Lowi - American Business, Public Policy, Case-Studies, and Political Theory

Trustees of Princeton University
Review: American Business, Public Policy, Case-Studies, and Political Theory

Reviewed Work(s): American Business and Public Policy: The Politics of Foreign Trade. by
Raymond A. Bauer, Ithiel de Sola Pool and Lewis A. Dexter
Review by: Theodore J. Lowi
Source: World Politics, Vol. 16, No. 4 (Jul., 1964), pp. 677-
693+695+697+699+701+703+705+707+709+711+713+715
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/2009452
Accessed: 07-10-2016 16:07 UTC
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
Cambridge University Press, Trustees of Princeton University are collaborating with JSTOR to
digitize, preserve and extend access to World Politics
This content downloaded from 200.18.49.10 on Fri, 07 Oct 2016 16:07:10 UTC
All use subject to http://about.jstor.org/terms
AMERICAN BUSINESS,
PUBLIC POLICY, CASE-STUDIES,
AND POLITICAL THEORY
By THEODORE J. LOWI
Raymond A. Bauer, Ithiel de Sola Pool, and Lewis A. Dexter, American

Business and Public Policy: The Politics of Foreign Trade, New York,
Atherton Press, i963, 499 pp. $8.95.
C ASE-STUDIES of the policy-making process constitute one of the

more important methods of political science analysis. Beginning
with Schattschneider, Herring, and others in the I930's, case-studies
have been conducted on a great variety of decisions. They have varied
in subject-matter and format, in scope and rigor, but they form a
distinguishable body of literature which continues to grow year by
year. The most recent addition, a book-length study by Raymond
Bauer and his associates, stands with Robert A. Dahl's prize-winning
Who Governs? (New Haven i96i) as the best yet to appear. With its
publication a new level of sophistication has been reached. The stand-
ards of research its authors have set will indeed be difficult to uphold
in the future. American Business and Public Policy is an analysis of
political relationships within the context of a single, well-defined issue-
foreign trade. It is an analysis of business attitudes, strategies, com-
munications and, through these, business relationships in politics. The
analysis makes use of the best behavioral research techniques without
losing sight of the rich context of policies, traditions, and institutions.
Thus, it does not, in Dahl's words, exchange relevance for rigor; rather
it is standing proof that the two-relevance and rigor-are not
mutually exclusive goals.
But what do all the case-studies, including American Business and
Public Policy, add up to? As a result of these case materials, how much
farther along the road of political theory are we? What questions have
the authors of these studies raised, and what non-obvious hypotheses
and generalizations about "who rules and why" would we have lacked
without them?1 Because of what it does, what it implies and what it
does not do, American Business and Public Policy provides a proper
occasion for asking these questions and for attempting once again to
1 For similar questions and a critique, see Herbert Kaufman, "The Next Step in
Case Studies," Public Administration Review, xviii (Winter 1958), 52-59.
678 WORLD POLITICS
formulate theories that will c

into elements that can be ass
what theories have we now, and how does this significant new study
relate to them ?
I. EXISTING NOnONS: THE NON-THEORIES OF POWER IN AMERICA

It was inevitable that some general notions about power and public
policy would develop out of the case-study literature. Together, these
notions form what is variously called the group theory, the pressure-
group, or the pluralist model of the democratic political system (a
model recently also applied to non-democratic systems). No theory or
approach has ever come closer to defining and unifying the field of
political science than pluralism, perhaps because it fitted so nicely both
the outlook of revered Federalist #io and the observables of the New
Deal. Group theory provided a rationale for the weakness of parties
and the electoral process. It provided an appropriate defense for the
particular programs pursued by the New Deal and successive Admin-
istrations. And, more importantly, it seemed to provide an instant
explanation, in more or less generalizable terms, of the politics of each
decision. Analysis requires simply an inventory of the group participants
and their strategies, usually in chronological form-for, after all, politics
is a process. Each group participant is a datum, and power is attributed
in terms of inferred patterns of advantage and indulgence in the final
decision. The extremists have treated government ("formal institu-
tions") as a tabula rasa, with policy as the residue of the "interplay of
forces" measurable as a "parallelogram." More sophisticated analysts
avoided the government-as-blank-key approach by treating officials as
simply other units in the group process, where Congressman and
bureaucrat were brokers but with their own interests and resources.
In group theory, all resources are treated as equivalent and inter-
changeable. And all the varieties of interaction among groups and
between groups and officials are also treated as equivalent, to such an
extent that only one term is employed for all forms of political inter-
action: the coalition. Coalitions, so the argument goes, form around
"shared attitudes" and are extended by expansion of the stakes of the
controversy. Two types of strategies comprise the dynamics of the
process: internal and external. The first refers to the problem of
cohesion in the midst of overlapping memberships; cohesion is a de-
terminant of full use of group resources. The second refers to expansion
of the coalition and the strategy of its use. Large coalitions beat small
coalitions. System equilibrium (of unquestionably high-priority value
PUBLIC POLICY AND POLITICAL THEORY 679
to pluralists) is maintained by the requirement

tions, which are extremely difficult to create bu
virtually from scratch for each issue. Thus, pow
ized, fluid, and situational. There is no single elit
system in which the centers exist in a conflict-
to each other.
As an argument that the group must be the m
pluralism excites little controversy. But controversy is unavoidable
insofar as the pluralist model implies a theory of power or power
distribution. Most importantly, the pluralist model has, until recently,
failed to take into account the general economic and political structure
within which the group process takes place.2 On this basis, the leading
type of critique of the pluralist model is a set of explicit propositions
about power structure and elites. The typical answer to pluralism is a
straightforward Marxian assumption that there is a one-for-one relation
between socioeconomic status and power over public decisions. Perhaps
the more sophisticated version is a combination Marx-Weber approach
which specifies the particular status bases most closely related to
power-i.e., the major "orders" of society (stinden) in our day are the
military, the industrial, and the political hierarchies.
This is no place to enter into an elaborate critique of either of these
approaches or of the pluralist approach itself. Suffice it to say that
while the pluralist model has failed to take the abiding, institutional
factors sufficiently into account, the "social stratification" and "power
elite" schools wrongly assume a simple relation between status and
power. Both these latter schools mistake the resources of power for
power itself, and escape analytic and empirical problems by the route
of definition.3 There is no denying, however, that the social-stratification
or power-elite approaches can explain certain important outcomes in a
more intuitively satisfactory manner than the pluralist model precisely
because each emphasizes that, while coalition-forming may be uni-
versal, not all coalitions are equivalent. For certain types of issues
(without accepting Mills's argument that these are all the "key" issues),
2David Truman's rather weak and diffuse "potential interest group" is a doff of
the hat in this direction, but this concept is so non-directive and non-observable as to
be disregarded even by its creator.
3 The best critiques and analyses of all the various currents of thought are found
in Nelson W. Polsby, Community Power and Political Theory (New Haven i963);
Daniel Bell, "The Power Elite-Revisited," American journal of Sociology, LXIV
(November 1958), 238-50; Robert A. Dahl, "Critique of the Ruling Elite Model,"
American Political Science Review, LII (June 1958), 463-69; and Raymond Wolfinger,
"Reputation and Reality in the Study of Community Power," American Sociological
Review, xxv (October i960), 636-44.
680 WORLD POLITICS
it seems clear that decisions

"officials" in virtually a pub
vacuum. One does not have to
behind all apparent conflict th
on specific major policy goal
pluralist is equally unwise w
post" positions in all orders of
the recruitment and grooming
sible a reduction in the num
equally possible (i) many stab
(2) accommodation to conflic
("through channels") than c
conflict by more informal m
debates and votes) that maint
There is still a third approach to power and policy-making, no
less important than the others, which has not been self-consciously
employed since its creation in I935 because it was mistakenly taken as
a case of pluralism. I refer to E. E. Schattschneider's conclusions in
Politics, Pressures and the Tariff (New York I935). Schattschneider
observed a multiplicity of groups in a decentralized and bargaining
arena, but the nature of relations among participants was not in the
strictest sense pluralistic. The pluralist model stresses conflict and
conflict resolution through bargaining among groups and coalitions
organized around shared interests. The elitists stress conflict reduction
among formal officeholders in a much more restricted, centralized, and
stable arena. What Schattschneider saw was neither, but contained
elements of both. His political arena was decentralized and multi-
centered, but relationships among participants were based upon "mutual
non-interference" among uncommon interests. The "power structure"
was stabilized toward the "command posts" (in this case, the House
Ways and Means Committee), not because the officials were above pres-
sure groups, but because the pattern of access led to supportive relations
between pressure groups and officials. What may appear to one ob-
server as evidence of a power elite appears to another as decentralized
pluralism (to such an extent, indeed, that Schattschneider is often
credited with an important share in the founding of pluralist political
analysis). Schattschneider's masterful case-study actually reveals neither.
At one point he concludes: "A policy that is so hospitable and catholic
as the protective tariff disorganizes the opposition."4 In many important
4 Politics, Pressures and the Taritj, 88. The fact that Schattschneider holds his gen-
cases completely unrelated to the tariff and much more recent than
I930, we can find plenty of evidence to support this third or fourth5
approach to a "theory" of power and policy-making. But as a general
theory Schattschneider's conclusions would be no more satisfactory
than any one approach identified earlier.
The main trouble with all these approaches is that they do not
generate related propositions that can be tested by research and
experience. Moreover, the findings of studies based upon any one of
them are not cumulative. Finally, in the absence of logical relations
between the "theory" and the propositions, the "theory" becomes self-
directing and self-supportive. This is why I have employed the term
"theory" only with grave reservations and quotation marks.
The pluralist approach has generated case-study after case-study that
"proves" the model with findings directed by the approach itself.
Issues are chosen for research because conflict made them public; group
influence is found because in public conflict groups participate whether
they are influential or not. Group influence can be attributed because
groups so often share in the definition of the issue and have taken
positions that are more or less directly congruent with the outcomes.
An indulged group was influential, and a deprived group was un-
influential; but that leaves no room for group irrelevancy.
The elitist approach is no less without a means of self-assessment.
If power distributions are defined as "inherently hierarchical,"' then a
case of coalition politics either represents non-exhaustive research or
concerns an issue that is not fundamental and so only involves the
"middle levels of power."' One need not look for theoretical weaknesses
in Schattschneider's approach because his interpretation was mistakenly
thrown in with the pluralists. This is most unfortunate, because if the
differences between Schattschneider's discoveries (especially his insights
eralizations to the particular policy in question should be noted here as a point central
to my later arguments.
5 There are four approaches here if the "social stratification" school is kept separate
from the "power elite" school. While both make the same kinds of errors, each leads
to different kinds of propositions. In some hands they are, of course, indistinguishable
and, for good reason, Polsby in Community Power and Political Theory treats the
two as one. Since the distinction, once made, is not important here, I will more or
less follow Polsby's lead.
6 Milton Gordon, quoted in Polsby, 103.
7 Cf. C. Wright Mills, The Power Elite (New York 1956), 245: ". . . the political
analyst is generally on the middle levels of power himself. He knows the top only
by gossip; the bottom, if at all, only by 'research.'" Thus, Mills continues, the pro-
fessor and free-lance intellectual are "at home with the leaders of the middle level,
and . . . focus upon the middle levels and their balances because they are closer to
them."
682 WORLD POLITICS
into a different type of coal

been recognized, a more soph
have resulted. This is coming close to the approach I will presently
propose.
The controversies among the approaches to policy analysis, as well
as the logical and empirical weaknesses I have identified, have engen-
dered some of the best empirical work to be found in the literature. But
since most of this work is based on the local community, only a col-
lateral attack can be mounted from these findings against assumptions
and propositions concerning national power. Only similarly careful
and systematic studies of national political processes will lead to a
balanced and well-founded attack and to a more productive approach
to theory. The first step is certainly American Business and Public
Policy. While the authors do not offer an adequate theoretical alterna-
tive to established approaches, their findings, either explicitly or by
implication, are worthy of review in any effort to build such a theory.
American Business and Public Policy starts out as a study of "the
flow of communications to and within groups involved in the foreign-
policy decision-making process" (p. 5). But while this is carried through
for every type of participant and phase in the process, the authors are
in a position to deal as well with the substance of the communications,
perceptions of them by the recipients and, therefore, with some of the
vital aspects of influence. They have looked into the little black box of
political relationships among all the influence-seekers which the plural-
ists say contains the reality of influence and the elitists say doesn't really
exist anyhow. What they have found severely tests the assumptions of
both schools of thought. Without significant adjustments, the pluralist
framework (called by the authors the "pressure-group model") will be
able to encompass a great many of their propositions and findings only
with extreme difficulty. For example:
(i) There is no one-for-one nexus between the policies at stake and
interest-group activation (chap. 9). Protectionists were more highly
activated than anti-protectionists.
(2) The two "sides" of the controversy never met in a face-to-face
conflict with a settlement by compromise. In large part, the basis of
the protectionist coalition was a series of interests in individual tariff
items, in a relation of "uncommon interest" rather than of "tangential
interest." The basis of the anti-protectionist position was general
opposition in line "with the ideology of the times" (pp. I5off. and 209).
The outcome depended not upon compromise between the two sides in
Congress but upon whose definition of the situation prevailed. If tariff
is an instrument of foreign policy and genera

national purposes, the anti-protectionists win; i
tion of tariff as an aid to ioo,ooo individual
protectionists win.'
(3) Cohesion was found to be directly (not inversely) related to
overlapping membership (pp. 333 and 338). Most of the well-organized
"associations" were so heterogeneous on the tariff question that they
took no stand at all (pp. 334-38). Those groups that were at all cohesive
maintained their cohesion "by use of multiple group memberships for
purposes which might produce conflict within a single given group"
(p. 332).'
(4) Despite the strong saliency of the trade issues to many groups,
little in the way of direct attempts to influence policy-makers could be
found. Most group activity "involved interaction with people on the
same side." "Scholars have assumed that interest groups had clear
interests, of which they were aware. A number of the campaigns we
have described show clearly that a pressure group's function is fre-
quently to define the interest of its partisans" (p. 398).
(5) Massive activity, expenditure, and influence by monolithic inter-
est groups appeared to constitute an image that opposing groups create
of each other rather than the reality of pressure practices (Part iv, esp.
chap. 28).
(6) Most Congressmen were found to have their own independent
ideologies and interests and tended to read constituency support into
8 Thus, with no major decline in the numbers of industries and groups with a self-
interest in protection-indeed, with the defection of virtually the entire South from
the free-trade cause-liberal trade lost consistently until i962. But in i962 it won be-
cause tariff had finally lost its traditional definition. Lest it be concluded that the
Administration won merely because of its use of traditional log-rolling strategies in
its textile concessions, note that on the crucial votes in both Houses most of the
Democratic protectionist vote was Southern: in the House vote on the Mason motion
for recommittal, 37 of the 44 protectionist Democrats were Southern, mainly from
the textile and oil states of North and South Carolina, Texas, and Oklahoma. Ken-
nedy's strategy probably got him only Georgia's votes. In the Senate, the crucial vote
was on the "peril point," with the Southerners splitting io to io. Despite Kennedy's
concessions, the defectors included two from Georgia, South Carolina, Mississippi, and
Virginia and one from North Carolina.
9I obtain from this and other propositions in the book a conclusion not reached
by the authors but strongly supported by their findings: there are probably several
kinds of coalitions (rather than "a politics of coalition" where all coalitions are equiv-
alent in every way except size and value of resources). Each type of coalition is ap-
propriate for certain types of issues and each has extreme significance for outcomes,
perhaps as much as cohesion and access. More on this presently.
10Again suggesting that a distinct kind of coalition is involved here, the authors
say: ". . . neither the interest of the New England textile manufacturers in supporting
oil imports nor that of the Eastern railroads in opposing these imports were in any
way self-evident. Offered the proper coalition, they both might well have been per-
suaded that their interest was in the opposite direction" (p. 398).
684 WORLD POLITICS
their mail rather than to refle

from constituents and groups are so numerous and conflicting that
Congressmen have a maximum-choice situation. And, groups were
found to hold fairly strictly to the rule of contacting only those
Congressmen already known to agree with the group's aims. The
predominant group role was found by the authors to be as "service
bureaus.""
(7) Congress displayed no capacity for, nor did Congressmen display
any desire for, retaining the power of tariff and trade decision-making
(esp. pp. 35, 37-38, I97, and 455-56).
But for every single problem that the tariff decisions of a decade
pose for the pluralist model, they pose several for the elitist or the
stratification approaches:
(i) There was never, throughout the entire modern history of tariff
and foreign trade, any evidence that the top military played a role of
any sort. This is true despite the fact that the effectiveness of military
aid and military installations abroad was deeply influenced by the
economic health of the recipient countries. The Millsian complex of
military-industrial intimacy would make military participation in
foreign trade decisions all the more probable.
(2) Only a minority of businessmen large and small ever felt them-
selves aligned with either side of the issue (p. I25). (This finding hangs
heavy upon the elitists who argue that the top business leaders care and
control, and upon the pluralists who assume that most business leaders
are interested and compete.)
(3) Of I28 classifiable responses by big business leaders in their
sample, the authors report that 56 per cent favored freer trade, 3 per
cent favored raising tariffs, and the remainder favored leaving tariffs
"as is" (p. iii). A more conservative estimate of the distribution of
attitudes was a 3-tO-I preference for freer trade (chap. 8). Yet the
protectionists won year after year following World War II.
(4) Many of the leaders of the largest firms refused to generalize at
all and remained actively inactive throughout the decade because their
firms were too diversified to have a clear interest. General Electric and
-- Access to and information about Congress and Congressmen was so poor that the
liberal CNTP's list of Congressmen's positions was almost a third incorrect for both
Houses in the case of those whose positions had been identified at all. Nearly 15 per cent
of the members of Congress were listed as "undecided," and in most cases this meant
no contact at all had been made. Many pluralist writers before Bauer, Pool, and
Dexter have recognized the "service bureau" role, but rather than reexamine their
premises, they usually catalogue this as a "form of influence."
DuPont were identified as "typical of large fir

individual division managers to lobby or not as
(5) Most other industrialists in the top "comm
favored freer trade. The leading liberal trade gro
a National Trade Policy, was composed of many
of the "establishment": Harry Bullis of General M
Gillette, John McCloy of Chase National, John Co
and Charles Taft and George Ball for themselv
Their leading opposition, the Nation-wide Comm
Industry and Labor on Export-Import Policy w
"middle levels of power" interest-group leaders as
UMW, National Coal Association, Manufacturin
tion, and the Window Glass Cutters' League o
The CNTP and its official parent, the Randall C
Steel, etc.), had gone out of existence well befo
successes.
(6) The final outcome in the Trade Act of i962 was the direct result
of classic President-Congress interplay, involving many of the kinds of
compromises that the most vulgar pluralist view of American politics
would lead one to expect. Congressmen appeared to be more inde-
pendent of constituents and groups than the pressure-group model
would allow. But, on the other hand, the degree to which the Executive
would be able to use trade as an instrument of international politics was
determined in Congress; and it seemed that such a relationship would
continue (pp. 73-79).
Elitists might accommodate to the extraordinarily careful analysis
and unimpeachable findings of American Business and Public Policy
by treating the trade issues of the whole postwar period as "middle-
level." But this avoids all the interesting questions and begs a few as
well. It was the large industrialists and the high members of the
political directorate themselves who defined the issue as fundamental,
particularly in the late i950's after the Common Market was seen to be
an immediate rather than a future reality. These were the men who kept
trade on the agenda for over ten years. Yet there is nothing in the
politics of trade decisions to support the notion that power follows
directly from selected, established institutional positions in industry or
military and public administration.
Bauer, Pool, and Dexter discovered that the pluralist model was of
little use to them. It seems equally clear that the elitist model would
have been of even less assistance. But this is because neither approach
686 WORLD POLITICS
is a model. Each is, if anything, a self-validating standpoint; the

pluralist approach suggests what to look for and the elitist model
suggests perhaps what not to look for. Since neither is a theory, neither
has much bearing on specific cases. At the end of an empirical study,
neither approach affords a means for cumulating the data and findings
in coherent and logical abstractions with other findings; they merely
provide the basis for repeating the assumptions of the beginning. The
following example is drawn from one of the very best case studies: "The
realization of national policy in the United States depends upon the
formation and maintenance of coalitions (often of a temporary nature)
overcoming the separation of powers, overcoming the fragmentation
of power within the legislative and executive branches of government,
overcoming weak cohesion by joining elements of both parties and
many interests, to accomplish a desired objective. There is no small
group of men who, if they agree among themselves, can assure favorable
decisions on all or most of the important national policies. Power in
Congress is fragmented and dispersed. Bits and pieces of influence are
scattered (unequally to be sure) among committee chairmen, party
leaders, and many others. . . National policy is approved or rejected
by building a coalition majority through bargaining and the proposal
of objectives appealing to a wide variety of interests.... 12
The comment "This is where we came in" would be appropriate not
only for these general remarks in Wildavsky's study, but also for vir-
tually all case-studies and their generalizations since Bentley. In many
respects, one can say of both pluralists and elitists what was said of
Debussy, but say it here more meaningfully: they opened up a couple
of new streets that have turned out to be blind alleys.
II. ARENAS OF POWER: AN INTERPRETATIVE SCHEME FOR CASES

American Business and Public Policy, despite its richness as a case-
study in the politics of foreign trade, suffers the one debilitating handi-
cap of all case-studies, the problem of uniqueness. The case casts seri-
ous doubts on the perspectives employed in most case-studies because
of the rare comprehensiveness and exhaustiveness of its analysis, which
goes far beyond the storybook and paste-pot relations among partici-
pants usually contained in case-studies. But still, how are we to know
the extent to which the patterns the authors discover apply to all cases?
How are we to know at least toward what class of cases we can gen-
eralize ? The authors have too little to say about this. The subtitle of the
12 Aaron Wildavsky, Dixon-Yates: A Study in Power Politics (New Haven i962),
3II.
book, "The Politics of Foreign Trade," implies that everything they

say obtains only for this class of cases. However, at one point in the
book they argue that the case belongs to a much larger class, a class
which probably includes all domestic decisions except "some really
new and unexpected issue" (p. 46i): "So, on the whole, our case
seems to us typical of any issue having economic implications for a
number of industries and where, through a historical process, insti-
tutional alignments and expectations have been established" (pp. 460-
6i). On the other hand, one cannot be entirely sure how inclusive a
category the authors had in mind, for at another point they say:
". . . the more an organization represents the business community as a
whole, the more unlikely it is to become committed on such an issue as
foreign-trade policy" (p. 339). The qualification ("on such an issue
as") would hardly be necessary if they took foreign-trade policy as
typical of all on-going policy processes.
Thus, the authors are unusually clear about their research aim, and
self-conscious and skillful as to appropriate techniques of research and
data manipulation; but they are far less than clear or self-conscious
and skillful about defining their case in terms of some general scheme
of classification. If their findings are to be judged for their applicability
to all of national domestic policy-making, then we are about where
we were before the book was published, except for some healthy nega-
tive insights as to prevailing "theories." If foreign trade is part of a
smaller category of cases, then we would have to know what the class
is, how it is distinguished from others, and what attributes are special
to that class and what are general to all classes. Rather than dealing
with these problems, the authors came out in favor of cashiering the
present approaches altogether and of adopting a communications or
"transactional" approach.13 So, despite the fact that Bauer, Pool, and
Dexter have managed an outstanding empirical study, the job of
"making something of" the findings still remains, as it has remained
for some time.
It seems to me that the reason for lack of interesting and non-obvious
generalization from cases and other specific empirical studies is clearly
that the broad-gauged theories of politics are not related, perhaps are
not relatable, to observable cases. In general, American political science
seems to be subject to a continuing fission of theory and research, in
13 "Inter alia, the transactional analysis here employed makes the concept of 'power'
and 'pressure' in the ordinary political-science sense of the terms somewhat more
difficult to employ; Arthur F. Bentley himself pointed out to one of us in 1936 that
he had long since abandoned these notions as not useful for systematic analysis"
(p. 460). See also p. 456.
688 WORLD POLITICS
which the empiricist is not suf

builder and the system-builder is not sufficiently mindful (if at all)
of the role that theory is supposed to play. What is needed is a basis for
cumulating, comparing, and contrasting diverse findings. Such a
framework or interpretative scheme would bring the diverse cases
and findings into a more consistent relation to each other and would
begin to suggest generalizations sufficiently close to the data to be
relevant and sufficiently abstract to be subject to more broadly theoreti-
cal treatment.
An attempt at such a framework follows. For over two years prior
to the publication of American Business and Public Policy I had
been working on a general interpretative scheme.14 The hypotheses
drawn from the scheme have so far anticipated most of the patterns
described in existing case literature, and few of those patterns not
anticipated have been found to be inconsistent with a logical exten-
sion of the scheme. A review article as the first published use of the
scheme for national politics seemed appropriate because Bauer, Pool,
and Dexter's case-study is the most elaborate case yet published, and
it appeared long after most of my hypotheses had been developed.15
The scheme is based upon the following argument: (i) The types
of relationships to be found among people are determined by their
expectations-by what they hope to achieve or get from relating to
others. (2) In politics, expectations are determined by governmental
outputs or policies. (3) Therefore, a political relationship is determined
by the type of policy at stake, so that for every type of policy there is
likely to be a distinctive type of political relationship. If power is defined
as a share in the making of policy, or authoritative allocations, then
the political relationship in question is a power relationship or, over
time, a power structure. As Dahl would say, one must ask, "Power
for what?" One must control for the scope of power and look for
elites, power structures, and the like within each of the predefined
scopes or "issue areas."1 My analysis moves in this direction, but
14 The first formulation, developed for urban politics, appears in my study, At the
Pleasure of the Mayor (New York i964), chaps. 6 and 7. The scheme for national
politics which is presented in this article is an adaptation of the national "arenas of
power" discussed in a book now in preparation.
15 Their study is of further interest because they are dealing with a type of policy
which made a transition from one of my "arenas" to another between 1930 and i962.
That the politics of tariff changed accordingly is the best test I have yet found for my
scheme. The very differences they find between i950's patterns and those reported so
exceedingly well by Schattschneider-and which are branded by Bauer, Pool, and
Dexter as inconsistent with Schattschneider or as due vaguely to the "changing times"
-were both consistent with and anticipated by my scheme.
16 Dahl, Who Governs?, and Polsby, Community Power, esp. chap. 6.
farther. Issues as such are too ephemeral; it is on

expectations and a history of earlier governmen
type that single issues are fought out. The study
vides a good test of hypotheses about structure, but the hypotheses
must be arrived at in some other, independent way.
Obviously, the major analytic problem is that of identifying types
of outputs or policies. The approach I have taken is to define policies
in terms of their impact or expected impact on the society. When
policies are defined this way, there are only a limited number of types;
when all is said and done, there are only a limited number of functions
that governments can perform. This approach cashiers the "politics
of agriculture" and the "politics of education" or, even more narrowly
but typically, "the politics of the ARA bill" or "the politics of the i956
Aid to Education bill," in which the composition and strategy of the
participants are fairly well-known before the study is begun. But it
maintains the pluralist's resistance to the assumption that there is only
one power structure for every political system. My approach replaces
the descriptive, subject-matter categories of the pluralists with func-
tional categories. There is no need to argue that the classification
scheme exhausts all the possibilities even among domestic policies; it
is sufficient if most policies and the agencies that implement them can
be categorized with little, if any, damage to the nuances.
There are three major categories of public policies in the scheme:
distribution, regulation, and redistribution. These types are historically
as well as functionally distinct, distribution being almost the exclu-
sive type of national domestic policy from I789 until virtually i890.
Agitation for regulatory and redistributive policies began at about
the same time, but regulation had become an established fact before
any headway at all was made in redistribution."7
These categories are not mere contrivances for purposes of simplifi-
cation. They are meant to correspond to real phenomena-so much
so that the major hypotheses of the scheme follow directly from the
categories and their definitions. Thus, these areas of policy or govern-
ment activity constitute real arenas of power. Each arena tends to de-
velop its own characteristic political structure, political process, elites,
17 Foreign policy, for which no appropriate "-tion" word has been found, is obviously
a fourth category. It is not dealt with here for two reasons. First, it overly extends the
analysis. Second, and of greater importance, it is in many ways not part of the same
universe, because in foreign policy-making America is only a subsystem. Winston
Churchill, among other foreigners, has consistently participated in our foreign policy
decisions. Of course, those aspects of foreign and military policy that have direct do-
mestic implications are included in my scheme.
690 WORLD POLITICS
and group relations. What remains is to identify these arenas, to

formulate hypotheses about the attributes of each, and to test the
scheme by how many empirical relationships it can anticipate and
explain.
AREAS OF POLICY DEFINED
(i) In the long run, all governmental policies may be considered

redistributive, because in the long run some people pay in taxes more
than they receive in services. Or, all may be thought regulatory be-
cause, in the long run, a governmental decision on the use of resources
can only displace a private decision about the same resource or at least
reduce private alternatives about the resource. But politics works in
the short run, and in the short run certain kinds of government de-
cisions can be made without regard to limited resources. Policies of
this kind are called "distributive," a term first coined for nineteenth-
century land policies, but easily extended to include most contempo-
rary public land and resource policies; rivers and harbors ("pork
barrel") programs; defense procurement and R & D; labor, business,
and agricultural "clientele" services; and the traditional tariff. Dis-
tributive policies are characterized by the ease with which they can
be disaggregated and dispensed unit by small unit, each unit more
or less in isolation from other units and from any general rule. "Patron-
age" in the fullest meaning of the word can be taken as a synonym
for "distributive." These are policies that are virtually not policies at
all but are highly individualized decisions that only by accumulation
can be called a policy. They are policies in which the indulged and
the deprived, the loser and the recipient, need never come into direct
confrontation. Indeed, in many instances of distributive policy, the de-
prived cannot as a class be identified, because the most influential
among them can be accommodated by further disaggregation of the
stakes.
(2) Regulatory policies are also specific and individual in their
impact, but they are not capable of the almost infinite amount of dis-
aggregation typical of distributive policies. Although the laws are
stated in general terms ("Arrange the transportation system artisti-
cally." "Thou shalt not show favoritism in pricing."), the impact of
regulatory decisions is clearly one of directly raising costs and/or re-
ducing or expanding the alternatives of private individuals ("Get off
the grass!" "Produce kosher if you advertise kosher!"). Regulatory
policies are distinguishable from distributive in that in the short run
the regulatory decision involves a direct choice as to who will be in-
dulged and who deprived. Not all applicants for a single television
channel or an overseas air route can be propitiated. Enforcement of
an unfair labor practice on the part of management weakens manage-
ment in its dealings with labor. So, while implementation is firm-by-
firm and case-by-case, policies cannot be disaggregated to the level of
the individual or the single firm (as in distribution), because individ-
ual decisions must be made by application of a general rule and there-
fore become interrelated within the broader standards of law. De-
cisions cumulate among all individuals affected by the law in roughly
the same way. Since the most stable lines of perceived common impact
are the basic sectors of the economy, regulatory decisions are cumula-
tive largely along sectoral lines; regulatory policies are usually dis-
aggregable only down to the sector level.'8
(3) Redistributive policies are like regulatory policies in the sense
that relations among broad categories of private individuals are in-
volved and, hence, individual decisions must be interrelated. But on
all other counts there are great differences in the nature of impact.
The categories of impact are much broader, approaching social classes.
They are, crudely speaking, haves and have-nots, bigness and small-
ness, bourgeoisie and proletariat. The aim involved is not use of prop-
erty but property itself, not equal treatment but equal possession, not
behavior but being. The fact that our income tax is in reality only
mildly redistributive does not alter the fact of the aims and the stakes
involved in income tax policies. The same goes for our various "wel-
fare state" programs, which are redistributive only for those who en-
tered retirement or unemployment rolls without having contributed
at all. The nature of a redistributive issue is not determined by the
outcome of a battle over how redistributive a policy is going to be.
Expectations about what it can be, what it threatens to be, are de-
terminative.
ARENAS OF POWER
Once one posits the general tendency of these areas of

governmental activity to develop characteristic political s
number of hypotheses become compelling. And when the various
18 A "sector" refers to any set of common or substitutable commodities or services
or any other form of established economic interaction. Sectors therefore vary in size
because of natural economic forces and because of the different ways they are
identified by economists or businessmen. They vary in size also because they are
sometimes defined a priori by the observer's assessment of what constitutes a common
product and at other times are defined a posteriori by the trade associations that repre-
sent the identification of a sector by economic actors themselves.
692 WORLD POLITICS
hypotheses are accumulated, th

arenas begin quickly to resemble, respectively, the three "general"
theories of political process identified earlier. The arena that develops
around distributive policies is best characterized in the terms of Schatt-
schneider's findings. The regulatory arena corresponds to the pluralist
school, and the school's general notions are found to be limited pretty
much to this one arena. The redistributive arena most closely approx-
imates, with some adaptation, an elitist view of the political process.
(i) The distributive arena can be identified in considerable detail
from Schattschneider's case-study alone. What he and his pluralist
successors did not see was that the traditional structure of tariff politics
is also in largest part the structure of politics of all those diverse pol-
icies identified earlier as distributive. The arena is "pluralistic" only
in the sense that a large number of small, intensely organized interests
are operating. In fact, there is even greater multiplicity of participants
here than the pressure-group model can account for, because essentially
it is a politics of every man for himself. The single person and the
single firm are the major activists. Bauer, Pool, and Dexter, for in-
stance, are led to question seriously the "pressure-group model" be-
cause of the ineffectiveness of virtually all the groups that should have
been most active and effective.
Although a generation removed, Schattschneider's conclusions about
the politics of the Smoot-Hawley Tariff are almost one-for-one appli-
cable to rivers and harbors and land development policies, tax exemp-
tions, defense procurement, area redevelopment, and government
"services." Since there is no real basis for discriminating between those
who should and those who should not be protected [indulged], says
Schattschneider, Congress seeks political support by "giving a limited
protection [indulgence] to all interests strong enough to furnish for-
midable resistance." Decision-makers become "responsive to consid-
erations of equality, consistency, impartiality, uniformity, precedent,
and moderation, however formal and insubstantial these may be."1"
Furthermore, a "policy that is so hospitable and catholic . . . disor-
ganizes the opposition."20
When a billion-dollar issue can be disaggregated into many millions
of nickel-dime items and each item can be dealt with without regard
to the others, multiplication of interests and of access is inevitable,
and so is reduction of conflict. All of this has the greatest of bearing
on the relations among participants and, therefore, the "power struc-
19 Politics, Pressures, 85.
20 Ibid., 88.
ture." Indeed, coalitions must be built to pass legislation and "make

policy," but what of the nature and basis of the coalitions? In the
distributive arena, political relationships approximate what Schatt-
schneider called "mutual non-interference"-"a mutuality under which
it is proper for each to seek duties [indulgences] for himself but im-
proper and unfair to oppose duties [indulgences] sought by others."21
In the area of rivers and harbors, references are made to "pork barrel"
and "log-rolling," but these colloquialisms have not been taken suf-
ficiently seriously. A log-rolling coalition is not one forged of con-
flict, compromise, and tangential interest but, on the contrary, one
composed of members who have absolutely nothing in common; and
this is possible because the "pork barrel" is a container for unrelated
items. This is the typical form of relationship in the distributive arena.
The structure of these log-rolling relationships leads typically, though
not always, to Congress; and the structure is relatively stable because
all who have access of any sort usually support whoever are the leaders.
And there tend to be "elites" of a peculiar sort in the Congressional
committees whose jurisdictions include the subject-matter in question.
Until recently, for instance, on tariff matters the House Ways and
Means Committee was virtually the government. Much the same can
be said for Public Works on rivers and harbors.22 It is a broker leader-
ship, but "policy" is best understood as cooptation rather than conflict
and compromise.
Bauer, Pool, and Dexter are astonished to discover trade associa-
tions and other groups suffering from lack of funds and support. They
see as paradoxical the fact that "protectionism" as a policy could win out
time after time even when a majority of businessmen and Congressmen
seemed on principle to favor freer trade. (There are instances of this run-
ning clear back to the i890's.) They see as purposive Congress's "giving
up" tariff-making because the "power to dole out favors is not worth the
price of having to beat off and placate the insistent pleas of petitioners"
(p. 37). Astonishment and the detection of paradox and a "Congres-
sional group mind" are evidences of an insufficiently broad point of
21 Ibid., I35-36.
22 The stable, intimate interlocking of Congressional committeemen and their sup-
port groups in the Rivers and Harbors Congress and the Corps of Engineers has been
made famous by Arthur Maass; see Muddy Waters: The Army Engineers and the
Nation's Rivers (Cambridge, Mass., 1951), and especially "Congress and Water Re-
sources," American Political Science Review, XLIV (September 1950), 576-92, reprinted
in my reader, Legislative Politics USA (Boston I962). Cited widely as an example of
interest-group strategy and access, this case has not until now, as far as I know, been
given its proper significance. That significance comes clear within my scheme. The
pattern approaches that of the tariff but not of regulatory situations.
view. There are good and theoretically interesting

these phenomena. Distributive issues individualize c
the basis for highly stable coalitions that are virtu
larger policy outcomes; thousands of obscure decisions are merely
accumulated into a "policy" of protection or of natural-resources de-
velopment or of defense subcontracting. And Congress did not "give
up" the tariff; as the tariff became a matter of regulation (see below),
committee elites lost their power to contain the participants because
obscure decisions became interrelated, therefore less obscure, and more
controversy became built in and unavoidable.23
(2) The regulatory arena could hardly be better identified than in
the thousands of pages written for the whole polity by the pluralists.
But, unfortunately, some translation is necessary to accommodate
pluralism to its more limited universe. The regulatory arena appears
to be composed of a multiplicity of groups organized around tangential
relations or David Truman's "shared attitudes." Within this narrower
context of regulatory decisions, one can even go so far as to accept the
most extreme pluralist statement that policy tends to be a residue of
the interplay of group conflict. This statement can be severely criti-
cized only by use of examples drawn from non-regulatory decisions.
As I argued before, there is no way for regulatory policies to be dis-
aggregated into very large numbers of unrelated items. Because in-
dividual regulatory decisions involve direct confrontations of indulged
and deprived, the typical political coalition is born of conflict and
compromise among tangential interests that usually involve a total
sector of the economy. Thus, while the typical basis for coalition in
distributive politics is uncommon interests (log-rolling), an entirely
different basis is typical in regulatory politics. The pluralist went wrong
only in assuming the regulatory type of coalition is the coalition.24
23 Schattschneider, in his more recent book The Semi-sovereign People (New York
i960), offers some fascinating propositions about the "scope of conflict" which can
easily be subsumed within the scheme offered here.
24 I was surprised and pleased on rereading Truman's The Governmental Process
(New York 1951), after completing the first draft of this article, to find that he
identified two types of "mutual assistance," alliances and log-rolling (pp. 362-68). In
my scheme, as will soon be clear, there are two types of "alliance," tangential interest
and ideology. But what is of interest here is that Truman supports his distinction with
examples perfectly congruent with my theory. His case of the alliance is the aggrega-
tion of interests around the 1946 Employment Act (redistribution, even if a peculiar
"law"). The typical log-rolling situation he identifies with rivers and harbors ap-
propriations (distribution). The difference between us is that my scheme considers
these patterns of coalition as revealing fundamental political relations that are limited
to certain types of issues, while Truman implies that they are two strategies in all
inventory of strategies more or less appropriate to any issue.
One of the most significant differences between the pluralists and

Bauer, Pool, and Dexter-the treatment of the phenomenon and ef-
fects of overlapping membership-becomes consistent and supportive
within this scheme. In fact, it helps to clarify the distinctions I am try-
ing to draw here. Truman, for instance, stresses overlapping member-
ship as a source of conflict, the function of overlapping membership
as the reduction of cohesion in any given group. In contrast, Bauer,
Pool, and Dexter found that in tariff politics this very overlapping of
membership was a condition for cohesion: ". . . unanimity (or co-
hesion) is maintained by the use of multiple group memberships for
purposes that might produce conflict within a single given group"
(p. 332n). They observed that overlapping is a form of specialization
allowing individual firms, or special constituent groups within larger
associations, the freedom to pursue outside the association the goals
that are contrary to other associated groups. Meanwhile the cohesion
of the larger group is preserved for the goals that all the constituent
groups share. The fact appears to be that both positions are correct.
Owing to the unrelatedness of issues in distributive politics, the activities
of single participants need not be related but rather can be specialized
as the situation warrants it. But the relatedness of regulatory issues, at
least up to the sector level of the trade association, leads to the con-
tainment of all these within the association and, therefore, to the
dynamic situation ascribed erroneously by Truman to all intergroup
relations in all issues. When all the stakes are contained in one organ-
ization, constituents have no alternative but to fight against each other
to shape the policies of that organization or actually to abandon it.
What this suggests is that the typical power structure in regulatory
politics is far less stable than that in the distributive arena. Since co-
alitions form around shared interests, the coalitions will shift as the
interests change or as conflicts of interest emerge. With such group-
based and shifting patterns of conflict built into every regulatory issue,
it is in most cases impossible for a Congressional committee, an ad-
ministrative agency, a peak association governing board, or a social
elite to contain all the participants long enough to establish a stable
power elite. Policy outcomes seem inevitably to be the residue remain-
ing after all the reductions of demands by all participants have been
made in order to extend support to majority size. But a majority-sized
coalition of shared interests on one issue could not possibly be entirely
appropriate for some other issue. In regulatory decision-making, re-
lationships among group leadership elements and between them on
any one or more points of governmental access are too unstable to
form a single policy-making elite. As a consequence, decision-making

tends to pass from administrative agencies and Congressional com-
mittees to Congress, the place where uncertainties in the policy process
have always been settled. Congress as an institution is the last resort
for breakdowns in bargaining over policy, just as in the case of parties
the primary is a last resort for breakdowns in bargaining over nomina-
tions. No one leadership group can contain the conflict by an almost
infinite subdivision and distribution of the stakes. In the regulatory
political process, Congress and the "balance of power" seem to play the
classic role attributed to them by the pluralists, attacked as a theory by
C. Wright Mills, and at least seriously questioned by Bauer, Pool, and
Dexter.
The most interesting thing about the work of Bauer, Pool, and
Dexter, from the standpoint of my scheme, is that they studied a policy
that was undergoing a transition from distribution to regulation. It is,
I feel, for this reason that they find some support for the pressure-group
model but not enough to convince them of its utility. But it is this very
transition that makes their case-study so interesting. Beginning with
reciprocity in the I930'S, the tariff began to lose its capacity for infinite
disaggregation because it slowly underwent redefinition, moving away
from its purely domestic significance towards that of an instrument of
international politics. In brief, the tariff, especially following World
War II and our assumption of peacetime international leadership,
became a means of regulating the domestic economy for international
purposes. The significant feature here is not the international but the
regulatory part of the redefinition. As the process of redefinition took
place, a number of significant shifts in power relations took place as
well, because it was no longer possible to deal with each dutiable item
in isolation. Everything in Bauer, Pool, and Dexter points toward the
expansion of relationships to the level of the sector. The political
problem of the South was the concentration of textile industry there.
Coal, oil, and rails came closer and closer to coalition. The final shift
came with the i962 Trade Expansion Act, which enabled the President
for the first time to deal with broad categories (to the sector) rather
than individual commodities.
Certain elements of distributive politics remain, for two obvious
reasons. First, there are always efforts on the part of political leaders to
disaggregate policies because this is the best way to spread the patronage
and to avoid conflict. (Political actors, like economic actors, probably
view open competition as a necessary evil or a last resort to be avoided
at almost any cost.) Second, until i962, the basic tariff law and
schedules were still contained in the Smoot-Hawley Act. This act was
amended by Reciprocal Trade but only to the extent of allowing ne-
gotiated reductions rather than reductions based on comparative costs.
Until i962, tariff politics continued to be based on commodity-by-
commodity transactions, and thus until then tariff coalitions could be
based upon individual firms (or even branches of large and diversified
firms) and log-rolling, unrelated interests. The escape clause and peril
point were maintained in the i950's so that transactions could be made
on individual items even within reciprocity. And the coalitions of
strange bedfellows continued: "Offered the proper coalition, they
both [New England textiles and Eastern railroads] might well have
been persuaded that their interest was in the opposite direction"
(p. 398).
But despite the persistence of certain distributive features, the true
nature of tariff in the i960's emerges as regulatory policy with a de-
veloping regulatory arena. Already we can see some changes in Congress
even more clearly than the few already observed in the group structure.
Out of a committee (House Ways and Means) elite, we can see the
emergence of Congress in a pluralist setting. Even as early as I954-I955,
the compromises eventually ratified by Congress were worked out, not
in committee through direct cooptation of interests, but in the Randall
Commission, a collection of the major interests in conflict (p. 47).
Those issues that could not be thrashed out through the "group process"
also could not be thrashed out in committee but had to pass on to
Congress and the floor. After I954 the battle centered on major
categories of goods (even to the extent of a textile management-union
entente) and the battle took place more or less openly on the floor
(e.g., pp. 6o, 62, and 67). The weakening of the Ways and Means
Committee as the tariff elite is seen in the fact that in i955 Chairman
Cooper was unable to push a closed rule through. The Rules Com-
mittee, "in line with tradition," granted a closed rule but the House
voted it down 207-I78 (p. 63).25 Bauer, Pool, and Dexter saw this as a
victory for protectionism, but it is also evidence of the emerging
regulatory arena-arising from the difficulty of containing conflict and
policy within the governing committee. The last effort to keep the
tariff as a traditional instrument of distributive politics-a motion by
25 Sam Rayburn made one of his rare trips from rostrum to floor to support the
closed rule and the integrity of Ways and Means: "Only once in the history of the
House, in forty-two years in my memory, has a bill of this kind and character been
considered except under a closed rule . . ." (p. 64, emphasis added). It was on the
following morning that Rayburn expressed his now-famous warning to the frosh: "If
you want to get along, go along" (p. 64).
Reed to recommit, with instructions to write in a provision that

Tariff Commission rulings under the escape clause be final except where
the President finds the national security to be involved-was voted
down 206-i99 (pp. 64-65). After that, right up to i962, it was clear that
tariff decisions would not be made piecemeal. Tariff became a regula-
tory policy in i962; all that remains of distributive politics now are
quotas and subsidies for producers of specific commodities injured by
general tariff reductions.
(3) If Bauer, Pool, and Dexter had chosen a line of cases from the
redistributive arena for their intensive analysis, most assuredly they
would have found themselves in an altogether different universe,
proposing different generalizations, expressing different doubts. The
same would have been true of Schattschneider and of the pluralist
students of regulatory cases. Compared particularly with the regulatory
area, very few case-studies of redistributive decisions have ever been
published. This in itself is a significant datum-which Mills attributes
to the middle-level character of the issues that have gotten attention.
But, whatever the reasons, it reduces the opportunities for elaborating
upon and testing the scheme. Most of the propositions to follow are
illustrated by a single case, the "welfare state" battle of the I930'S. But
this case is a complex of many decisions that became one of the most
important acts of policy ever achieved in the United States. A brief
review of the facts of the case will be helpful.26 Other cases will be
referred to in less detail from time to time.
As the I934 mid-term elections approached, pressures for a federal
social security system began to mount. The Townsend Plan and the
Lundeen Bill had become nationally prominent and were gathering
widespread support. Both schemes were severely redistributive, giving
all citizens access to government-based insurance as a matter of right.
In response, the President created in June of I934 a Committee on
Economic Security (CES) composed of top cabinet members with
Secretary of Labor Perkins as chairman. In turn, they set up an
Advisory Council and a Technical Board, which held hearings, con-
ducted massive studies, and emerged on January I7, I935, with a bill.
The insiders around the CES were representatives of large industries,
business associations, unions, and the most interested government bu-
reaucracies. And the detailed legislative histories reveal that virtually all
26 The facts and events are taken from Paul H. Douglas, Social Security in the United
States (New York I936); Edwin E. Witte, The Development of the Social Security Act
(Madison, Wis., i962); Committee on Economic Security, Report to the President
(Washington, GPO, I935); and Frances Perkins, The Roosevelt I Knew (New York
I946).
of the debate was contained within the CES and

mature bill emerged. Since not all of the major
in the CES's bill, its members turned to Congress with far from a
common front. But the role of Congress was still not what would have
been expected. Except for a short fight over committee jurisdiction
(won by the more conservative Finance and Ways and Means com-
mittees) the legislative process was extraordinarily quiet, despite the
import of the issues. Hearings in both Houses brought forth very few
witnesses, and these were primarily CES members supporting the bill,
and Treasury Department officials, led by Morgenthau, opposing it with
"constructive criticism."
The Congressional battle was quiet because the real struggle was
taking place elsewhere, essentially between the Hopkins-Perkins bu-
reaucracies and the Treasury. The changes made in the CES bill had
all been proposed by Morgenthau (the most important one being the
principle of contribution, which took away the redistributive sting).
And the final victory for Treasury and mild redistribution came with
the removal of administrative responsibility from both Labor and
Hopkins's FERA. Throughout all of this some public expressions of
opinion were to be heard from the peak associations, but their efforts
were mainly expended in the quieter proceedings in the bureaucracies.
The Congress's role seems largely to have been one of ratifying agree-
ments that arose out of the bureaucracies and the class agents repre-
sented there. Revisions attributable to Congress concerned such matters
as exceptions in coverage, which are part of the distributive game that
Congress plays at every opportunity. The principle of the Act was set
in an interplay involving (quietly) top executives and business and
labor leaders.
With only slight changes in the left-right positions of the participants,
the same pattern has been observed in income tax decisions.27 Professor
Surrey notes: "The question, 'Who speaks for tax equity and tax
fairness?,' is answered today largely in terms of only the Treasury
Department" (p. ii64). "Thus, in tax bouts . . . it is the Treasury
versus percentage legislation, the Treasury versus capital gains, the
Treasury versus this constituent, the Treasury versus that private group.
. . . As a consequence, the congressman . . . [sees] a dispute . . . only
as a contest between a private group and a government department"
(pp. ii65-66). Congress, says Surrey, "occupies the role of mediator
between the tax views of the executive and the demands of the pressure
27 Stanley S. Surrey, "The Congress and the Tax Lobbyist: How Special Tax Pro-
visions Get Enacted," Harvard Law Review, LXX (May I957), II45-82.
groups" (p. 1154). And when the tax issues "are at a major political
level, as are tax rates or personal exemptions, then pressure groups,
labor organizations, the Chamber of Commerce, the National As-
sociation of Manufacturers, and the others, become concerned" (p.
ii66). The "average congressman does not basically believe in the
present income tax in the upper brackets" (p. ii50), but rather than
touch the principle he deals in "special hardship" and "penalizing" and
waits for decisions on principle to come from abroad. Amidst the I954-
I955 tax controversies, for example, Ways and Means members decided
to allow each member one bill to be favorably reported if the bill met
with unanimous agreement (p. II57).
Issues that involve28 redistribution cut closer than any others along
class lines and activate interests in what are roughly class terms. If
there is ever any cohesion within the peak associations, it occurs on
redistributive issues, and their rhetoric suggests that they occupy them-
selves most of the time with these.29 In a ten-year period just before and
after, but not including, the war years, the Manufacturers' Association
of Connecticut, for example, expressed itself overwhelmingly more
often on redistributive than on any other types of issues.30 Table i
summarizes the pattern, showing that expressions on generalized issues
involving basic relations between bourgeoisie and proletariat outnum-
bered expressions on regulation of business practices by 870 to 418,
despite the larger number of issues in the latter category.3' This pattern
goes contrary to the one observed by Bauer, Pool, and Dexter in tariff
politics, where they discovered, much to their surprise, that self-interest
did not activate both "sides" equally. Rather, they found, the con-
28 "Involve" may appear to be a weasel word, but it is used advisedly. As I argued
earlier when defining redistribution, it is not the actual outcomes but the expectations
as to what the outcomes can be that shape the issues and determine their politics. One
of the important strategies in any controversial issue is to attempt to define it in re-
distributive terms in order to broaden the base of opposition or support.
29 In personal conversations, Andrew Biemiller of AFL-CIO has observed that this
is true even of his group. He estimates that perhaps from 8o to go per cent of their
formal policy expressions deal with welfare and general rights of collective bargaining
and that only occasionally does the central board touch specific regulatory issues.
30 Robert E. Lane, The Regulation of Businessmen (New Haven I953), 38ff.
31 Note also in the table the fairly drastic contrast in the proportion of references
that expressed approval. Similarly drastic differences are revealed in Lane's figures on
the reasons given for expressing disapproval. On those issues I call redistributive, the
overwhelmingly most important reason is "coerciveness." In contrast, this reason was
given for about io per cent of general trade regulation and anti-trust references, 3 per
cent of the basing-point negative references, and not once when Miller-Tydings and
Robinson-Patman were denounced. For regulatory issues, the reason for disapproval
given most frequently was that the policy was confused and that it failed to achieve its
purposes. And there were equally high percentages of residual or "other" responses,
suggesting a widespread lack of agreement as to the very meaning of the policy.
TABLE I. PUBLISHED EXPRESSIONS OF MANUFACTURERS' ASSOCIATION

OF CONNECTICUT ON SELECTED ISSUES
Number of References Per Cent of

in Ten-year Period Favorable
(1934-40, 1946-48) References
i. Unspecified regulation 378 7.7%/O
2. Labor relations, general 297 0.0
3. Wages and hours 195 0.5
Total expressions, redistribution 870
4. Trade practices 119 13.8
5. Robinson-Patman 103 18.4
6. Anti-trust 72 26.4
7. Basing points 55 20.0
8. Fair-trade (Miller-Tydings) 69 45.5
Total expressions, regulation 4I8
SOURCE: Lane, Regulation of Businessmen, 38ff. The fig
ment is mine.
creteness and specificity of protectionist intere

more often and intensely than did the general
the liberal-traders (pp. I92-93). This was true in tariff, as they say,
because there the "structure of the communications system favored the
propagation of particular demands" (p. i9i). But there is also a struc-
ture of communications favoring generalized and ideological demands;
this structure consists of the peak associations (which were seen as
ineffective in tariffs-pp. 334, 335-36, 337-38, and 340); and it is highly
effective when the issues are generalizable. This is the case consistently
for redistributive issues, almost never for distributive issues, and only
seldom for regulatory issues.
As the pluralists would argue, there will be a vast array of organized
interests for any item on the policy agenda. But the relations among
the interests and between them and government vary, and the nature
of and conditions for this variation are what our political analyses
should be concerned with. Let us say, in brief, that on Monday night
the big associations meet in agreement and considerable cohesion on
"the problem of government," the income tax, the Welfare State. On
Tuesday, facing regulatory issues, the big associations break up into
their constituent trade and other specialized groups, each prepared to
deal with special problems in its own special ways, usually along
subject-matter lines. On Wednesday night still another fission takes
place as the pork barrel and the other forms of subsidy and policy
patronage come under consideration. The parent groups and "catalytic
groups" still exist, but by Wednesday night they have little identity. As
Bauer, Pool, and Dexter would say, they have prese

through overlapping memberships. They gain iden
they can define the issues in redistributive terms.
issues are more salient in sectoral or geographic o
the common or generalized factor will be lost in abstractness and
diffuseness. This is what happened to the liberal trade groups in the
tariff battles of the i950's, when "the protectionist position was more
firmly grounded in direct business considerations and . . . the liberal-
trade position fitted better with the ideology of the times.. ." (p. I50).
Where the peak associations, led by elements of Mr. Mills's power
elite, have reality, their resources and access are bound to affect power
relations. Owing to their stability and the impasse (or equilibrium) in
relations among broad classes of the entire society, the political structure
of the redistributive arena seems to be highly stabilized, virtually
institutionalized. Its stability, unlike that of the distributive arena,
derives from shared interests. But in contrast to the regulatory arena,
these shared interests are sufficiently stable and clear and consistent to
provide the foundation for ideologies. Table 2 summarizes the hypothe-
sized differences in political relationships drawn above.
Many of the other distinctive characteristics of this arena are related
to, perhaps follow from, the special role of the peak associations. The
cohesion of peak associations means that the special differences among
related but competing groups are likely to be settled long before the
policies reach the governmental agenda. In many respects the upper-
class directors perform the functions in the redistributive arena that
are performed by Congressional committees in the distributive arena
and by committees and Congress in the regulatory arena. But the dif-
ferences are crucial. In distributive policies there are as many "sides"
as there are tariff items, bridges and dams to be built, parcels of public
land to be given away or leased, and so on. And there are probably as
many elites as there are Congressional committees and subcommittees
which have jurisdiction over distributive policies. In redistribution,
there will never be more than two sides and the sides are clear, stable,
and consistent. Negotiation is possible, but only for the purpose of
strengthening or softening the impact of redistribution. And there is
probably one elite for each side. The elites do not correspond directly
to bourgeoisie and proletariat; they are better understood under Wal-
lace Sayre's designation of "money-providing" and "service-demand-
ing" groups. Nonetheless, the basis for coalition is broad, and it centers
around those individuals most respected and best known for worth
and wealth. If the top leaders did not know each other and develop
bo S-4 UU Uj
b 140- 4 7V~4- %4-4 V
bo -c ~ U 0 S
U U z 4-0 ~ ~ -
U -4 U 7~~~~~~~~~~~
~~~~ ~~~~~5 bo 4 )V ..
~~~~~ V~~~U, ) b m 5
~~ *V VO ~~~~~~~~~~~V V 4 "
VVU o -
0 bj
V - V 4 --0
04
.fl VV~~~~~~~~-
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4b
V ~ ~ I
H ~ ~ ~ ~~ ~ ~ ~ ~ .~~~~~ ~~~~~~4j-' -
V~~~~0
0 I 0.~~~~~~~~~~~~~~~~~~~~~~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~0 o 7
4- j cn --,
~~~~~~~~~ ~~~~~~~~~~~~~~~~~ _~~~~~~~~~~~~4 0 c
0 4~~~~~~~~~
bo 0 Ix00 '
0 -4
4 y -4 m U m U 04-'L~ S
common perspectives as a result of common schooling, as Mills would

argue, these commonalities could easily develop later in life because
the kinds of stakes involved in redistributive issues are always the
same. So institutionalized does the conflict become that governmental
bureaucracies themselves begin to reflect them, as do national party
leaders and Administrations. Finally, just as the nature of redistribu-
tive policies influences politics towards the centralization and stabili-
zation of conflict, so does it further influence the removal of decision-
making from Congress. A decentralized and bargaining Congress can
cumulate but it cannot balance, and redistributive policies require com-
plex balancing on a very large scale. What Riker has said of budget-
making applies here: ". . . legislative governments cannot endure a
budget. Its finances must be totted up by party leaders in the legisla-
ture itself. In a complex fiscal system, however, haphazard legislative
judgments cannot bring revenue into even rough alignment with sup-
ply. So budgeting is introduced-which transfers financial control to
the budget maker. .*.." Congress can provide exceptions to prin-
ciples and it can implement those principles with elaborate standards
of implementation as a condition for the concessions that money-
providers will make. But the makers of principles of redistribution
seem to be the holders of the "command posts."
None of this suggests a power elite such as Mills would have had
us believe existed, but it does suggest a type of stable and continual
conflict that can only be understood in class terms. The foundation
upon which the social-stratification and power-elite school rested,
especially when dealing with national power, was so conceptually
weak and empirically unsupported that its critics were led to err in
the opposite direction by denying the direct relevance of social and
institutional positions and the probability of stable decision-making
elites. But the relevance of that approach becomes stronger as the scope
of its application is reduced and as the standards for identifying the
scope are clarified. But this is equally true of the pluralist school and
of those approaches based on a "politics of this-or-that policy."
To date, no study of policy can equal American Business and Public
Policy in care, rigor, and exhaustiveness. But its very empirical superi-
ority tends to emphasize its theoretical weaknesses. Data of given
amount and quality become a richer and richer source as the context
of those data is better and better understood. It is just such a context
that I have attempted to sketch out here.
32 William H. Riker, Democracy in the United States (New York I953), 2i6.

Lowi - American Business, Public Policy, Case-Studies, and Political Theory

Uploaded by

Copyright:

Available Formats

Lowi - American Business, Public Policy, Case-Studies, and Political Theory

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lowi - American Business, Public Policy, Case-Studies, and Political Theory

Uploaded by

Copyright:

Available Formats

Trustees of Princeton University

Review: American Business, Public Policy, Case-Studies, and Political Theory

PUBLIC POLICY, CASE-STUDIES,

AND POLITICAL THEORY

Raymond A. Bauer, Ithiel de Sola Pool, and Lewis A. Dexter, American

C ASE-STUDIES of the policy-making process constitute one of the

formulate theories that will c

I. EXISTING NOnONS: THE NON-THEORIES OF POWER IN AMERICA

to pluralists) is maintained by the requirement

it seems clear that decisions

into a different type of coal

is an instrument of foreign policy and genera

their mail rather than to refle

DuPont were identified as "typical of large fir

is a model. Each is, if anything, a self-validating standpoint; the

II. ARENAS OF POWER: AN INTERPRETATIVE SCHEME FOR CASES

book, "The Politics of Foreign Trade," implies that everything they

which the empiricist is not suf

farther. Issues as such are too ephemeral; it is on

and group relations. What remains is to identify these arenas, to

AREAS OF POLICY DEFINED

(i) In the long run, all governmental policies may be considered

Once one posits the general tendency of these areas of

hypotheses are accumulated, th

ture." Indeed, coalitions must be built to pass legislation and "make

view. There are good and theoretically interesting

One of the most significant differences between the pluralists and

form a single policy-making elite. As a consequence, decision-making

Reed to recommit, with instructions to write in a provision that

of the debate was contained within the CES and

TABLE I. PUBLISHED EXPRESSIONS OF MANUFACTURERS' ASSOCIATION

Number of References Per Cent of

creteness and specificity of protectionist intere

Bauer, Pool, and Dexter would say, they have prese

b 140- 4 7V~4- %4-4 V

common perspectives as a result of common schooling, as Mills would

You might also like