You are on page 1of 11

r

Nondecisions and Power:


The Two Faces of Bachrach and Baratz
GEOFFREY DEBNAM
University of Otago, New Zealand

Since Bachrach and Baratz published their two community.5 The most authoritative voice in the
articles, Two Faces of Power,1 and Decisions and debate thus far has been that of Robert Dahl,
Nondecisions: An Analytical Framework? the con- whose decision-making, or issue analysis, method
cept "nondecision" has become increasingly fa- of studying community power involves isolating
miliar,3 and now occupies a prominent position in selected issue areas and showing how crucial de-
the hagiology of community power. It was put cisions are reached within those areas. Since this
forward as a means of filling what the authors saw approach attempts to reconstruct the actual be-
as a gap in the conceptual armory of community havior of participants in various important areas
studies that had resulted in a misrepresentation of of community life, it is held to provide a reliable
American local politics. This paper looks very and representative body of data about what actu-
briefly at the limitations Bachrach and Baratz ally goes on in the community. The pluralist con-
noted in the work of Robert Dahl,4 and then at clusions drawn from this, that power tends to be
the two forms of nondecision making their solu- dispersed among a variety of issue-oriented elites,
tion appears to take in. The conclusion reached are therefore held to be closer to the reality of
here is that, in its present form, the concept has no community politics in America than are the find-
empirical value even if it may legitimately be re- ings of Floyd Hunter,6 whose suggestion that
tained to point up a weakness in the literature. Atlanta was dominated politically by a business
But it is only part of their solution. Although elite was the spark which set the whole debate
"nondecision" is held to identify the gap, it is not going. It was argued that Hunter's elitist conclu-
advanced as explanation of it. In their view, this sions could have been predicted from his reputa-
has occurred because observers have not differ- tional methodology.7 He had, or so he thought,
entiated among the various means of resolving located the "real" holders of power by the simple
relationships, of which power is only one form. and inexpensive device of asking strategically
This paper examines their proposals and con- located citizens for their opinion of who was
cludes with a more limited view of the significance powerful, thus confusing hearsay with fact.
of power for empirical research.
How far this apparent conflict is a function of
methodology, and how far it reflects real differ-
Dahl's Approach to Community Politics ences in the communities studied, is another
8
The debate over community power has been al- matter. It is referred to here simply to put Dahl's
most exclusively an American affair. The domi- 5
nant theme, understandably enough, is the at- See, for example, the article by Nelson W. Polsby
on the "Study of Community Power," International
tempt to come to terms, in a systematic way, with Encyclopaedia of the Social Sciences (New York:
the nature of political reality in the American Macmillan and The Free Press, 1968) III, 157-163,
which he opens with the statement that "contemporary
1 research on community power is distinguished by: (1)
American Political Science Review, 56 (December,
1962), 947-952. a concern with characterizing as a whole the political
2 order of an entire community (generally an American
American Political Science Review, 57 (September,
1963), 632-642. local community)" [p. 157, emphasis added].
3 6
See for example Matthew A. Crenson, The Un- Floyd Hunter, Community Power Structure.
Politics of Air Pollution: A Study of Non-Decision- (Chapel Hill: University of North Carolina Press,
making in the Cities (Baltimore: The Johns Hopkins 1953).
7
Press, 1971); Frederick Frey, "Comment: Issues and Critics of Hunter are legion—see for example Nel-
Nonissues in the Study of Community Power," Ameri- son Polsby, Community Power and Political Theory
can Political Science Review, 55 (December, 1971), (New Haven: Yale University Press, 1963); and
1081-1101; K. Newton, "Democracy, Community Arnold Rose, The Power Structure: Political Process
Power and Non-decisionmaking," Political Studies, 20 in American Society (New York: Oxford University
(December, 1972), 484-547. For a dissenting view see Press, 1967).
Raymond Wolfinger, "Non-decisions and the study of 8
Rose, pp. 296-297 notes evidence that the size of a
Local Politics," American Political Science Review, community, and its interrelationships with the national
55 (December, 1971), 1063-1080; and see also Wol- economy, may significantly influence the nature of its
finger's "Rejoinder" in the same volume, pp. 1102- power structure. This is considered in more detail by
1104. John Walton, "The Vertical Axis of Community Or-
4
Robert Dahl Who Governs? Democracy and Power ganization and the Structure of Power," (Southwest-
in an American Community (New Haven: Yale Uni- ern) Social Science Quarterly, 48 (1967), 353-368,
versity Press, 1961). which has been most recently reprinted in Com-

889
890 The American Political Science Review Vol. 69

work in context. Hunter's approach devalued the likely to be small; if he sees an adverse effect, he will
significance of the political process. In a putatively avoid the issue if he can. . . . Politicians may not see
democratic context this was clearly a challenge. how they can gain by taking a position on an issue;
action by government may seem wholly inappropriate;
Dahl concluded his study of New Haven by argu- policies intended to cope with dissatisfaction may be
ing that blocked; solutions may be improperly designed;
neither the prevailing consensus, the [democratic] indeed, politicians may even find it politically profit-
creed, nor even the political system itself are immu- able to maintain a shaky coalition by keeping tension
table products of democratic ideas, beliefs, and insti- and discontent alive and deflecting attention to irrele-
tutions inherited from the past. For better or worse, vant "solutions" or alternative issues.13
they are always open, in some measure, to alteration
through those complex processes of symbiosis and This sort of maneuvering is quite legitimate within
change that constitute the relations of citizens and a democratic context. There will be a variety of
leaders in a pluralist democracy.9 issues on which the views of constituents are con-
fused, or conflicting. Where there is no clear lead,
The central feature of the democratic political or support, from the public, the politician must
community for Dahl, then, is that the relationship act on the basis of his own calculations.
between leaders and citizens guarantees the con- Yet despite the prevalence of inertia among the
tinuing possibility of change because it is the ful- "apolitical stratum," and the opportunities for
crum of power. Any attempt to describe and independent action amongst the leaders, the sig-
analyze must look at this relationship within a nificance of the relationship can continue to be
context where both sides are motivated to act and, assumed, Dahl argues. The majority of citizens
therefore, display the dynamics of the relation- may be politically inactive, but, "although their
ship. This can be achieved by observing the de- influence is low, their indirect collective influence
cision-making process at work. There are two is high."14 And the politicians' actions and beliefs
difficulties which Dahl sees, at first sight, as re- "are all constrained by the wide adherence to the
ducing the significance of this relationship. And, [democratic] creed that exists throughout the
of course, if its significance can be questioned, it community."15 Dahl plainly believes that this re-
would diminish its relevance to any description of straint is significant because "to reject the demo-
community power. cratic creed is in effect to refuse to be an Ameri-
In the first place, citizens are largely apolitical, can."16 So the creed inhibits the politician and
"strongly influenced by inertia, habit, unexamined causes him to attend to the wishes of the politically
loyalties, personal attachment, emotions, transient inert citizen, who is thus endowed with indirect
impulses."10 A significant political relationship is influence.
difficult to envisage where one of the presumed Dahl does not say explicitly that New Haven is
parties to it turns out to be so plainly lacking in America in microcosm. He acknowledges that his
interest. Dahl notes that the citizen will engage in "data on New Haven are not wholly adequate for
political activity when "the actions or inac- the task at hand," which is a general discussion of
tions of government may threaten . . . (his) pri- "Stability, Change and the Democratic Creed."17
mary goals."11 But even though he examined three He nevertheless asserts that "New Haven will pro-
issue areas favorable to such activity in his study vide a convenient reference point" without show-
of New Haven, he found few signs of it. He is ing in what way the data are inadequate. Since he
quite justified, it would appear, in asserting that does not question the relevance of his findings to
"Homo civicus is not, by nature, a political the wider context, the implication is that they
animal."12 may be taken as a description of it. Indeed, the
The second factor which may reduce the sig- innuendo in the statement "the data are not wholly
nificance of a leader-citizen relationship, says adequate" (emphasis added) surely amounts to an
Dahl, is the relative freedom of action available assertion that, in large measure, they are. Al-
to the politician because of the complexity of the though many observers felt that his conclusions
political process and the variety of interests to be were justified, others like Parenti,18 Anton,19
considered. Thus the calculations of the politician 13
Ibid., p. 93.
may be critical. "Ibid., p. 233.
15
If a party politician sees no pay-off, his interest is Ibid., p. 325.
"Ibid., p. 317.
"Ibid., pp. 311-325.
munity Politics: A Behavioral Approach, ed. Charles 18
Michael Parenti, "Power and Pluralism: A View
M. Bonjean, Terry Clark and Robert Lineberry (New from the Bottom," Journal of Politics, 32 (August,
York: The Free Press, 1971), pp. 188-97. 1970), 501-530.
' Dahl, Who Governs? p. 325.
10
10
Thomas Anton, "Power, Pluralism, and Local Poli-
Ibid., pp. 90-91. tics," Administrative Science Quarterly, 7 (March,
"Ibid., p. 225. 1963), 425-457; "Rejoinder" Admin. Science Quar-
12
Ibid., p. 225. terly, 8 (September, 1963), 257-268.
1975 The Two Faces of Bachrach and Baratz 891
Bachrach and Baratz,20 and Crenson,21 believe tures of political life—covert control and the
that they are deceptively optimistic, particularly mobilization of bias.
when they are generalized from the local evidence In point of fact, they do not treat these as sepa-
to American society at large, where, it is clear, the rate forms. The distinction can be inferred from
possibility of change is less significant for many their writing, but it is nowhere made explicit. The
than their total inability to effect it. result is that such methodological proposals as
Dahl argues, then, that statements about power they make are uninformed by what should have
in a community should be based on direct ob- been fundamental to their case. The general thrust
servation, and not on hearsay or supposition. But of their writing is that covert control and mo-
since the leader-citizen relationship is regarded as bilization of bias, however they intend these to be
the linchpin of pluralist democracy, and the core jointly or separately interpreted, may lead to what
of community politics, and since there is no sig- they call a "nondecision." They define this as "a
nificant and observable interaction between its decision that results in suppression or thwarting
two elements, the relationship has to be rescued of a latent or manifest challenge to the values or
by two nonobservables; the leaders' commitment interests of the decision maker."22 This, they
to the democratic creed, and the citizens' indirect argue, is what an issue analysis approach inevita-
influence on the leaders. Once these are accepted bly overlooks through focusing on "concrete de-
as valid, however, they provide the complete cisions."23 By these one must assume they mean
justification for focusing on the actions of leaders formal, authoritative decisions arrived at by a
in pursuit of various objectives in various conten- constitutionally recognized organization or insti-
tious issue areas, and for ignoring the apparently tution.24 They believe that such decisions reflect
settled aspects of community life as well as the only one face of power.
aspirations of the apolitical. The viability, and The question that needs to be asked, then, is
value, of the leader-citizen relationship is demon- whether the nondecision-making concept is a use-
strated only in those areas determined by public ful addition to the power study vocabulary. It is
action. Short of public demonstrations to the con- also worth considering whether the criticisms
trary, Dahl can assume that what the leaders do prompting Bachrach and Baratz's disquisition are
is more or less what the citizens want. In these properly attributable to a decision-making ap-
terms his restricted focus is justified. proach, or whether they are the consequence of
Dahl's own partial interests.
Nondecisions
The Response of Bachrach and Baratz. Bachrach Nondecisions by Covert Control. The first objec-
and Baratz believe that the significance of the re- tion that Bachrach and Baratz make, then, is that
lationship between leader and citizen may be con- the "symbiotic" relationship between citizens and
siderably reduced for either of two reasons. First, leaders, upon which Dahl places so much empha-
when politicians are able to operate out of the sis, may turn out to be one in which the leaders,
public eye, they may not feel themselves bound by or some hidden elite which in turn controls them,
any abstract democratic creed, or may so interpret exercise a form of covert control over the whole
it that they are freely able to ignore the interests process. The leaders may not need to take any
of any group of constituents, and that these in- account of citizen interests. They may be able to
stances may be more significant than politicians' 23
formal public actions. Alternatively, the most im- NoteBachrach and Baratz, Power and Poverty, p. 44.
that this definition of nondecisions is more re-
portant fact may be neither the full flowering of strictive than what is implied by mobilization of bias
the leader-citizen relationship, nor the covert since it denies the possibility of such bias operating
machinations of an elite, but the imperatives of against "the values or interests of the decision makers."
the structural setting. Some bias that is inherent Clearly the mobilization of bias may not itself discrimi-
nate how it operates, and against whom—as is of the
in the political process itself, or the political cul- essence of covert control. The failure to distinguish
ture, may be more significant than the strivings of between these two forms of nondecision making cre-
either leaders or citizens. Bachrach and Baratz proach,ates the confusion which lies at the heart of their ap-
and is the justification for Merelman describ-
argue, then, that Dahl misses two important fea- ing them as "neo-elitists." Richard Merelman, "On
the Neo-Elitist Critique of Community Power," Ameri-
20
Both the Bachrach and Baratz articles (see notes can Political Science Review, 62 (June, 1968), 451-
1 and 2 above) have been reprinted, together with a 460.
23
sketchy report of an associated empirical study, in Bachrach and Baratz, p. 9.
24
their Power and Poverty: Theory and Practice (New They would appear to mean the same class of
York: Oxford University Press, 1970). All further events as Polsby when he refers to "specific outcomes,"
Bachrach and Baratz references will be to this book. and cites examples such as party nomination, urban
21
See Crenson, especially pp. 181-182 for a brief, development programme, and public education. See
but very penetrating analysis of the logical difficulties Polsby, Community Power and Political Theory, pp.
posed by the concept of indirect influence. 113-114.
892 The American Political Science Review Vol. 69

get their way by behind the scenes manipulation, Bachrach and Baratz argue, however, that "the
thus provoking no hostile citizen reaction. The [pluralist] model provides no objective criteria for
effect of this is that distinguishing between 'important' and 'unimpor-
demands for change in the existing allocation of tant' issues arising in the political arena."29 Yet
benefits and privileges in the community can b>2 suf- they do not establish objective criteria either for
focated before they are even voiced; or kept covert; selecting nondecisions or for identifying them.
or killed before they gain access to the relevant deci- Nor do they follow the pluralists' modest precau-
sion-making arena; or failing all these things, maimed tion of studying a selection of issue areas.30 In this
or destroyed in the decision-implementing stage of the respect their scheme offers no advantages over the
policy process.26
one they criticize.
Merelman dismisses this argument on the For example, it is quite clear that this first form
grounds that it presumes an elite in advance of of nondecision making, where the barrier to cer-
any evidence that one exists. Neither misfortune, tain forms of action is consciously, but covertly,
frustration, nor inequality can necessarily be erected by the "status quo defender," is likely to be
taken as the work of a hidden hand—which is an very difficult to distinguish from the calculations
implication of Bachrach and Baratz's approach. of politicians in confused and complex situations,
Certainly the violence of their description can be and leading to inaction, such as Dahl referred to.
dismissed as merely tendentious. After all, the Methodologically the two classes of events make
"suffocating, hiding, killing, maiming and de- the same demands. In both cases one is concerned
stroying" could equally well be directed toward with explaining something which does not neces-
demands for stability in the existing allocation of sarily have a public face. In fact, "deflecting atten-
benefits, where action is necessary to adjust to tion to irrelevant 'solutions' or alternative is-
changing conditions. Or, as Merelman argues, sues,"31 as Dahl puts it, is precisely, although not
"the neo-elitists formulation of non-decision- solely, what Bachrach and Baratz mean by non-
making . . . emphasizes only those values which decision making. They argue, though, that Dahl
contribute to elite control and support the status is unable to deal with this because his methodol-
quo. But are there no 'dominant values,' no 'ac- ogy (and, by implication, all decision-making
cepted rules of the game' which favor the initiators analysis) requires him to focus on an "examina-
of issues ?"26 tion of a series of concrete cases where key de-
cisions are made."32 But the question is whether
The question can, of course, be answered only this limitation is a function of decision-making
by locating those instances where this form of analysis generally, or whether it is a peculiarity of
nondecision making actually occurs. And if the the line taken by Dahl. If it is the latter, then the
purpose of all this is a study of community power, answer to their problem may lie, simply, in a
it would require a reasonably comprehensive sensible application of decision-making analysis
statement covering a communitywide range of techniques—in which case they may have been
nondecisions. Without this one can make no as- indulging in a case of special pleading on behalf
sertion about the community significance of any of a useless neologism.
singular instance, unless one is prepared to sug-
gest that certain nondecisions may be identified In examining politics in New Haven, Dahl was
and selected for study because they are in some chiefly concerned with asking, "How are impor-
way representative.27 The issue analysis approach tant political decisions actually made?"33 and he
to the study of community power faces the same focused on "important decisions requiring the
problem, and resolves it by selecting what are
believed to be important issues. Polsby justifies
such a preoccupation with that seems to follow from
this by asking the rhetorical question," What sort the belief that the sharing of power is a crucial index of
of power elite asserts itself in relatively trivial democracy. (On this point, see L. J. Sharpe, "Ameri-
matters, but is inactive or ineffective in the most can Democracy Reconsidered: Part II and Conclu-
significant areas of community policy-making?"28 sions," British Journal of Political Science, 3 (April,
1973), 129-167, at p. 135). Such an approach tends to
25 ignore the style and content of politics. One should
Bachrach and Baratz, p. 44. not only consider who governs, but how, and in rela-
20
Merelman, p. 459. tion to what.
27
See Roy Forward, "Issue Analysis in Community 29
Bachrach and Baratz, p. 6.
Power Studies," Australian Journal of Politics and
History, 15 (December, 1969), 26-44, for an exhaus- ™ Polsby, p. 113. "More than a single issue area is
tive analysis of the varieties of issue that may occur always chosen . . . because of the presumption among
in a community. But drawing attention to variety is pluralist researchers that the same pattern of decision-
not in itself enough, since the profusion of alternatives making is highly unlikely to reproduce itself in more
simply defeats the objective of focused analysis. than one issue area."
31
28
Polsby, Community Power and Political Theory, Dahl, Who Governs, p. 93.
32
p. 114. This question makes sense only if one is con- Bachrach and Baratz, p. 9.
33
cerned with power comparability. A typical reason for Dahl, p. 7.
1975 The Two Faces of Bachrach and Baratz 893
formal assent of local government officials."34 In Baratz's concern in examining "political interac-
an Appendix he notes that the "three issue areas tion between the black poor in Baltimore and the
were chosen because they promised to cut across white-dominated 'Establishment',"40 and which
a wide variety of interests and participants . . . they sought to explain by using the concept "non-
and in each of these issue areas all the decisions decision making." But, if an observer wants to
that the participants regarded as the most impor- account for some problem in a community, he
tant since about 1950 were selected for detailed may do so either by reference to decisions taken,
study."35 Now it is clear that these research pro- or to some process which is the embodiment of
cedures do not necessarily follow from the initial past decisions, or to the fortuitous working of
question. While there is no denying the relevance circumstances. Our understanding is not ad-
of the issues selected, a definition of "important vanced by putting forward "nondecision mak-
political decisions" does not require that a de- ing," as if another category of possibilities existed.
cision not to act, or a behind the scenes maneuver, In this sense, then, the term "nondecision mak-
or supportive, as distinct from initiatory, be- ing" is superfluous.
havior, be ignored. That Dahl does ignore such
aspects of the political scene is incontestable.36 Nondecisions through Mobilization of Bias. Bach-
But this oversight was not determined by the gen- rach and Baratz's first form of nondecision mak-
eral questions with which he prefaces his inquiry. ing, then, adds nothing to Dahl's general discus-
The reason is not hard to suggest. Dahl was, sion of power. They are correct to point out that
perhaps, more concerned with producing a rea- his empirical work fails to exemplify the range of
soned response to the views of Hunter, who his discussion, but this applies with equal force to
pictured an American community as controlled their own efforts in this field. They add neither
by big business, than with establishing a general theoretical nor methodological perspective and
picture of community politics. "The question is," are wrong to assume that any limitation in Dahl's
he wrote two years after the publication of Who work is justification for a new approach.
Governs?, "whether businessmen or related groups, But it seems that they go further in their criti-
social strata or classes, also dominate the de- cisms and argue that the relationship is also dis-
cisions of public government... if not, what torted by the mobilization of bias. Both leaders
groups or strata have the most influence over the and citizens may be affected by this because, as
decisions of public governments."37 He justified they point out by quoting Schattschneider, "all
this particular focus by arguing that it was "the forms of political organization have a bias in
arena in which the controversy over 'community' favor of the exploitation of some kinds of conflict
power arises."38 But is community power simply a and the suppression of others because organiza-
function of group conflict over governmentally tion is the mobilization of bias. Some issues are
controlled decisions? This is as much a matter for organized into politics while others are organized
investigation as is the question of who governs. out."41 Yet their research strategy does not sug-
In a critique of Dahl's study Anton complains, gest any means of examining or assessing the
"Surely the study of power must involve more significance of a mobilization of bias. In fact,
than the actions of government agencies."39 What- there is some confusion in their thinking here. It
ever one may believe to be the case, the answer is easy to follow through their reference to killing,
needs demonstrating rather than merely asserting. maiming, suffocating, etc., by whatever means are
There is no reason, for example, why the de- available to the "status quo defenders."42 This, as
cision-making analyst should not decide that the noted above, can be put under the general heading
most important decision in community X, re- of "covert control." Since this is rational, pur-
putedly democratic, is that which has resulted in, posive, calculation it need not be connected in any
and continues to maintain, the exclusion of forty way with the mobilization of bias. It is not so
per cent of the citizens from any effective part in easy, though, to see the value of their reference to
the political process. This was Bachrach and the mobilization of bias except as a general
legitimation of their claim that there are more
"Ibid., p. 102. things in their political universe than are dreamt
35
Ibid., p. 333. of in Dahl's philosophy.
"See Anton, "Rejoinder," p. 265; and also Peter
Morriss, "Power in New Haven: A Reassessment of From what they have written it appears that
'Who Governs?'," British Journal of Political Science, the mobilization of bias is itself to be considered
2 (October, 1972), 457-465, at pp. 459-460. a form of nondecision making as well as, and
37
Robert Dahl, "Reply to Anton's Power, Pluralism 40
and Local Politics," Administrative Science Quarterly, Bachrach and Baratz, p. 53.
41
7 (September, 1963), 250-256, at p. 254. E. E. Schattsschneider, The Semi-Sovereign People
88
Dahl, "Reply to Anton," p. 254. (New York: Holt, Rinehart and Winston, Inc., 1960),
30
Anton, "Power, Pluralism and Local Politics," p. p. 71; cited in Bachrach and Baratz, p. 8.
42
453. Bachrach and Baratz, p. 43.
894 The American Political Science Review Vol. 69

separately from, the actions of individuals in sup- Identification of Nondecisions. Bachrach and
port of, or modifying it. For in discussing "such Baratz have argued, then, that Dahl has failed to
bias" they write, "To the extent that a person or establish sufficiently comprehensive criteria for
group—consciously or unconsciously—creates or selecting important decisions, and that this over-
reinforces barriers to the public airing of conflicts, sight leads him to ignore important aspects of
that person or group has power."43 If someone community politics. Crtnson, in his study The
does unconsciously create or reinforce a barrier, Un-Politics of Air Pollution: A Study of Non-
however, it is unrealistic to say that they are, for Decisionmaking in the Cities, in which he first sets
that reason, powerful. My style of speech or man- out a detailed case study of the differing responses
nerisms may, for example, inhibit someone else's of two neighboring American communities to al-
actions or opinions without my wishing this to be most identical air pollution problems, and then
so. In fact it may be my wish that they should not uses some very slender data from fifty-one cities
be inhibited. If that is so it would seem rather to test the propositions thrown up by the case
peculiar to say that, by unconsciously accomplish- studies, agrees with this argument. He believes
ing what I would not have wanted to happen, I that the "power to restrict the scope of the politi-
have therefore demonstrated my power. Where cal process . . . is not revealed by the investigation
my actions do have such an effect, whether I sub- of political activities in key issue areas because
sequently approve of the outcome or not, then the issues in which this kind of power is likely to
presumably it is because it is customary for some become significant are precisely the ones that
to respond in that sort of way to that sort of never become 'key'."45 Bi_t "which non-events are
stimulus; in which case, it is more proper to to be regarded as significant?," Polsby wants to
impute power to the customary basis of behavior
rather than to the behavior itself. In other words, know.46 Bachrach and Baratz say go and ask the
the mobilization of bias may itself be recognized locals.47 Crenson gets round the problem by ar-
as the effective agent. It is as tangible a fact of life guing that significant nonissues are—well, signifi-
as the overt actions of some who may exploit it cant in the way that "the rre /ention of elephant
for their own benefit. The question is how one stampedes or the persecution of witches"48 in con-
makes it the subject of a research exercise. temporary America is not. In other words, the
researcher should use his common sense. There is
Its existence, and its significance, can be de- everything to be said for this, but rather than
termined only by reference to some datum point pointing up the need for a new concept, it merely
since not only is organization the mobilization of stresses the cruciality of defining "key issue" in a
bias, but so, in a sense, is civilization itself. The way that is not restrictive.
observer must attack the problem from a perspec- Crenson attempts to establish the nature of the
tive which he determines himself, or which problem more firmly by defining nonissues as
emerges from the community he chooses to study. topics which have not been included on a "com-
He may thus seek to analyze community interac- munity's political agenda."49 These can be identi-
tion in terms of some notion of justice or equality, fied because "only when advocacy or opposition
or in terms of the pursuit of a neglected issue he has made an appearance arrong community lead-
nominates as of some significance. In addition he ers will we say that a topic \ a t .tken its p!ace on
may analyze unsettled grievances held by citizens, the local agenia."50 "Community political sys-
perhaps established by a sample survey. In neither tems," he writes, "may be ' m penetrable' where
case is anything proposed here which has not al- certain issues are concerned."61 In other words,
ready been touched on by Bachrach and Baratz, nondecision making preven s bsues crossing the
and Crenson. Either way the observer has to focus boundary from the nonpolitical to the political
on specific aspects which are believed, on what- system.52 Since Crenson's definition throws an em-
ever grounds, to be a manifestation of bias.
Now asking the question, "Why did this, or "Crenson, p. 21.
46
that, not happen?" makes the same methodologi- Po!sby, p. 97.
" Bachrach and Baratz, pp. 47-51.
cal demands as its opposite. There is, of course, 43
Crenson, p. 26.
the possibility that the mobilization of bias may "Ibid., p. 29.
m
be so pervasive, so diffuse and insidious that it 51
Ibid., p. 30.
simply cannot be captured in action, rule, or Ibid., p. 17.
52
The notion of "boundary" is descriptively useful so
speech. But this resurrects, in a different form, the long as one need not be precise about where it actually
"principle of infinite regress" and may be safely is. It is useful, that is, as a means of drawing attention
left to the metaphysicians.44 to gross distinctions between general classes of variables.
Nondecision making, however, requires that the pre-
"Ibid., p. 8. cise boundary between political and nonpolitical be
"See Robert Dahl, "Critique of the Ruling Elite determined for it is only at this point, in Crenson's
Model," American Political Science Review, 52 (June, terms, that it can operate. For the difficulties of
1958), 463-469; and Polsby, Community Power and boundary definition see Samuel E. Finer, "Almond's
Political Theory, p. 34. Concept of 'The Political System': A Textual Critique,"
1975 The Two Faces of Bachrach and Baratz 895
barrassingly large area of public discussion into proach. But that is, of course, merely a method of
the "nonissue" category, we obviously need rather inquiry. In examining power in the community55
more selective criteria to determine what topics the kinds of data collected should be determined
are kept off the "political agenda" through non- by our understanding of the concept "power" and
decision making. After all, every issue, whether not by the methodology employed.
eventually successful or not, must go through One can at least say of the reputational ap-
some perinatal obscurity. We have no means of proach to the study of corrmun'.ty power that it
determining that the alleged nondecision has been starts from a fairly clear idea of both those linked
kept out and is not, in fact, going through a concepts. Power, for examp'e, i lvolves "the acts
lengthy process of legitimation. Drawing the po- of men going about the busi; ess of moving other
litical boundary in one place rather than another men to act in relation to tht mse ves or in relation
has the effect of creating two classes of event to organic or inorganic tlings.' 5e And commu-
where there may be only one.53 nity is the complex of socioeconcnic relationships
If we look more closely at Crenson's argument within which such a faculty is sustained. Polsby
that nondecision making power can be significant correctly points out that such a view presupposes
only where issues do not become "key," we can that "power is a subsidiary aspect of the com-
recognize the weakness of the nondecision case in munity's social structure."57 It is an attempt,
its present form. Since Bachrach and Baratz, and nevertheless, to relate political activity to its com-
Crenson in his case studies, have all focused on munity environment. "The first and most basic
what clearly have become key issues in their com- presupposition of the pluralist approach," writes
munities, it is plain that, in the terms of Crenson's Polsby, however, "is that nothing categorical can
definition, none of them can have been examining be assumed about power in any community."58
significant nondecision-making power. Both issues Dahl assumes, instead, the priority of rational
were, as Polsby would argue, eminently research- self-interest within the vaguely limiting frame-
able by decision-making analysis techniques.54 If work of the democratic creed. His approach to
Crenson wishes to establish the lexical convention power, for example, is that "A has power over
B to the extent that he can get B to do something
of calling neglected, or nascent, topics "non- that B would not otherwise do."59 The effect of
issues," there can be no objection. What is objec- this approach, since it focuses on the act of indi-
tionable is the claim that this provides a new and vidual choice, is to imply that such choice is the
empirically viable form of political analysis. most significant aspect of the polity. The com-
Power munity exists only as "an aggregation of indi-
viduals."60 As much may be said of Crenson who
The concept "nondecision" does not provide accepts the pluralist view of power61 and ignores
any new means of identifying and studying the less community entirely. Bachrach and Baratz have
apparent features of community life that Dahl has not sought to come to terms with the community
been accused of ignoring. It offers nothing that is
not already provided by a decision-making ap- M
For sociologists, community has been a key con-
cept, although Colin Bell and Howard Newby, Com-
Government and Opposition, 5 (Winter, 1969-70), 3- munity Studies (London: Allen and Unwin, 1971), point
21. to a current disenchantment with the term (pp. 48-53).
83
It is, of course, a perfectly legitimate undertaking Political scientists have not, however, tended to regard
to inquire why a community did not do this or that. this as a problem requiring much attention. Polsby,
If it can be shown that inaction was intentional, then "Study of Community Power," p. 157, notes that com-
it concerns community power. But the result would be munity power researchers have adopted a conventionl
a community study only in the most limited sense. perspective by defining community "as a population
Crenson's critique of the pluralist position, particularly living within legally established city limits." This view
of the muddy notion of indirect influence, is extremely was also taken by L. T. Hobhouse, Social Develop-
well argued. And he displays considerable ingenuity ment (London: Allen and Unwin, 1924) who regarded
in his study of factors influencing the pollution issue "all populations living under a common rule as political
in fifty-one cities. But since these cities are reduced communities, though they have only the bare bones of
to little more than statistical artifacts, he entirely dis- a common life" (pp. 41-42). Since "community implies
poses of the community baby with the pluralist bath having something in common" (Ronald Frankenberg,
water. Of course, in one sense, his is not a community Communities in Britain: Social Life in Town and
study. But it is so clearly advanced as part of a con- Country [London: Penguin, 1966], p. 238), it seems
tinuing debate on community power, seeking to vali- reasonable to take the common element as the defining
date the current focal concept of that debate, that it characteristic.
must be judged in terms of its contribution in that m
Hunter, Community Power Structure, pp. 2-3.
direction, and not simply on the more limited terms "Polsby, p. 7.
of its subtitle, i.e., "Non-Decisionmaking in the Cities." "Ibid., p. 113.
54 m
"Insofar as . . . goals are in some way explicitly Robert Dahl, "The Concept of Power," Behavioral
pursued by people in the community, the method of Science, 2 (July, 1957), 201-215, at pp. 202-203.
M
study used in New Haven has a reasonable chance of Anton, "Power, Pluralism and Local Politics," p.
capturing them." Polsby, Community Power and 445.
Political Theory, p. 97. " Crenson, p. 34.
896 The American Political Science Review Vol. 69

concept. But they have made an elaborate attempt existence or when only the less powerful party is
to resolve the power problem, an attempt which aware of it."64 But, having made the general state-
falls between the two extremes offered by Hunter ment, Bachrach and Baratz do not satisfactorily
and Dahl. It avoids putting forward definitions at answer the second question that needs asking.
the community level of generality, but goes be- How does one then identify conflict ? Instead, in
yond Dahl's focus on the act of power by insisting a section misleadingly entitled "Empirical Identi-
that there are a variety of types of relationship fication of Nondecisions," they shift the question
which shape the character of the community. to "what persons or groups in the community are
They argue that the failure to recognize this va- especially disfavored under the existing distribu-
riety has caused observers to overlook what is tion of benefits and privileges?"65 This requires
implicit in their general notion of nondecision an entirely different set of questions being asked
making. about the scope and purposes of government.
Thus they leave the matter unresolved. The final
The Bachrach and Baratz Proposals. In examining question to be put is whether conflict is a necessary
the two faces of power, Bachrach and Baratz ar- indicator of power. This can be best answered by
gue, one must distinguish among power, author- posing a counter question. Do Bachrach and
ity, influence, force and manipulation, because Baratz mean to say that power may not be
"while decision-making frequently does involve exerted to prevent conflict? Merely asking the
power relationships, it very often does not."62 question suggests the inadequacy of such a view,
Now, the point of interest here does not lie in the
definitions that they create to sustain these differ- therefore this first indicator of power may be
ences, but how these are to be applied in an em- summarily dismissed.
pirical study. For this purpose Bachrach and
Baratz specifically refer to several indicators, all (ii) Sanctions. Bachrach and Baratz do not make
of them positively associated with power, and it clear if they intend to place their several con-
either positively or negatively associated with the cepts on some notional "sanctions scale," ranging
other concepts. The conditions that must be pres- from the case of manipulation (to which sanctions
ent, for example, before they would agree to the are held to be not relevant), through influence and
use of the concept "power" to describe the means authority where there are "no severe sanctions,"
of arriving at some conclusion to a situation and power, where the "threat of severe sanctions"
would be as follows. applies, to the case of force which requires the
actual "application of severe sanctions."66 Else-
(1) There is a conflict of values between A where they imply that influence involves no sanc-
and.B; tions at all. For example, "the exercise of power
(2) A threatens B with severe sanctions in the depends upon potential sanctions, while the exer-
event of Z?'s noncompliance; cise of influence does not."67 In their discussion of
68
(3) A's demand, and the severity of the sanc- authority no reference is made to the subject.
tions, are rationally perceived by B; and In view of their emphasis on severity, however,
(4) A gets his way with B's compliance. it is reasonable to assume that they mean to imply
a hierarchy of sanctions. But this raises the whole
They feel impelled, however, to make explicit question of establishing an acceptable "severity
the most general feature of power from which scale" against which the interpretation of the
these more specific conditions follow. Power, they sanctions by the "patient" in the power relation-
argue, is a relational concept.63 The consequences ship can be measured. This is crucial to their de-
of drawing particular attention to relationality sign since severity of sanction is the only signifi-
will be discussed after the value of each of these cant means of distinguishing power from author-
four indicators—conflict, sanctions, rationality, ity and influence.69
and compliance—has been examined briefly. Dahl discusses this point in these terms. "Ex-
(i) Conflict. Thefirstquestion that needs asking is actly what constitutes a 'severe' loss or depriva-
whether we should be concerned only with the tion is, to be sure, somewhat arbitrary. No doubt
overt signs of conflict, the public words and ac- what a person regards as severe varies a good deal
tions that express an opposition of interests; or with his experiences, culture, bodily conditions,
whether we should also include the psychological and so on. Nevertheless, probably among all
state, the mere awareness of opposed interests. 64
Bachrach and Baratz include both in their scheme; Ibid., p. 50.
05
Ibid., p. 50.
"a power struggle exists, overtly or covertly, either '"These references are all to Bachrach and Baratz,
when both sets of contestants are aware of its p. 6737, Table 1.
02 Ibid., pp. 30-31.
3
Bachrach and Baratz, p. 42. 68
89
Ibid., pp. 32-36.
* Ibid., pp. 19-21. Ibid., p. 37.
1975 The Two Faces of Bachrach and Baratz 897
peoples exile, imprisonment and death would be cherished values and is, therefore, rational; while
considered severe punishments." 70 If such an ap- rationality in their example of influence seems to
proach, with its disarming "somewhat," "no mean no more than abject self-abasement. The
doubt," and "probably," is of any use, it certainly resolution of a situation in any of these terms is
gives no help to the student of community power. held to be rational, whereas resolution by force is
But it is the sort of limp conclusion that one u not because, presumably, the act of force does not
driven to if one insists on a necessary connection depend for its success on B's perception.
between power and severity of sanction. It also Notice that, in the case of power, authority, and
places an unjustifiable limitation on the way we influence, B is alleged to be rational because he
use the word "power." For if A is able to get his has chosen to comply. There is an opportunity
way without resorting to such sanctions, it would after the stimulus for B to behave in whichever
be decidedly anomalous to say that he was, there- way he wishes; but there must, by definition, be
fore, not powerful.71 some limitation on this freedom implicit in the
initial stimulus. The rationality of subsequent be-
(iii) Rational Perception. Bachrach and Baratz havior is, therefore, relative to the restricted defi-
maintain that nition of the situation contained in the stimulus.
power has a rational attribute: for it to exist, the In the case of power, authority, and influence,
person threatened must comprehend the alternatives however, B has no control over the stimulus,
which face him. . . . In a situation involving power, B whereas in the case of force the stimulus is, in a
is rational in the sense that he chooses compliance in- sense, B's choice. It is B's definition of the situa-
stead of defiance because it seems the less of two evils.
In a situation involving authority, B complies because tion in as real a sense as agreement is B's choice in
he recognizes that the command is reasonable in terms the case of the other concepts. To deny rationality
of his own values; in other words, B defers to A, not in this case, while claiming it for any emotional
because he fears severe deprivations, but because his and submissive response, is highly tendentious.
decision can be rationalized.72 The purpose of this brief discussion, however,
is not only to demonstrate that Bachrach and
As far as influence is concerned (the only other Baratz use the term "rational" in a rather inade-
concept characterized by rationality), Bachrach
quate manner, but also to suggest that there is
and Baratz are less helpful. They do state ex-
little point in trying to base a definition of ra-
plicitly that "power and influence are alike in that
each has both rational and relational attributes." 73
tionality on the subjective state of the particular
But we are left to infer the nature of such rational-
actors in a specific situation, because there are,
ity from the example of "the ambitious young then, no valid grounds for distinguishing between
man who submits unhappily to the every dictate rational and irrational behavior. Yet as Frohock
of his rich uncle . . . because he admires wealthy points out, "If criteria are introduced to allow the
men (influence)."74 rational-irrational distinction, then the analytic
construct gets away from that which it is trying to
They are arguing, then, that rationality is a suffi-
explain, which is the way in which actors ascribe
ciently notable characteristic of these three con-
cepts to merit particular mention. But consider
meaning to what they do."75 To specify rationality
the significance of what they are saying. In the
as a condition of power is unnecessary, then, be-
case of a "power relationship" B is rational be- cause it raises greater problems than it may solve.
cause he is dominated by fear; in the case of
authority B enjoys the vicarious gratification of (iv) Compliance. One of the requirements that
must be borne in mind in creating definitions of
70
Robert Dahl, Modern Political Analysis, 2nd ed. abstract concepts is how far the terms of the defi-
(Englewood Cliffs, New Jersey: Prentice Hall, 1970), nition vary from conventional usage. A definition
pp. 32-33. may, of course, be as arbitrary as the special pur-
71
We continue to associate power with the ability to
inflict severe sanctions, but it is not a necessary associ- poses of the analyst require. But the more com-
ation. As Anthony de Crespigny, "Power and its monplace the concept, the less such semantic
Forms," Political Studies, 16 (June, 1968), 192-205, deviance is acceptable. Power is a commonplace.
points out, "If it is wished to make 'power' a technical It is part of this commonplace to recognize that a
term in the social sciences, it must be stripped of its
dyslogistic associations. It must be used without any man who habitually gets his way is powerful. One
limitations concerning the ways in which power may be does not delay such recognition for the testimony
said to be exercised" (p. 193). This analysis has bene- of compliant former opponents. It is, thus, a
fited considerably from de Crespigny's preliminary further limitation on power to suppose that it can
discussion of power, although his major concern, to
distinguish between various types of power, is not felt apply only to those cases where the patient is
to advance the cause of empirical political analysis.
" Bachrach and Baratz, pp. 22 and 34.
n
"Frederick Frohock, The Nature of Political In-
lbid., p. 30. quiry (Homewood, Illinois: The Dorsey Press, 1967),
M
Ibid., p. 31. p. 137.
898 The American Political Science Review Vol. 69

compliant. This would reduce success, which we banal point that the political scientist does so by
intuitively suppose to be the chief feature of choosing not to study matter but people in their
power, to a subordinate status. political relationships.
If compliance is made into a condition, then it
follows that power is a term describing a whole Intention and Effect: The Contextual Alternative.
developing sequence, spread over whatever period If these unnecessary associations are removed
of time, whose two boundary points, conflict and from the discussion of power, we are left with an
compliance, are separated by threat of severe agent and an effect, and the presumption that the
sanctions and rational perception of these. This is agent intended the effect. In this view, then, power
unwieldy, particularly since it requires that the is simply "the production of intended effects."78
power relationship ceases to exist, for lack of con- If a term is used to describe an effect, however, it
flict, the moment the patient complies with a com- cannot also be used to explain its cause. For exam-
mand. It creates the logically difficult position ple, while we use the word "bang" to describe the
that compliance both creates a power relationship effect of detonating an explosive, only a child
and terminates it. would use the same word to describe the explosive
itself. We develop the ability to distinguish be-
(v) Relationality. The foregoing discussion has tween cause and effect. There is no reason to sup-
indicated the difficulties raised by these indicators pose that power should be treated any differently.
of power. More than this there is no sign that they If it is, then one is faced with the question of what
are specifically directed toward the initial prob- creates power, and the answer would have to be
lem, which involved an attempted clarification of power. In which case one would be driven to con-
what Bachrach and Baratz describe as the "Two clude that power is an inherent property of the
Faces of Power." They have built these concepts power resource, which is an absurdity. Although
on a set of terms which describe an individual we say, for example, that money is power, the ex-
state rather than a political relationship—hence pression is clearly an ellipsis. We know quite well
the need to emphasize relationality. that some other agency must intervene before the
connection is made. This would be either an indi-
The objection to such an emphasis is, quite vidual who actively uses money as a resource, or
simply, that there is no concept in the social sci- the institutional ation of money which may re-
ences that is not relational in some sense. Socio- sult in others responding to what they believe to
political reality cannot be conceived in terms other be the wishes of the wealthy without any overt
than those which involve, pertain to, or imply, a prompting to do so by the wealthy themselves.
relationship. It is therefore, quite worthless to Money is, potentially at least, power. But power
propose relationality as a distinguishing character- potential cannot be the same as power, by defini-
istic. Bachrach and Baratz argue that manipula- tion.79 If we continue using the same word to de-
tion does not involve a relationship.76 But all they scribe both we confound understanding by ob-
can mean by this is that they believe a relationship scuring what is to be explained.
exists only when both parties are conscious of each
other, and also of the terms which are held to con- Any approach to power which ignores the dis-
stitute the relationship. Such symmetry is, how- tinction between cause and effect, then, is bound
ever, neither necessary to comply with conven- to lead to difficulties. But in conventional usage
tional usage, which would not insist on a direct we frequently describe someone as powerful out-
link being established, nor reasonable in the de- side the context of any particular situation. We
mands it makes for empirical analysis. Relational- may thus describe an agent as powerful where
ity is emphasized, of course, as a rebuttal of the known to have bean successful in the past, and
view that power is a commodity that can be pos- believed to retain the capacity for success. We
sessed by one person outside the context of any would say that he "possesses" power, or "is
relationship. But why go to such lengths to refute powerful." In his case, we would imply, power is
what is, after all, irrefutable, and equally un- a cause of success, and it is because of this that he
demonstrable ? can get his way. Yet such an imputation is justified
It is, then, impossible to regard relationality as by past actions, or by our ready assumption of
a characteristic of specific "power events" in an what, in those circumstances, ought to constitute
otherwise nonrelational political field. This latter power. We could not apply these terms to anyone
is, to steal a phrase from Bentham, "nonsense on who, although possessing all the objectively ap-
stilts." Bachrach and Baratz argue that one must propriate resources, has never 80
appeared to have
"distinguish clearly between power over people been intentionally successful.
and power over matter."77 They overlook the rather 78
Bertrand Russell, Power: A New Social Analysis
(London: Allen and Unwin, 1938), p. 35.
78
"Bachrach and Baratz, p. 37. See Rose, The Power Structure, pp. 45-53.
77 80
Ibid., p. 19. See de Crespigny, pp. 192-196.
1975 The Two Faces of Bachrach and Baratz 899
So although conventional usage refers to power tional as well as individual, must be given equal
as cause as well as effect, it does so only where status with the problem of effects—which has, so
there appears to be some known, or assumed, far, received the major emphasis.83 This implies, of
record of success. It is a convenient simplification course, that one must establish a causal relation-
where one does not know, or one does not choose ship between intention and effect before saying
to spell out, why such success has been possible. with any certainty that one has identified an in-
That general usage will be retained here. Power stance of power. While this may be an ideal posi-
may be variously referred to as being created, tion to work toward, one can never demonstrate
sought after, possessed, exercised, structured, conclusively in the social sciences that any given
eroded, fragmented or lost. Nothing more than factor is both necessary and sufficient for the pro-
the processes related to the "production of in- duction of any given effect. This difficulty indi-
tended effects" is implied. Power, then, has the cates the importance of circumstantial evidence.
same limited, but graphic, value as "bang." Rec- It is more useful to formulate an approach which
ognizing the value of a term which points up the deals with the context of power than to attempt
salient features of our concern does not mean that to capture power inflagrante delicto, although one
we should be led into believing that that term is naturally anticipates that specific conclusions may
sufficient explanation in itself. It is more impor- be drawn from such a general account.
tant to consider the context in which power may
be discussed. Conclusion. This discussion suggests that the con-
One could argue that this is little different from cept "nondecision" is of no practical value in the
the approach adopted by Dahl, which also iso- analysis of community power, because it masks
lates a reference point and attempts to encapsulate the useful distinction that can be made between
it within a wide range of explanatory data.81 How- covert control and mobilization of bias, and offers
ever, the similarity should not mask the divergent nothing which is not already available through
consequences for empirical analysis. decision-making analysis. The problem it was
Dahl's approach to power is that "A has power directed toward can be resolved only by adopting
over B to the extent that he can get B to do some- a more modest approach to the concept "power"
thing that B would not otherwise do."82 If this is which does not require the observer to focus on
adopted as a guide to research, one must first individual initiatives within the context of an ob-
identify two parties in interaction, and then estab- servable relationship. Such an unnecessary restric-
lish a change of position by either participant tion devalues the significance of the context to the
which can be attributed to the efforts of the other. description of power. It is only by capturing some-
The inevitable result is that one must focus on the thing of this that the observer can hope to under-
dynamics of a concretely verifiable relationship. stand the forces shaping the relationships that
It is not possible for such an approach to deal constitute the community. There is no shortcut to
with the problem of inaction, which is just as rele- be gained by elaborating a set of stipulative defini-
vant to the discussion of power. tions, as Bachrach and Baratz have attempted.
The result is merely to inhibit the observer's
The definition adopted here shifts the focus in- sensitivity to his surroundings. Since their work
stead to the intention to produce an effect, as well can be taken primarily as an appeal to this, it is
as the effect itself. It does not require that move- unfortunate that the "Two Faces" they have put
ment be demonstrated in the position of any other on their concern should both look inward onto a
party. This means that empirical analysis is freed definitional maze.
from a preoccupation with the intricacies of an
ascertainable relationship. It means also that the 83
A definition of "intentions" and "effects" poses
intentions of actors in the community, institu- no more problems than does a definition of "issue"
or "decision." In both cases the problem of establish-
81
1 am grateful to Alan Alexander of Reading Uni- ing a requisite level of significance arises, but can
versity for his forceful probing on this point. be dealt with only by the observer using stated cri-
82
Dahl, "Concept of Power," pp. 202-203. teria within the context of a specific study.

You might also like