
tween urban police forces and the black com- previous time been responsive. And cities with
munity. a greater percentage may be said to be more
Initial competition for Model Cities funds responsive in this regard than cities with a
were intense. Nearly 200 cities submitted appli- smaller percentage.
cations for the first round of planning grants of Both the existence of a Model Cities pro-
which only 75 were chosen during the winter of gram and the percentage of minority group po-
licemen are positively related to the incidence
4 This was set up in terms of a Model Cities grant/ of protest (.23 and .20, respectively), but only
no Model Cities grant dummy variable. The source for
the former is statistically significant. Black pro-
this information is 1968 HUD Statistical Yearbook
(Washington, D.C.: Department of Housing and
test is also slightly related to Model Cities, but
Urban Development, Government Printing Office, the coefficient falls just short of our criterion of
1968). There are no data concerning those cities which significance. Black protest is not at all related
applied for Model Cities funds but were turned down.
to the percentage of uniformed minority police-
46 Municipal Yearbook 1970 (Chicago: International
City Management Association, 1970) provided the
men.
source for these data. These relationships, while small and in most
41 Survey data I have collected in the city of Mil- cases not significant, lend more support to the
waukee indicate that housing ranks consistently first
for members of both races as "the most important 48The Model Cities Program: A Comparativ e Analy-
problem facing this city." This is a finding duplicated sis of the Planning Process in Eleven Cities (Washing-
in other surveys. See, for example, Bernard J. Frieden, ton, D.C.: Department of Housing and Urban Devel-
"Housing and National Urban Goals: Old Policies and opment, Government Printing Office, 1970), p. 7.
New Realities," in The Metropolitan Enigma, ed. 49 Paul Delaney, "Recruiting of Negro Police Is a
James Q. Wilson, (Cambridge, Mass.: Harvard Uni- Failure in Most Cities," New York Times, January 25,
versity Press, 1968), p. 162. 1971, p. 1, cols. 2, 3, p. 14, cols. 3-8.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
22 The American Political Science Review Vol. 67

curvilinear model than to the linear one. That Table 10. Model Cities, Riots, and
is, protest increases with responsiveness. Average Number of Protests
It should be said, however, that a more de-
tailed examination of these two variables would Model Cities No Model Cities
probably reveal that they are complex and am- Grant Grant
biguous measures of responsiveness. Some of
Average Number of
the cities in this sample which have Model Cit-
protests per city 3.2 2.1
ies programs were more committed than others
to lay involvement in planning.50 And literature Model Cities No Model Cities
on the tensions experienced by black police- Grant and Riot Grant and Riot
men51 reveals that the cross-pressures felt by
such men may drive them to brutal behavior Average Number of
against other blacks as a way of demonstrating protests per city 3.6 2.7
professional loyalty in the face of race demands.
Both Model Cities and the black policeman may
represent mixed blessings in some sense, at least tion is convincing. Of the twelve cities which
for the black community. In other words, no had neither a riot nor a Model Cities grant,
Model Cities grant and no black policemen seven are located in the South. They are char-
would be signs of a failure to respond; but hav- acterized by large black populations (potential
ing such programs is not necessarily an unam- protest participants) and the usual range of ur-
biguous sign of responsiveness.52 Such pro- ban ills.
grams are likely, then, to represent a mix of re- The most plausible explanation is that the
sponsive and nonresponsive impulses, a mix political systems in the cities which did not par-
characteristic apparently of the conditions un- ticipate in Model Cities are so closed or so un-
der which protests take place. responsive that protest was either discouraged
In those cities which did not obtain Model or felt to be futile. The data tend to support
Cities funds, protest was not a frequent occur- such speculation, even in the case of those cities
rence. Several explanations are plausible. One which experienced riots (Table 10).
is that the demand for such a program did not The degree to which political opportunity
exist. This is unlikely, however, since at least structures are open or closed is not only a func-
seven of the 19 cities experienced mass racial tion of formal governmental structure, the dis-
disorders in 1967 or 1968,53 sufficient indica- tribution of certain skills and status, representa-
tion that conditions in those cities posed at least tion, and governmental responsiveness. Oppor-
an implicit demand for massive governmental tunity is also related to the social stability of
response. the potentially mobilizable population. Where
Another explanation is that in those cities population groups exhibit high social disorgani-
which had neither riots nor Model Cities zation, it might be expected that opportunities
grants, conditions were such during the time for elites seeking to mobilize political consti-
period of this study that the population felt no tuencies would be low. As a measure of social
grievances deeply enough to participate in pro- disorganization, this study employed the rate of
tests. It is also possible that the population those crimes reported in the FBI Uniform
groups from which protesters are generally Crime Reports 1968 (so-called "index crimes,"
drawn is simply too small to organize sustained because they comprise the national crime in-
protest actions in these cities. Neither explana- dex) per 1000 population.54 The incidence of
-"For example, Milwaukee, Atlanta, and Gary, all
protest, however, is positively related at a sig-
cities in our sample, were less likely than some of the nificant level to high crime rate (r = .34).
other cities to share power over the program with local Black protest is related less strongly to crime
residents. See The Model Cities Program. . . rate and the correlation is just short of signifi-
"' Nicholas Alex, Black in Blue: A Study of the
Negro Policeman (New York: Appleton-Century-
cance. Gurr cites evidence which indicates that
Crofts, 1969), p. 161. during periods of protest activity, protest dis-
32 Naturally, such an argument means that the open places aggressive black crime.55 The aggregate
system cannot be measured by these variables: they
only indicate closed and mixed systems. This is true 54Variations in reporting practices make interjuris-
when the variables are treated as gross indicators of dictional crime rate comparisons a delicate task, yet
responsiveness. If they could be combined with a lay I have used such a measure in the absence of other
participation index and a black policeman brutality readily available summary measures.
index, for example, then the variables would offer 30 F. Solomon et al., "Civil Rights Activity and Re-
greater discriminatory power. duction in Crime Among Negroes," Archives of Gen-
53 These included Phoenix, Elizabeth, Sacramento, eral Psychiatry, 12 (March, 1965), 227-236, cited in
Long Beach, Rockford, Las Vegas, and Miami. Gurr, Why Men Rebel, p. 310.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 1973 The Conditions of Protest Behavior in American Cities 23

data presented here cannot assess changes in of social conditions which breed such deviance.
the crime rate which might occur during peri- In addition, protest is associated with a concen-
ods of protest activity, but as a static relation- tration of occupational status, which may be
ship, crime and protest appear to occur to- construed as one measure of the distribution of
gether. Such a finding indicates that the degree potential power resources. The structure of po-
of social disorganization reflected at least by litical opportunities cannot be said to be open to
crime rates does little to impair the ability of the fullest extent under such conditions. Yet
elites to organize protests. the opportunities for gaining access to and for-
In a similar way the incidence of both pro- mal representation in government, characteris-
test and black protest is also related to the oc- tics of an open system, are good in cities which
currence of ghetto riots. Cities were coded ac- experience protest. And finally protest is mod-
cording to whether they had experienced no ri- erately associated with certain very specific
ots, one riot, or more than one during the peak government responses in the form of Model
years of rioting, 1967 and 1968.56 The inci- Cities and the recruitment of minority group po-
dence of all protest and-of black protest in licemen, but these responses do not indicate un-
particular-is slightly related to riot violence, ambiguously that the structure of opportunities
but neither coefficient (r = .22 and .23, respec- is an open one. In short, the curvilinear model
tively) is significant. To the slight extent that describing the conditions of protest seems more
protest and rioting occur together, it would be accurate than the simple linear model. In the
possible to conclude that the ghetto violence of final section an attempt is made to elaborate on
1967-68 did not displace protest activity. The this notion and to interpret the considerable
time period of this study occurs after the last gaps left by unexplained variance in the data.
wave of rioting, which followed Martin Luther Before commencing this effort, however, we
King's assassination in April, 1968. The impul- must consider the problem of the intensity of
ses to protest were not exhausted either by that protest and the political environment.
violence or the earlier riots of 1967, for protest
Intensity of Protest
continued after the major outbreaks. To the de-
gree that the recent ghetto violence was a form Protest incidents differ in the intensity of the
of political expression, it was in all likelihood a concerns they express and the reactions they
mode of expression qualitatively different from generate. The participants' and bystanders' pas-
protest and not the logical outcome of the lat- sions associated, for example, with the fair
ter. housing marches led by Father Groppi or Mar-
In a speculative way, these findings on the tin Luther King into working class ethnic
relationship of crime, rioting, and protest are neighborhoods represent a quantum leap in in-
suggestive for characterizing the political sys- tensity over those feelings associated with the
tem itself: insofar as crime is an individualistic small protest delegation which meets with the
response to poor social conditions and mass vi- city welfare commissioner far from the public
olence is a collective response, high crime and eye. As a unit of analysis, then, one protest in-
riot cities are those with serious problems. Sub- cident is not necessarily strictly comparable to
stantial problems indicate relative governmental another.
impotence, even in those cities with the most This is not to say that the foregoing exercise
responsive administrations. Impotent govern- has been misguided. In one sense protests may
ment is one mark of a closed structure of politi- be treated as comparable in assessing the inten-
cal opportunities: if government is powerless to sity of a city's protest environment. As we shall
act, then political action to influence govern- see presently, the mere number of protests is as
ment is futile. good an indicator as any of the general state of
To summarize, this section has shown that tension and challenge wrought by protest be-
protest occurs most frequently in cities whose havior in any given city. A city which experi-
structure of political opportunities reflects a enced five protests during the time period of this
mix of open and closed characterisitcs. The ex- study had a more intense protest environment
istence of both a high crime rate and mass vio- than the city which had only one incident. Even
lence signifies either a generalized governmen- if the single protest was large and passionate, it
tal impotence or unresponsiveness in the face may be possible to view it as an anomaly, after
which politics settles down or returns to its more
I For riots in 1967 the list on pages 158-159 of the conventional patterns. But five incidents bespeak
Report of the National Advisory Commission on Civil
a relatively consistent state of crisis, an indica-
Disorders (New York: Bantam Books, 1968) was used.
For 1968 riots, the New York Times Index supplied tion that things are more often felt to be wrong
the data. and remediable by protest action by more groups

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 24 The American Political Science Review Vol. 67

of different people than in cities experiencing The intensity of the protest environment
a single protest. score was calculated by the formula:
Nevertheless, the mere number of protests
duration + sites + size
tells nothing of the attention any single protest
incident might have engaged or the scope of its number of protests
impact as a focus for mobilization. Conceiv- + number of protests
ably, the conditions which give rise to one ex-
The formula for figuring the average intensity
tremely intense protest might differ from those
of protest incidents was as follows:
which are associated with milder but more nu-
merous protests. duration + sites + size
To measure intensity two basis scores were
number of protests
computed. One was an indicator of the inten-
sity of the protest environment, and it was em- The intensity of the protest environment
ployed largely to test the reliability of the find- score is, of course, closely related to the num-
ings using the simple frequency of protest. The ber of protests (r .90), but it offers a some-
second measure was an indicator of the average what finer measure. Average intensity is not
intensity of individual protest incidents in a so strongly related to the frequency of pro-
city. tests (r = .41). The two scores are themselves
Intensity was conceived for measurement related (r = .75).
purposes as a function of the number of partici- The intensity of protest environment correla-
pants in a protest action, the duration in con- tions tend to corroborate and strengthen the
tinuous days, and the number of separate sites conclusions reached in the previous section. Ta-
within the city at which the protest simultane- ble 11 provides a comparison of the correla-
ously occurred. The assumption was that as tions between simple frequency and environ-
these factors increased, the passions of the ment variables on the one hand and intensity of
principals and observers would grow deeper.57 the protest environment and environment vari-
For example, a school boycott and picketing ables on the other.
which lasted four days and affected three high Using the intensity of the protest environ-
schools was more intense than one which lasted ment score instead of frequency does not in-
four days but affected only one school. The crease our understanding of the conditions as-
scores are crude in that they offer no direct sociated with protest in a significant way. The
measure of emotional intensity (which must be average intensity of protest incidents is of
inferred). Another problem is that they assign greater interest, however. Whereas most of the
the same values to size, duration, and number relationships between average intensity and the
of sites. Thus no conclusion about relative in- political environment achieve the same levels of
tensity is possible for two protests of the same association as those between simply frequency
size, one of which lasted one day at two sites and the environment variables, several are dif-
and the other of which lasted two days at one ferent. For example, the frequency of protest is
site. related to population size of the city, but aver-
Protests were scored in the following man- age intensity is not. Frequency is also related to
ner: 1) duration: one day, one point; two days,
two points; three or more days, three points; 2)
number of sites: one site, one point; two sites, Table 11. Comparison of Intensity
two points; three or more sites, three points; 3) and Frequency Correlations
size of protest: 5-75 participants, one point;
76-200 participants, two points; 201 or more Intensity of
participants, three points.8 A low intensity pro- Simple Protest
Frequency Environment
test scored a minimum of three points; the
Score
maximum score of a high intensity incident was
nine points. Mayor-council government .35 .33
Ward aldermanic elections -.01 -.06
Partisan elections .07 .02
" There is some point of diminishing returns, after
which a long and large protest simply becomes routine, MPO ratio -.30 -.32
and passions settle down. The measure of intensity City size .46 .41
used here takes this phenomenon into account in an Model Cities .26 .38
arbitrary way by scoring equally any duration over Riot .23 .31
three days. Minority policemen .20 .24
5 The mean number of participants per protest was Index crimes per 1000 .34 .31
180.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
1973 The Conditions of Protest Behavior in American Cities 25

mayoral government, but, again, average inten- Table 12. Comparison of Selected Average
sity shows no such correlation. Intensity and Frequency Correlations
On the other hand, the frequency of black
protests is unrelated to the percentage or the Simple Average
number of blacks in the city's population, but Frequency Intensity
of all of
average intensity shows slight positive rela-
Protests Protests
tionships with both of these variables. These
figures suggest that while the size of the black Population .46 .14
population has little effect on the likelihood Mayor-council government .35 .14
that protests will occur, it does affect the inten-
sity associated with protests. Protests involving Simple Average
blacks are likely to last longer, be larger in
it. Frequency of
Intensity
Black of
scale, and take place at more sites than those
involving whites. This finding simply reinforces Protests Protests

the notion that urban protest is a tactic more

Number of Blacks .14 .25
fully exploited by blacks as a major form of
Percentage of Blacks -.02 .21
political activity, in which the demands for
commitment of resources made on individual
participants are substantial. White protest is not in relation to others in the society is likely to
quite so enduring, nor so broad in its attraction result in aggressive political behavior. The ex-
as a focus for mobilization. planation is attractive for the fact that it sug-
gests that socioeconomic status and the propen-
The Conditions of Protest: Some Conclusions
sity to engage in protest do not covary in a lin-
The data support the conclusion that the in- ear relationship. Middle-class homeowners
cidence of protest is mildly related to the na- threatened by freeway construction may per-
ture of a city's political opportunity structure, ceive themselves as deprived in relation to
which I have conceived as a function of the de- those homeowners whose property is not
gree to which groups are likely to be able to threatened, just as poor blacks see themselves
gain access to power and to manipulate the po- as deprived in the context of an affluent society.
litical system. The configuration of those rela- Bowen et al. are able to demonstrate that
tionships affords a basis for some theoretical those blacks who perceive the greatest gap be-
speculation concerning the nature of the envi- tween "the best possible life" and their own po-
ronmental conditions most fertile for protest, sition on a Cantril Free Self-Anchoring Scale
but before turning to that task it is appropriate are those most likely to evince great approval
to sketch some additional explanations of the for protest forms of behavior.60 They do not
factors which give rise to protest as a way of show, however, that the perception of depriva-
interpreting the unexplained variance in the tion and behavior are linked. One attempt to
data presented here. These factors are not mu- relate relative deprivation of blacks, using ob-
tually exclusive of one another, nor of the na- jective indicators of social welfare for 49 cities,
ture of the political environment. Rather, they to the incidence of mass ghetto violence proved
are probably all at work to one degree or an- fruitless,61 suggesting that the various causal
other, enabling us to understand protest behav- leaps from objective relative deprivation to the
ior, like voting, as a product of multiple com- perception of that deprivation to violent behav-
plex causes and stimuli. ior in politics need further and more careful
One major factor which undoubtedly con- charting.62 Deprivation surely plays a role in
tributes to the tendency to engage in protest has the motivations of protesters, yet its nature re-
to do with the capacity of the political system mains to be explored systematically.
to fulfill value expectations. In other words, de- Another explanation which might contribute
privation, a condition which may exist in both to an understanding of why groups engage in
open and closed structures of political opportu- protest has to do with the particular array of
nity, may give rise to protest.59
0 Don R. Bowen, Elinor R. Bowen, Sheldon
This explanation assumes that the perception
Gawiser, and Louis H. Masotti, "Deprivation, Mobil-
of deprivation, whether objective or subjective, ity, and Orientation Toward Protest of the Urban
Poor," American Behavioral Scientist, 2 (March-April
59The argument of this paper would have it that 1968), 20-24.
discontent over deprivation may be expressed more 61 Palley and Palley, "Social Welfare Indicators as
easily through conventional political strategies in an Predictors of Racial Disorders. . ."
open system, whereas protest is often necessary to 62 Lupsha, "On Theories of Urban Violence," pp.
communicate discontent in a closed system. 285, 288.

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
 26 The American Political Science Review Vol. 67

Table 13. Partial Correlations of Governmental Form and Protest

Independent variables Simple r Partial r Dependent Variable

Mayor-council government .35 .41

Ward elections -.01 -.21 Frequency of protest
Partisanship .07 . 15

organizational needs which protest can fulfill. means for forcing consideration of the extraor-
For example, protest action is frequently suc- dinary, the unpopular, or the nonroutine. De-
cessful as a strategy for mass mobilization. Pro- mands for student power in the university or
test may often be undertaken primarily as a re- welfare client representation on policy-making
cruiting activity for organizations, for it is a boards are apt examples. Protest, then, occurs
way of cutting through communal apathy and when certain demands (and indeed demanders)
attracting membership through its sheer excite- represent such major departures from the con-
ment. Protest also helps elites to manipulate ventional mainstream of politics that they are
constituents' understanding of issues, for such not considered as a routine matter of institu-
mass actions and their associated rhetoric offer tionalized practice.
participants an easily comprehended Mani- Discontent over deprivation, organizational
chean explanation of the political world.63 The needs, and extraordinary demands all occur in
we-they dichotomy between protestors and tar- and are in some degree shaped by the nature of
gets is sharply drawn in protest actions and the political system. What this study has done
serves to foster group identity and cohesion. Fi- is to provide some preliminary clues to the na-
nally, protest is action, a means of demonstrat- ture of those urban political systems most con-
ing to group constituents that the organization ducive to the development of persistent protest.
can do something. By taking such action, the The data have suggested in effect that protest
protest may serve to bolster morale and confi- flourishes in a system marked by paradox. The
dence in the organization itself. In short, pro- paradox of protest is that while on the one
test may be employed by elites as a means to hand it appears to be a response to certain
sustain and enlarge their organization; pressur- closed system characteristics, it only takes place
ing targets for action may be entirely secon- on a persistent basis in systems in which other
dary. characteristics are open. The incidence of pro-
Still a third explanation of the factors lead- test was associated, for example, with a rela-
ing to protest has to do with the types of de- tively small managerial work force, with a high
mands people might make. Insofar as the issues crime rate, and with the ambiguous variables of
treated in community decision making tend to Model Cities and minority uniformed police.
be limited by the biases and predispositions of The former two are indicative of a restricted
those in powerful positions-the setters of the opportunity structure, while the latter two po-
agenda-certain types of demands will not tentially signify both open and closed charac-
command routine attention."64 Protest is a teristics.
63 Frances Fox Piven has observed that "social pro- On the other hand protest occurred more
test actions, because they offer simple and dramatic frequently in mayor-council governments
definitions of problems, may penetrate apathy and than in manager governments, and black
override puzzled disengagement bred of lack of in-
protest took place most frequently where
formation." "Participation of Residents in Neighbor-
hood Community Action Programs," Social Work, 11 blacks had gained formal representation in pro-
(January, 1966), 78. portion to their population. Both of these fac-
64This draws on the well-known argument of Peter tors characterize the system as an open one.
Bachrach and Morton S. Baratz, "Two Faces of
Power," American Political Science Review, 57 (De- The data also showed that violence did not ac-
cember, 1962), 947-952. company most protests, and that in a majority

Table 14. Partial and Multiple Correlations of Open System Characteristics and Black Protest

Independent cluster Simple r Partial r Multiple r Dependent Variable

Mayor-council government .42 .34

Model Cities grant .23 .06 .46 Frequency of black protest
Representation ratio .30 .21

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
1973 The Conditions of Protest Behavior in American Cities 27

of cases the protestors won at least the right to willingness or even inability to suppress protest
a hearing for their demands. by force, may serve as the functional equiva-
Several speculative observations can be based lent of license to protest. Such license repre-
on these findings. Those who pursue protest as sents an opportunity in the whole structure of
an ongoing tactic must in effect gain license opportunities: protest offers a chance to gain a
from the authorities in the system to do so. hearing in public councils. The openness of the
That is to say, protest will probably not be used system, in other words, is conducive to protest.
in contemporary American cities where it is In a similar way the system which responds
suppressed by violence. Violent reactions by to protest is likely by its very responsiveness to
the authorities will likely stimulate violence by encourage protest. Elites who attempt to mobi-
the potential protest population or will cause lize people to protest will fail eventually to re-
withdrawal from aggressively demonstrative cruit participants if protests are never success-
politics. Official tolerance, signified by the un- ful. Protestors must gain satisfaction through

OPEN

mayor
ward xx
partisan
mayor
at-large ward - x xx
partisan

mayor
ward x X
nonpartisan

mayor
at-large ward - xx
nonpartisan
mayor
at-large _ x x x x
partisan

mayor
at-large _ xx
nonpartisan
manager
ward _ x
partisan
manager
at-large ward -
partisan
manager
ward x x
nonpartisan
manager
at-large ward x x
non-partisan

manager
at-large
partisan
manager
at-large xx xxx xx xx xx x x
nonpartisan /

CLOSED

No. of Protests 0 1 2 3 4 5 6 7 8 9 10
Figure 1. Scattergram Distribution of Protest Fre

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms
28 The American Political Science Review Vol. 67

protest on occasion or they will stop using it as the drama of protest. Using a cluster of open
an instrumental tactic. Some survey data indi- characteristic variables to predict the incidence
cate that those who take part more often in of protest indicates at least that the rate of in-
protest are more likely to believe that protest crease in predictive value tails off sharply. If
works than those who take part less frequently.65 those mayoral governments which procured
Protest, then feeds on the responsiveness it Model Cities funds and in which blacks have a
succeeds in eliciting. System responsiveness is substantial degree of proportional representa-
an opportunity in the sense that people are tion are run against frequency of black protest,
more likely to get what they want in responsive the increase in variance explained by the multi-
political systems than in unresponsive ones. ple correlation coefficient amounts to less than
Protest is more likely to flourish in relatively four per cent over that explained solely by
open systems where it elicits responses. mayoral government.
If the paradox of protest is that it appears to The data presented here provide a strong ba-
occur in systems exhibiting both open and sis for rejecting the hypothesis that protest is
closed characteristics, then it is possible to ar- associated with closed structural characteristics
gue theoretically that the incidence of protest is and a moderately suggestive basis for speculat-
related to the openness of the structure of oppor- ing on a theory of protest. The conditions
tunities in a curvilinear fashion. which give rise to protest are many and com-
Consider the scattergram presented in Figure plex, and the nature of the structure of political
1. Forms of government combining the nature opportunities, insofar as this is measurable by
of the chief executive, aldermanic electoral aggregate indicators, plays only one small part.
method, and partisanship are ranged on the left Yet it would appear that the incidence of pro-
from those which offer the greatest chance for test does vary with the nature of that structure
access and representation for minority and out- in a curvilinear fashion.
groups to those which offer the least. The distri- To conclude, protest seems to be an activity
bution of cities by frequency of protest appears which marks the political life of contemporary
to follow a curvilinear pattern. The most fre- American cities at a stage when they are be-
quent occurrence of protest takes place in cities coming more, not less, responsive to minority
with mixed open and closed characteristics. demands. Protest is a sign that the opportunity
The cities with the most incidents have mayors structure is flexible and vulnerable to the politi-
but they also have at-large electoral systems. cal assaults of excluded groups. As such, pro-
This finding is borne out by the partial corre- test signifies changes not only among previously
lation coefficients obtained when each of three quiescent or conventionally oriented groups but
governmental form variables is run against fre- also in the political system itself.
quency of protest controlling for the other two.
While protest is positively related to mayoral APPENDIX
government, controlling for aldermanic
City Sample with Number of Protests
elections and partisanship, it is negatively re-
lated to ward systems of aldermanic election, Anaheim 4 Miami 2
Atlanta 1 Mobile 3
when the other two variables are controlled.
Boston 9 Newark 4
Both mayoral government and ward election
Bridgeport 1 New Orleans 3
are open characteristics.
Charlotte 1 Newport News 0
Protest is neither a viable nor a fruitful strat- Cincinnati 3 Norfolk 1
egy in extremely closed systems. Not only is Columbus 1 Oakland S
protest likely to be an inadequate tactic for en- Dallas 2 Oklahoma City 3
hancing political opportunities in a closed sys- Denver 3 Phoenix 0
tem but it is not likely to be tolerated. Protest Des Moines 2 Raleigh 2
will not flourish where its use finds neither tol- Elizabeth 4 Richmond 1
Evansville 2 Roanoke 2
erance nor elicits favorable responses.
Fort Lauderdale 0 Rockford 1
In a highly open system, on the other hand,
Fresno 3 San Francisco 7
where government is not only responsive but Gary 5 Sacramento 6
anticipates needs and meets them, if any such Grand Rapids 2 Santa Ana 1
urban system exists in America, protest will be Las Vegas 2 Seattle 6
unnecessary. In an open system, groups have Little Rock 1 South Bend 5
easy access to decision makers without resort to Long Beach 4 Spokane 1
Milwaukee 11 Tuscon 1
65 Eisinger, "Protest Behavior and the Integration Minneapolis
of 5 Winston-Salem 0
Urban Political Systems," p. 990. Madison 0

This content downloaded from 193.198.215.72 on Mon, 13 Mar 2017 10:11:08 UTC
All use subject to http://about.jstor.org/terms