On the Theory of Party Organization

Author(s): Joseph A. Schlesinger

Source: The Journal of Politics , May, 1984, Vol. 46, No. 2 (May, 1984), pp. 369-400
Published by: The University of Chicago Press on behalf of the Southern Political
Science Association

Stable URL: https://www.jstor.org/stable/2130967

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
https://about.jstor.org/terms

The University of Chicago Press and Southern Political Science Association are collaborating
with JSTOR to digitize, preserve and extend access to The Journal of Politics

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 On the
Theory
of
Party Organization

Joseph A. Schlesinger
Michigan State University

Assessments of the condition of our political parties reveal as much about the disarray in
our theory of parties as they do about the actual disarray in the parties themselves. This
paper argues that in the literature there exists a cumulative theory that explains most of what
we know about parties. At the level of party competition this has been laid down by Downs
and his successors. The rational-choice model, however, says little about party organization.
A theory of party organization can be developed from similar assumptions, thus providing a
realistic yet compatible theory of parties. The character of the political party emerges from
the fact that it combines three facets of organization in a unique manner: (1) the party main-
tains itself through market-exchange, making it similar to the business firm and thus unlike
the interest group or public bureau; (2) the party's output consists of collective goods, making
it similar to the interest group and public bureau yet unlike the business; and (3) the party
compensates its participants indirectly, making it similar to the interest group yet unlike the
bureau or business. Party is thus a structure that must be understood on its own terms. The
model explains why the Downsian hypothesis that parties put forth policies in order to gain
office rather than gain office to pursue policies is correct. It also explains why parties are
oligarchies and how parties maintain their organizations despite an inevitable sense of
dissatisfaction among many of their participants.

T he current concern for the condition of America's political parties

raises the uncomfortable question: just how well do we understand parties
in democratic politics? A nearly universal chorus of academics, jour-
nalists, and professional politicians tell us of the weakening, decline, and
decomposition of our parties. ' Their evidence seems compelling: a

I The journalist David Broder first noticed that "the party was over" in 1971. The decom-
position thesis has been developed by Walter Burnham (1970). Nie et al. (1976) have
developed the most influential evidence of the decline in party voting. Jack Dennis (1975)
has charted the decline in public support for the parties. A good statement is that of Pomper
(1977). Crotty and Jacobson (1980) have put together the most complete set of evidence in
various contexts. Today no analysis of any problem in contemporary politics appears com-

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 370 THE JOURNAL OF POLITICS, VOL. 46, 1984

declining proportion of citizens willing to identify with any party; a

growth in split-ticket voting; a severely reduced role in the nominating
process for party officials; the emergence of media specialists in the direc-
tion of electoral campaigns; the growth of political action groups outside
the regular party apparatus as a major source for campaign funds; and
the difficulties party leaders have in marshaling support for their policies.
At the same time, it is not difficult to amass equally compelling
evidence suggesting that today's parties are stronger than ever. The two
major parties, for example, dominate elective offices at all levels of
government to a degree unprecedented in American history.2 They also
compete over a wider range of constituencies than they have since before
the Civil War. Whereas Franklin Roosevelt could describe the
Democratic national headquarters of the 1920s as "two ladies occupying
one room in a Washington office building" (Key, 1964, p. 322), today
both national parties have sizable staffs. Twenty years ago most state
parties had no permanent office and unpaid chairs; today all that is
changed.3 Furthermore, whereas one could find little evidence in the

plete without some reference to how its solution is made more difficult by the decline of par-
ties. See, for example, Everson (1982) and Sundquist (1982).
2 Note in the figure that there has been only the merest trace of minor party or indepen-
dent strength in Congress in the last three decades. Almost all of that is accounted for by
former senators Harry Byrd, Jr. (Ind.) of Virginia and James Buckley (Cons.) of New York.

10

8 A Average Percentoge (by decodes) of Minor

7 \ | \ Party and Independent Members of Congress
6 (House and Senate)

4-

3-

0 0 0 0 0 , . . AC,
R Se' I' I' I'I' I' 0) 0)V A ' Qe 0) 0) 0)
3Compare the findings of Agranoff and Cooke (1964) or of Wiggins and Turk (1970) on
state party chairs with those of Huckshorn (1976). Cotter and colleagues (1980) have charted

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 THEORY OF PARTY ORGANIZATION 371

past of any systematic concern by the national apparatus for state and
local partisan affairs, today both parties maintain a close watch on and
are actively involved in such critical activities as the recruitment of can-
didates, campaign strategy, even the arcana of state legislative redistrict-
ing.4
If one assesses a party's condition by whether or not it stands for
anything, then it would be difficult to find earlier periods when the two
major parties were any more distinct in program and policy than they are
today. In the past twenty years the presidential candidacies of
Goldwater, McGovern, and Reagan hardly represent a muffling of par-
tisan choice. In the Congress, the erosion of the southern conservative
bloc within the Democratic party has given partisanship a sharper mean-
ing in that body. And the weakening of the powers of seniority in the
distribution of power in Congress has reduced the impact of what was
surely one of the most antipartisan rules under the old system.5
What then do we make of parties that win all the elections yet do not
control their nominations, parties that take distinct policy positions yet
whose leaders have little influence over their members, parties whose
organizations have decomposed or atrophied yet whose personnel and
payrolls have blossomed, parties whose support by the electorate has
declined yet which win more and more of the elections? If we look to the
literature we find little help. Students of parties have never even come to
an agreement on what a political party is, much less on how to tell
whether one is strong or weak, decaying or blossoming. The sole collec-
tive effort to state how parties ought to look, the 1950 report of the
American Political Science Association committee on parties, was roundly
criticized in its time, and is only occasionally revived today.6 The domi-
nant, textbook mode of analysis is what I would call the piecemeal ap-
proach or sequence of chapters dealing with what V. 0. Key, Jr. dubbed
the "party-in-the-electorate," the "party-in-government," and the "party
organization" (Key, 1964, pp. 163-65). Pieces are treated as though each
leads a life of its own with little attention to what if anything holds them

much of the growth of state party apparatus in recent years. Much of this is discussed in a
general context by Huckshorn and Bibby (1982). Evidences of change and renewal are in
Pomper (1981).
4 See, for example, Ehrenhalt (1983) on congressional campaign committees. The New
York Times of 26 June 1983 reports the launching of "Campaign '84," a program aimed at
strengthening the Democratic party at the state level.
5 The impact of congressional reforms on parties is discussed by Ornstein and Rohde
(1978).
6 Kirkpatrick (1971) reviews the history of the report and its critical reception. See
Fiorina (1980) for a recent call for some return to the notion of responsible parties.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
372 THE JOURNAL OF POLITICS, VOL. 46, 1984

together. Party is a lot of different things or, in Eldersveld's term, dif-

ferent "images" (Eldersveld, 1964).
This piecemeal approach worked, or rather was not obviously inade-
quate, as long as party-related matters did not change very much or
changed gradually. From the close of the Civil War until the 1950s,
American parties did not change markedly. The Civil War remained the
most dominant influence upon the system of competition between the two
parties. The regional pattern set by the war was reinforced after 1896;
even after 1932 it was critical to understanding American party politics.
The presidential parties underwent few structural changes in their pro-
cedures. The biggest change came in 1936 when the Democrats dropped
their two-thirds rule for nomination in national conventions. Within the
states and localities dominant officeholders had little difficulty in using
the direct primary and other reforms of the progressive era for their own
advantage. William Jennings Bryan or Theodore Roosevelt might have
found much to marvel at in the America of 1948, but they easily could
have understood the partisan structures and politics of the 1948 presiden-
tial campaign.
After a century of stability, we need to consider the shifts of the past
thirty years. Changes in communications, in the technology of organiza-
tion, in population mobility, and in political attitudes have transformed
each part of the sum of what we have called the parties. Lacking any
theory of what holds parties together or of how they might adapt to
change, lacking even an agreement on what a party is, the only observa-
tion we can make if our attention is riveted to the separate parts is to see
decline, weakening, and eventual disappearance.
To say that we lack a theory of party is to overstate the case. Rather, a
theory exists embedded in most of our writings on parties, but we seldom
see it as a whole. The situation in political writing today is similar to that
described over twenty years ago by Charles Lindblom in his review of
Downs's Economic Theory of Democracy (Lindblom, 1957). Lindblom
argued that political science, in contrast with his own field of economics,
appeared "thin" because political scientists refused to adopt the kind of
common approach to a subject which would assure the development of a
cumulative theory and findings. We cannot predict how a political
scientist might go about answering the question: "are parties declining?"
not because students of politics are incompetent, but because they have no
agreed upon way to approach such a question. Students of voting
behavior will answer the question one way, advocates of responsible par-
ties another, and students of party organization, depending on their
assumptions, in any of several ways. In contrast, Lindblom argued,
economists did have a guiding framework which, whatever the values or
approach employed, did direct their research. Lindblom was not argu-

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 373

ing then, and certainly would not now, that economists had developed a
solid body of truths. Tied to a common view of market forces, however,
economists of all persuasions have been able to develop a much more
cumulative discipline than have political scientists.
Lindblom went on to argue that embedded in the works of writers such
as E. E. Schattschneider, Pendleton Herring, David Truman, Robert
Dahl, as well as Austin Ranney and Willmoore Kendall, there was a com-
mon approach. What Downs had done was to draw much of this out,
demonstrate systematically that it did fit together, and explain much of
what was then known about party and electoral behavior. But Lindblom
was, I fear, overly sanguine in expecting that the Downs book would pro-
vide the impetus for the development of such a common approach in the
study of politics. Whatever influence Downs has had in fostering positive
or rational-choice theory and in stimulating the theory of competition
amongst candidates, his work has had little impact on the literature on
parties and party organization.
Lindblom's point is still valid. There does exist within the study of
politics a general framework for the study of parties, one which is used by
most students of parties, at least some of the time, if not self-consciously in
a systematic manner, then as a set of implicit assumptions. It underlies
common explanations of such phenomena as the impact of the electoral
college, the two-party system, the impact of reapportionment, and the
oligarchic character of parties.
In this paper I shall seek to lay out the parts of this framework. Most
parts are well recognized. What we ignore is that the parts fit logically
together. The theory of party competition as pioneered by Downs has
been the most systematically developed, and I shall say little about it
other than to draw from it our definition of party and the basic
hypothesis. More poorly developed is the theory of party organization
and its relation to the theory of competition. Since the development of a
cumulative theory of party rests upon our views of organization, and since
all of the discordant perceptions of what is happening to American parties
today flows around different views of organization, I shall devote the bulk
of this analysis to that problem.

THE DEFINITION OF PARTY

To theorize about parties requires a clear definition of what we mean

by party. Our inability to agree upon such a definition is surely the most
telling symptom of our malaise as we approach this subject. Part of the
problem arises from methodological conflicts such as the differences be-
tween functional and rational-choice approaches. Normative issues arise
over just what a party ought to be. Nothing I say here can resolve such

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
374 THE JOURNAL OF POLITICS, VOL. 46, 1984

conflicts. Yet, another part of the difficulty arises because many want to
define party in such a way to encompass the different kinds of political
organizations around the globe which call themselves parties.7 However
useful such a general theory of party would be, the theory I am discussing
is less ambitious. It seeks to explain only parties in freely working
democratic elections, and primarily those parties which are able over
time to win elections. The reason for this limitation is the assumption
that parties which do contest in free elections face an environment suffi-
ciently different from that of parties facing no such contest to justify a
distinctive theory. Similarly, parties which have no expectations of win-
ning elections cannot be assumed to be subject to the same tensions and
constraints as those parties which do.
Since our objective is to link together the theory of party competition
with that of party organization, we shall adopt the definition posited by
Downs in his theory of competition: a party is "a team seeking to control
the governing apparatus by gaining office in a duly constituted election"
(Downs, 1957, p. 25). Let us spell out some of the implications of that
definition.

IMPLICATIONS OF THE DEFINITION

First, the definition directs us to rational-choice explanations of party

behavior. In so doing it reflects both the view that politics is goal-
oriented and the assumption that political actors are rational in the ways
they seek to achieve their goals. Thus the definition directs us away from
psychological explanations, by which I mean a concern for the sources of
those goals or wants. Note also that the definition does not refer to the
functions that parties perform for the political system. Consequently,
those who prefer psychological or functional explanations of political
phenomena will be unsatisfied by this theory. However, the theory
neither denies that parties may perform broad political functions nor that
they respond to psychological needs. It rather asserts implicitly that if
parties do such things they do them as by-products of the pursuit of their
goals as stated in the definition.
Note also that in the manner of goal-oriented definitions and of their
accompanying rational-choice theories, the only elements sharply stated
are the goal (control of the governing apparatus) and the means of achiev-
ing it (winning office in a duly constituted election). The reason is that
rational-choice theories work best in the context of well-defined structures

7Alternative perceptions of party are explored by William E. Wright (1971). Kay

Lawson (1976, p. 3) defines party broadly to encompass governing groups that seek
authorization from the public, regardless of whether they face competitive elections. A
similarly wide net is cast by Kenneth Janda in his major survey (1980).

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 THEORY OF PARTY ORGANIZATION 375

such as markets or political institutions within which people can be

observed working to satisfy their wants. In effect, it is not the individual
who is defining the goal at all (that would throw us back on psychological
needs) but rather the institutions which are defining the goals. The self-
evident fact that people do make overt efforts to achieve such goals gives
us some confidence in our assumption that these are indeed what they
want. Similarly, neither the individual nor the party defines the basic
means of attaining the goals; rather it is the political structure through its
"duly constituted elections." Rational-choice theory, therefore, is really
a revival of the institutional approach in politics, and for that reason it is
compatible with much of the long tradition of the study of politics.8
The institutional basis implicit in this definition is another reason why
the theory cannot apply to parties which do not seek office through
democratic election. For such parties, neither goals nor means of attain-
ment are specified. In contrast, parties which do contest elections with
some hope of winning can be observed. Simple observation cannot tell
what other goals they may have, but this we do know: they are con-
strained by the institutionally defined goal of winning office and by the
institutionally defined mechanisms for attaining that goal.
On the other hand, the definition says nothing at all about what parties
look like other than to refer to them as teams. To be a team implies some
degree of deliberate cooperation among two or more individuals to
achieve some purpose. Since this is one classic definition of an organiza-
tion, we conclude that the political party is some kind of an organization
(Barnard, 1938, p. 4). Yet the character of that organization - how it is
arranged, what if any lines of authority it has, how disciplined it is, how
much division of labor exists - is not part of this definition of party. Such
omissions are important if the definition of party is to be flexible enough
to understand changes in the forms of organization.
Finally, the definition sets narrow yet imprecise limits for those in-
cluded within the party. They are narrow because only those who might
reasonably be considered part of the team are included; imprecise because
just who they are is not defined. The team surely includes the candidates
and those who actively work for their election as well as all officials who
join forces in government. Yet the definition excludes those who are
essentially choosers among competing parties, i.e., the voters. Between
candidates and voters lies a host of interest groups, media experts, and ac-
tivists of all sorts whose relationship to the party is problematical.
The exclusion of the voter from the party in this definition is a point
whose importance must be emphasized. Much of the difficulty political

8 The link between rational-choice theories and institutionalism is made by Rohde and
Shapsle (1978).

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
376 THE JOURNAL OF POLITICS, VOL. 46, 1984

scientists have had in developing a theory of party has come from not
knowing what to do with voters. In some measure this difficulty flows
from normative conflicts over just what the role of voters should be in a
democracy. Proponents of the participatory role see voters as integral to
any party formation, and thus they will inevitably be dissatisfied by this
definition. Others such as E. E. Schattschneider (1942), who see
democracy as based in competition among parties, see voters as choosers,
and thus as necessarily outside the parties. Again, nothing I say here will
resolve the issue. It is important however to recognize that this definition
reflects the latter perspective.
Beyond the normative issues, the position of the voter today poses
serious difficulties in our perceptions of and thus for our theory of party.
Systematic surveys permit us both to chart with great refinement the sorts
of people who vote for each party and also to probe their reasons. The
very richness of this data has had its impact on us as scholars and jour-
nalists; even politicians have come to draw their perspectives from survey
techniques rather than from the conditions and behavior of the parties
themselves. Thus it is common today to talk of parties as though they
were coalitions of voters, as in the terms the "New Deal coalition", or the
"party coalitions of the 1980s" (Lipset, 1978). Such coalitions are simply
constructs in the minds of observers who note that parties draw support
disproportionately from various population categories. None of these are
coalitions in the ordinary sense of conscious, explicit agreements by
members of these categories to pursue joint action. Indeed, if one ex-
amines the principal categories used in surveys - age, sex, income, educa-
tion, region, union membership, and religion (the exception is race) -we
find that none of these categories votes in blocs. A 60/40 split, or even
narrower divisions, reflects the size of the commitment of any of these
blocs to a particular coalition.9
More insidious for our theory of parties has been the psychological ap-
proach with its implication that party resides simply in the minds of
voters. In 1960 V. 0. Key, Jr., in a review of The American Voter (Key,
1960), expressed concerns about what he called the "politically relevant in
surveys." While Key doubted the relevance to politics of much of the
previous survey research he was more sanguine about the turn reflected in
The American Voter (Campbell et al., 1960). Certainly that work stands
as a landmark in our understanding of modern politics. Yet it did con-
tain the germ of an idea which has undermined our commonsense notions
of a political party, namely, the concept of party identification.
Initially party identification was useful as one of the basic attitudes

9 Ahramson, Aldrich, and Rohde (1982, Chap. 5) present a thorough analysis of how
assorted groups have voted in presidential elections since 1944.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 377

which, along with the voters' views toward the candidates and the issues,
produced the voting decision. In all of the early literature, including The
American Voter, there was no hint that voters who paid attention to the
candidates or to issues could not also be voting for a party. The members
of the triad of party, issues, and candidates were simply components of a
decision that necessarily was partisan for the simple reason that there is no
way that anyone could distinguish them in the real world of casting
ballots. More to the point, since it is the parties which nominate the can-
didates who take stands on issues, the distinction could never be between
partisan and nonpartisan factors but rather among different aspects of
partisan behavior, some of longstanding duration, and others of current
concern.
Yet as survey after survey has been run through the mills of secondary
analyses, each of the three components has taken on a life of its own. To
pay attention to a party's candidates or to be concerned with issues has
come to imply a lack of concern for party. In the widely cited book The
Changing American Voter (Nie, Verba, and Petrocik, 1976), the concept
of party is thus reduced to a nonrational sentiment residing in the minds
of voters. The authors chart the decline in what they call party voting as
the factor of party identification becomes less important in voting deci-
sions. Even the person who votes for all of a party's candidates all of the
time cannot, by their definition, be a party voter if that decision is based
primarily on the voter's judgments of the candidates or of their stands on
issues. Party has thus become a phenomenon completely separated from
the real world of party behavior.
If we would recapture the reality of party we must exclude the voter.
Voters are choosers among parties, not components of them. Surely the
reasons for voters' choices are vital to any understanding of party
behavior. It is a great advantage to a party to have large numbers of
voters inclined in its favor. Yet the ultimate test of a party's strength, at
least within the electoral context, lies in its ability to win elections, not in
its identification among the electorate. In a competitive party system a
weakening of partisan identification simply raises the level of competition
among the parties: what one party loses others must gain. Indeed the
growth of Republican strength in the Old South and that of Democrats in
former Republican bastions of the Midwest have depended precisely on
this erosion of partisan identification. This seeming paradox, parties
growing stronger while they are growing weaker, is a paradox only if we
make the error of equating partisan strength with partisan
identification. 10
The political party of our definition, then, is the Downsian team which

10 The strength of partisan identification may or may not have much to do with the stabil-

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
378 THE JOURNAL OF POLITICS, VOL. 46, 1984

seeks to control government by winning elective office. On this team are

those who seek and those who hold office in the party's name. Excluded
are those who see their role as essentially that of choosers among parties,
the voters. Yet the team must encompass many more people than just of-
ficeseekers as it seeks voters' support. Such people must be recruited and
their activities organized. Thus a theory of party must also be a theory of
party organization.
It is in this context that our theories are most diffuse. Rational-choice
theorists, because they have concentrated on candidate strategies, have
simply ignored the problem of organization." Students of party
organization, because they emphasize the complexities of motivation, see
the party of the definition as a gross simplification.12 The textbook or
piecemeal approach, on the other hand, treats the organization as though
it leads a life of its own, independent of officeholders, candidates, or even
campaign specialists. 13 We need to bring these parts together, to develop
a theory which both recognizes the unity of party and its organization and
makes similar assumptions about the motivations and decisions of all of
the actors within the party.

PARTY ORGANIZATION

The American Problem

It is one thing to assert the unity of party and organization, another to

demonstrate it. That is the task I have set in the remainder of this paper.
This is a peculiarly tricky matter in the American political system because
much of the activity aimed at winning office must take place outside the
confines of the formal party apparatus. This country is unique among
democracies in the extent to which it has sought to regulate and define
party organization. State laws have created a panoply of party offices

ity of democratic systems. For a discussion of the conflicting evidence on this issue see
Abramson, 1983, pp. 82-85.
11 Rational-choice theorists have paid some attention to the sequential character of deci-
sions in American parties, and of the need for candidates to appeal both to party followers
and to the general electorate. These discussions do not, for the most part, address the general
problem of party organization. See, for example, Coleman (1971), Aranson and Ordeshook
(1972), Wittman (1973, 1983), Schlesinger (1975), and Aldrich (1980).
12 See, for example, Eldersveld (1964, 1982), Wilson (1973), Schwartz and Rusnak (1982),
and Ware (1979).
13 Writers of texts on parties are fully aware that parties operate well beyond the bound-
aries of their formal apparatus. Yet they retain the term and idea of organization for that ap-
paratus. V. 0. Key (1964) set the tone. More recent texts have followed suit, e.g., Sorauf
(1976), Gelh and Palley (1975), Goodman (1980), Henderson (1976), Scott and Hrebenar
(1979), and Blank (1980).

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 THEORY OF PARTY ORGANIZATION 379

encompassing local, county, and state parties, most of which have little or
no mandate to take part in the nominating of candidates for any public
elective office, in running their campaigns, and certainly not in the ways
they operate in government. This is not to say that such officials are ir-
relevant to officeseeking or government. Rather in American politics
their relevance is an open matter determined in a multitude of ways.'4
The legal definition of the party apparatus has not prevented the
development of an array of efforts at seeking office outside its confines.'5
To equate the party organization with this official apparatus, therefore, is
to thrust into a never-never land of dis-organization much if not most of
the efforts put into winning office in the United States. We are left only
with discrete studies of such key elements in gaining office as recruitment,
the media, money gathering, and the like.
Rather than throwing up our hands in despair and asserting that parties
lack meaningful organization, we should reexamine our perspectives on
party organization itself. There is an advantage to looking at party
organization from the American perspective, where the formal structure
is obviously not the real organization. In countries such as Britain or
France, where parties have been free to organize as they see fit, there is a
less obvious disjunction between the formal party apparatus and the
operative organization. Nevertheless, a theory derived from the
American structure ought to be applicable to any democratic party
organization.
In order to perceive much of the diverse partisan activity in the United
States as taking place within an organization we must recognize the
distinctiveness of the party as an organization. While a party shares
characteristics with other types of organizations such as political interest
groups or voluntary associations, it also has characteristics which make it
very much unlike these groups. By the standards of any other kind of
organization a political party is bound to fall short. Just as we would
neither judge an interest group by the standards of a business, nor a public
bureau in the same way we would an interest group, our assessment of a
political party's organization must rest upon its peculiar characteristics.

The Model: The Properties of Organization

Our model rests upon the properties which characterize all organiza-

14 For example, Jewell and Olson (1982, pp. 112-20) examine the various procedures used
by formal structures to affect primary nominations. Huckshorn (1976) notes the varied roles
played by state party chairs.
15 Some of these extra-legal structures are formalized as clubs, associations, political action
committees, and the like. Informal coalitions have always existed. Contemporary financial
reporting laws have required many of these to give themselves titles and to keep records,
whereas in the past they operated more silently.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
380 THE JOURNAL OF POLITICS, VOL. 45, 1983

tions. Structural characteristics in the model such as the division of labor

and lines of authority are dependent upon how the organization combines
these properties. Each property reflects a solution to the basic problems
of all organizations: How does the organization maintain itself? What is
the principal work of the organization or what is its principal output?
How does the organization compensate its workers or participants? In
answering these questions I shall develop three dichotomous variables: (1)
maintenance of organization through market-exchange or through non-
market devices; (2) the primary output of the organization as a collective
or private good; and (3) compensation of participants either directly or in-
directly.
In the following discussion we shall see how each property of or answer
to the basic questions about organizations creates a force directing the
behavior of the organization. We shall look at each variable with respect
to the political party, comparing the party with other basic types of
organization: the business firm, the public agency or bureau, and the in-
terest group. We shall then examine the consequences for the party of
the way in which it answers each question. Once we have defined the ef-
fects of each variable, we shall be able to see the combination of proper-
ties which define the structure and behavior of political parties. We shall
then be able to test the model by looking at the consequences of altering
one or more of its assumptions.

Market- versus Nonmarket-Based Organizations

Our first distinction is that between organizations which do or do not

maintain themselves through market exchange. Market-based organiza-
tions trade their output or product with others in a direct, quid pro quo
exchange for money or other essential resources.'6I The business firm is
the prime example of such an organization. As long as it sells its product
for at least as much as it costs to produce, then the firm can maintain
itself. Many organizations, in contrast, are not market based. Thus the
salient characteristic of a public agency or bureau is that its maintenance
rests upon the political budgetary process. Some may, as in the case of
the post office or urban mass transit, sell their services in a market, but
the willingness of the state to make up their deficits shows that society
prefers not to let their services depend on market forces. Another
nonmarket-based organization is the political interest group. The group
may attract supporters who wish to receive the benefits sought by the
group, but it does not trade those policies directly for support. Interest

16 In this analysis I draw heavily on Anthony Downs's Inside Bureaucracy (1967). Downs's
propositions deal with bureaus. Since the political party is the antithesis of a bureau in all
respects (except for the fact that both produce collective benefits), I have turned Downs's
propositions around.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 381

groups may sustain themselves through side-payments by selling journals,

cookies, or calendars, yet no one considers these products to be their
primary output.
Is a political party market or nonmarket based? Elections are a type of
political market, in which parties offer their candidates and their policies
in exchange for the votes needed to gain office. In this market, parties
gain what is surely their key resource, control of public office. Just as a
business can maintain itself by selling its product at an adequate price, a
party able to win office has no difficulty in obtaining all the elements of a
vital organization: attractive candidates, willing workers, and money
givers. And, just as the economic market sends clear and unambiguous
messages to the business firm concerning the success or failure of its prod-
uct, the political market evaluates openly, automatically, externally, and
with exquisite numerical precision the output of the political party.
The political and economic markets are not, of course, identical. The
political market operates much more discontinuously in accord with the
electoral cycle. Much of the most important difference between the two
kinds of markets lies in the character of what is being traded. Elections
create an imbalanced market. Parties receive votes and the control of of-
fice. These are private benefits, or goods which go only to the party and
to its candidates. But the parties offer to voters candidates and policies,
benefits which go to everyone regardless of the votes cast. Thus they
offer collective benefits. This is in sharp contrast to the economic market
in which both sides to a transaction give and receive private or selective
goods.
Undoubtedly this imbalance in the political market is the reason we
have difficulty in recognizing the market character of elections. From
the voters' standpoint it is seldom clear just what it is that they are getting
from the exchange. The only well-defined consequence of the electoral
trade is the personality of the officeholder. Yet the president, senator, or
mayor becomes everyone's government and not just that of those who
"paid" for them. However, what the parties receive in this transaction is
always very clear. Since our concern is with a theory of party organiza-
tion and not of voter behavior, it is the clarity of the electoral market from
the party's standpoint that matters.
The fact that the parties' output consists of collective goods is of major
consequence to them as organizations. We shall consider that shortly. At
this point we must spell out the consequences for their organization of
their being market based.

Consequences of the market base - (a) goals. The major impact of the
market on organizations is that market goals dominate their decisions.
The reason is simply that success in the market is so closely tied to the

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 382 THE JOURNAL OF POLITICS, VOL. 46, 1984

maintenance of the organization. It is true that a major advance in the

theory of organizations has been the recognition of the complexity and in-
deed the ambiguity of organizational goals. 17 Even the business firm has
been shown to be less clearly driven by profits maximization than is
assumed by market theory. Nevertheless, that businesses operate in a
market provides them with a well-defined test of their success, a test that
reverberates throughout their organization. Whatever goals may com-
pete with market goals, a business cannot escape the fact that it must sell
its product to cover its costs. Failing that, the firm collapses and all of
the competing goals with it. Similarly, a political party which fails to
win election, and even more important, holds no prospects of winning in
the future, also faces collapse.
Beyond the kind of bottom-line control implicit in the market-based
organization, there are additional factors which lead to the dominance of
market goals. The clarity, lack of ambiguity, and automatic quality of
the market test also give its goals precedence over less clearly evaluated
goals or objectives. Whether a party has won an election and by how
much are matters of public record. In this sense a party is even more con-
trolled by its market than a business, for no amount of creative account-
ing can alter the size of a party's victory or defeat. The achievement of
none of the other goals of parties is ever as sharply defined: policies are
always a matter of more or less; patronage is inevitably quarrelsome. On
the proposition that clear and unambiguous marks of achievement
dominate those which are less clear, we can see that market goals gain
precedence.
Furthermore, office as well as money are neutral or value-free goals.
Both are instrumental to other objectives. They can be used for high or
low purposes. Herein lies the germ of rationalization which permits
politicians to retain their principles and ideals while adjusting to the elec-
toral market. As Alf Landon noted to the Republican convention of
1976, "The art of winning, of winning elections, comes before the art of
governing. "18
In contrast, organizations which are not market based are much less
tightly held to any particular goal. The two examples of nonmarket

17 Cyert and March (1963) demonstrate the complexities of decisions within business firms.
March and Olsen (1976), particularly with respect to academic bureaucracies, develop their
"garbage can" model of decision making. Schwartz and Rusnak (1982) apply this concept of
ambiguity to a political party organization. Given the inherent diffuseness of parties com-
pared with businesses or bureaucracies, my own view is that to approach them as ambiguous
structures is to pick up the stick from the wrong end. Despite their ambiguities and conflicts
the central question is how do they make decisions and persist?
18 Statement to the Republican National Convention, reported in the New York Times, 18
August 1976.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 383

organizations we have used - the bureau and the interest group - are free
to shape their goals in such a manner that the organization can always
make a claim for existence. The bottom line for a bureau is its budget,
and the bureau's task is one of convincing the political authorities of the
value of its actions. As the literature on policy analysis clearly indicates,
tests of success or failure are far from self-evident or automatic. '9 In the
same sense an interest group, however well defined its policy goals may
be, has substantial leeway in defining its objectives in such a way as to put
forth a strong claim for its existence. And should the group's goal actu-
ally be achieved, such groups are clearly capable of transforming their
goals to assure their continuity (Sills, 1957). The difference is evident. A
market-based organization cannot transform its goals and remain the
same kind of organization.
The difference between an interest group's and a party's ability to
define its goals can be illustrated by comparing a party with interest
groups operating in the electoral arena. In recent years there has been a
proliferation of electoral interest groups, such as NCPAC, which compile
hit lists of members of Congress. The targets of such groups are usually
highly vulnerable incumbents. As they go back to their supporters for
renewed contributions, the groups thus can usually claim a creditable
record of wins. But these groups can enter the electoral arena and choose
their goals in any way they wish. A true political party cannot do that. A
rational party will surely allocate its resources to achieve the maximum
impact, but it cannot measure its success by some limited, self-defined
goal. A party is tied tightly to the win and loss column provided by the
total electoral market.
In this sense the market is, to use Lindblom's imagery, a prison. Lind-
blom (1982) argues that the economic market imprisons all policymakers,
government officials, business personnel, and union leaders because all
are controlled by the market effects of their actions. In precisely the
same sense parties, as they nominate candidates and pursue policies, are
imprisoned by the electoral consequences of their actions.
Market goals thus dominate organizations sustained by a market. This
provides the rationale for the basic hypothesis in Downs's work: "Parties
formulate policies in order to win elections rather than win elections in
order to formulate policies" (1957, p. 28). Certainly there are conflicting
pressures within parties and even within the hearts and minds of can-
didates themselves. Yet because the imperatives of the political market
do not allow a party to treat all of these pressures equally, neither can a
theory of party organization treat them equally. Frequent elections and,
today, constant polling provide nearly continuous tests of the appeal of

"I On the difficulties of evaluating public policies see Wolf (1979).

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
384 THE JOURNAL OF POLITICS, VOL. 46, 1984

candidates, officeholders, and their policies. A party which does not re-
spond to the electoral market will by definition lose to parties which do,
and over the long run in a society where people are free to form new par-
ties, it will find itself supplanted by responsive parties.

Consequences of the market base - (b) influence within the organiza-

tion. The market also affects the pattern of influence within market-
based organizations. Because the market sends clear and unavoidable
signals about performance with respect to its particular goals, individuals
or units most responsible for market success can readily be identified.
Such individuals also gain influence. In a business, people who make
successful sales or who are associated with a successful product will rise to
power. Similarly for the political party the winning candidates, or even
those who look as though they might win, gain power. Indeed, because
the most clearly defined product of the party is its candidate, there is no
leeway for argument within the party as there is in business. Influence
within the party, therefore, will follow closely individual success and
failure in the electoral market.20
Note again how different are nonmarket-based organizations. Lack-
ing the clarity and unambiguousness of market-defined goals, interest
groups or bureaus must develop their own tests of success and failure.
Thus they are relatively free to distribute influence according to such non-
market criteria as seniority, loyalty, test'performance, birth, friendship,
or whatever according to the needs of the organization.

Collective Goods-Producing Organizations

In the previous section we saw how the mode of maintenance for an

organization (market or nonmarket base) affected goals and, through
them, the internal structure. We turn now to see how one type of
organizational goal-the output-affects the mode of maintenance for
the organization. The distinction we shall consider is between organiza-
tions producing collective goods and those producing private goods. The
private goods-producing organization poses few problems in this context.
It can use the private goods it receives (usually money) for its product as
the incentive to attract workers and other contributors. Its life and death

20 It follows from this argument that influence accrues to anyone in the party who is seen
as helping the party win elections. Losing candidates who do better than expected or those
who run in hopeless situations thus may actually gain in strength. Those who either give
money or who are good at gathering it, as well as those with desirable skills similarly can ex-
pect to exert influence (Kornberg, Smith, and Clarke, 1979, Chap. 7). Whatever influence
they have, however, derives not from their personal characteristics as such but from the fact
that others perceive them to be useful to the party in meeting its market-defined goal of gain-
ing office.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 THEORY OF PARTY ORGANIZATION 385

are of central concern to macroeconomic theory, but they are not of con-
cern to us in this context. Political parties along with interest groups and
public bureaus are organizations which produce collective goods.
Such organizations have always posed a problem in the theory of
organization. Since everyone receives the collective benefits, whether or
not they have done anything to achieve them, we are led to ask why
anyone supports these groups. Government agencies, of course, are sup-
ported by the coercion of taxing power. For parties and interest groups
which lack such coercive powers, the answer to their persistence is less ob-
vious.
There has always been, to be sure, the straightforward explanation: it
makes sense for people with common interests to band together in the pur-
suit of such interests. Yet that explanation has never taken account of the
fact that so few interests get organized. Even when organization does
form around a particular interest, few people who share that interest ac-
tually join. Finally, as has been widely observed, even among those who
join, few take an active part in pursuing their interest. Much of the
literature on political groups consists of efforts to resolve such questions.21
To a great extent the answers are given in nonrational terms; they are
rooted in expressions such as apathy, alienation, or a sense of inefficacy.
They arise, then, from assumptions incompatible with the rational-choice
theory of party implicit in our definition.
The theory which does explain the varied development of organizations
within the same rational framework is that of Mancur Olson. Olson has
demonstrated that in large groups, "rational self-interested individuals
will not act to achieve their common or group interests" (1965, p. 2).
Simply by voiding the apparently straightforward explanation that peo-
ple join organizations because of their collective purposes, Olson has
given us a theory that allows us to apply the same assumptions, perspec-
tives, and logic we commonly apply to private goods-producing organiza-
tions to organizations which produce collective benefits.
More important for our purposes, we have a theory of organization in-
herently compatible with the Downsian definition of party. The
theoretical gap between party organization and party behavior in com-
petitive elections has been closed. Olson's formulation allows us to
observe that in large groups, such as a political party, the increment in the
collective good (winning the election, achieving its policies) resulting
from a single participant's activity is so small that it will not be equal to or
greater than the cost of the effort to the individual, and therefore no ra-

21 The more penetrating analyses of these problems are in Truman (1951), and Schatt-
schneider (1960). A recent effort to put much of this together in light of evidence on the
development of a variety of groups is Walker (1983).

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
386 THE JOURNAL OF POLITICS, VOL. 46, 1984

tional person will participate to achieve the collective good. In other

words, the efforts of one individual will have such a small impact on the
achievement of the party's collective goods that, no matter how attached
to those goals, that individual will not contribute in order to achieve them
because the cost of the contribution will inevitably be worth more than its
marginal impact on the likelihood of achieving the goal. Since people
obviously do contribute to parties, in Olson's theory they must do so in
order to gain side payments, or private benefits they would have not
received had they not contributed.
Here we must reiterate an earlier point: rational-choice theories work
best in well-defined institutional settings where the benefits to be gained
are well marked and observable, and where the costs of achieving them
are easily calculated. Thus Olson's theory works best because it is most
predictive in economic groups like trade associations, labor unions, and
consumer organizations, where the participants can readily measure costs
and benefits in monetary terms.22 When we turn to groups such as par-
ties, in which neither benefits nor costs can readily be measured in the
same terms, Olson's theory is not invalidated. It is, however, less likely to
be completely predictive; that is because the ability to make cost-and-
benefit calculations is unevenly distributed in the society.
In order to behave as Olson predicts one needs information, first about
the costs of participating and second about the marginal effects of one's
contribution in producing the collective good. As Downs has stated, in-
formation itself is costly, and influence rests upon an uneven distribution
of information (Downs, 1957, chap. 6). In his latest work Olson notes
that information about costs and benefits of participation is itself a collec-
tive good, and thus is not likely to be supplied in optimal amounts (1982,
pp. 25-29). Furthermore, most of the free information available through
education and the media overstates the impact of individuals in the
political process. In imparting a sense of civic duty we spread the golden
lie exemplified in that old chestnut, "if only one more voter in every
precinct had voted for X, he would have won the election," and thus we
suppress the fact that there are thousands of precincts. In effect, the in-
formation needed to behave rationally in Olson's terms is not readily
available; for most people it must be gained from experience.
The theory thus predicts that those who do participate in political par-
ties in order to gain the collective benefits will have relatively little infor-
mation about the costs and benefits. These participants are the inex-
perienced and the young. With experience and therefore information
such individuals will drop out unless they find in political activity some
form of personal or noncollective satisfaction. The logic of collective ac-

22 Moe (1980) examines the Olson thesis.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 THEORY OF PARTY ORGANIZATION 387

tion thus predicts a high turnover in party personnel. Attractive can-

didates and new issues (collective goals) will bring in new activists, most
of whom fade away. Those who remain have found a sense of private
satisfaction. This seems to be a good description of most of what we
know about the fluctuations in partisan participation.23
Olson's logic removes one of the legs of the tripod of incentives con-
structed by Clark and Wilson (1961). Their distinction among material,
solidary, and purposive incentives, along with Wilson's elaboration in his
distinction between amateurs and professionals in politics, have given
direction to much of our research on party organization. Purposive in-
centives are of course collective. They are also central to the picture of
the amateur in politics, as well as to Wildavsky's picture of the "purists"
(1965). While not denying that these incentives do attract many ac-
tivists, the logic indicates that such incentives, unreinforced by private
benefits, have little staying power.
The incentive system in a party is thus complex. Leaving aside ambi-
tions for office, which clearly are vital to party development, the kinds of
private incentives range from material rewards of patronage through the
sense of satisfaction coming from involvement in great affairs. As in any
voluntary association, the sensitivities of people must be acknowledged to
prevent membership withdrawal. Because parties inherently must pit
people against each other in struggles over nominations and policies, it is
inevitable that many will feel injured. Thus parties share with other col-
lective goods-producing organizations all of the problems of incentives
exacerbated by personal and issue conflicts.
The logic of collective action further clarifies the phenomenon of
oligarchic control in political parties. Parties, along with other collective
goods-producing organizations in America such as professional associa-
tions, trade unions, and interest groups, fall into what David Truman
dubbed the "democratic mold" (1951, pp. 129-55). Since parties osten-
sibly attract people on the basis of their collective interests, the argument
that the members should govern these organizations is irresistible. Yet
the evidence points to a reality in which relatively few oligarchs or an "ac-
tive minority" run the organization. This is precisely what we should ex-
pect if the members are not there to attain the collective benefits. Attend-
ing meetings and running for office are costly, and will be done only by
those with a personal stake.
The weakness of collective benefits in providing the basis for party
organization is important to our understanding of how parties respond to
the electoral market. If, for example, policy goals were the prime incen-

23 On variations in participation see Eldersveld (1964), Verba and Nie (1972), and Korn-
berg, Smith, and Clarke (1970).

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
388 THE JOURNAL OF POLITICS, VOL. 46, 1984

tive for most participants, it is likely that they would seek to restrain the
party in its electoral maneuvering. That of course is precisely the argu-
ment of Wilson and Wildavsky. But since such collective incentives, un-
supported by private benefits, have a very short life, the party retains its
flexibility.
In a political party it is clear enough which people have the best de-
fined personal stake: those with ambitions for office. Their payoffs,
substantial and personal, are worth the costs of organization. Of-
ficeseekers thus are the entrepreneurs of party, which Salisbury (1969)
noted are essential for the development of collective goods-producing
organizations.

The Mode of Compensation of Participants

The third major distinction is that between organizations which com-

pensate their participants directly and those which do so indirectly. The
distinction is similar to that between voluntary and paid work. I prefer
the direct-indirect distinction first because much voluntary work for
charities, professional associations, chambers of commerce, interest
groups, and the like has its financial rewards in one's business or profes-
sion. Second, the distinction defines the degree of control through
payments that an organization can have over its work force. And third,
it establishes the extent to which that workforce is itself dependent for its
livelihood on the continued maintenance of the organization. These are
all factors of consequence for an organization.
Of the four types of organizations we have mentioned, it is the business
and the bureau which pay their work forces directly. Interest groups and
political parties, even though they may have some paid staff, depend
primarily on people whom they do not pay directly. In the political
party, even the most important leaders - its public officeholders - receive
their pay from the state and not the party.
The means of compensation defines the way participants behave within
any organization. Direct compensation permits a degree of rationaliza-
tion of activity denied to the organization which can compensate its
workers only indirectly. Direct compensation attracts people who show
up on time, who have skills designed for required tasks, and who are in-
different to and therefore unaffected in their work by the goals or output
of the organization. Indirect compensation, on the other hand, recruits
people from the leisure market or flexible occupations whose presence is
constantly at risk for the organization because of outside diversions. The
indirect-paying organization, therefore, must comply fully with its per-
sonnel from the time it asks them to work to the tasks it asks them to per-
form. Moreover its output must be acceptable to the participants. Disaf-
fection on the part of participants, whether due to the ways in which they

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 389

are treated or to the stands of the organization, poses a serious problem

for the maintenance of the indirect-paying organization.
Indirectly paid participants also have less of a stake in the maintenance
of an organization than do directly paid workers. This is of special
significance for political parties because it is not only the mass of activists
who are indirectly paid, but also the most important figures in the party,
the candidates and public officeholders. The latter, of course, are paid
by the state, and the party is important in gaining paid offices. But the
party cannot use that pay to direct the officials' behavior, nor can it deny
pay if the officeholders are elected without the party's support. This
distinguishes parties from interest groups which pay their leaders and
staff. Payment of staff is certainly a major factor leading to organiza-
tional continuity, even if that requires the transformation of the organiza-
tion's goals.

THE POLITICAL PARTY-A UNIQUE COMBINATION OF PROPERTIES

We are now in a position to bring together the three variables in our

model and look at the political party as a whole. The political party of-
fers collective benefits and compensates its participants indirectly, yet is
market based. From this unique combination of properties flows the
peculiar character of the party and its ability to adapt to a changing
world. Of course in the real world of politics pressures always exist
which tend to move parties, in one or another aspect, away from this
model. But the parties resist these changes. Before examining such
aberrations, let us first see where the model takes us in understanding
party organization.
The peculiar combination of properties creates in parties what, when
seen from the standpoint of other organizations, seem to be fatal
weaknesses. Compared with other market-based organizations, the
party is seriously deficient in its ability to discipline its participants in
order to meet the ever insistent market challenges. Compared with other
producers of collective goods, the party is incapable of adjusting its objec-
tives in order to maintain its tenuous supporters. It can neither narrow
nor broaden its objectives, muddy its purposes, nor alter its timing to
await the propitious moment to act.
These weaknesses are conspicuous to us because most of our experience
is with other types of organizations. We know that if these organizations
were to have the ways parties do they could not last. In the world of
business we recognize the importance of cost accounting, or efficiency ex-
perts, of the timeclock, and of hierarchic decision making. Those of us
who live in bureaus live with clearly defined structures, procedures, and
incentives. In our relationships with professional and voluntary associa-

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
390 THE JOURNAL OF POLITICS, VOL. 46, 1984

tions we see how and why they adjust their aims and activities to satisfy
and retain their members. Our perception of other organizations is so
clear that the deficiencies in parties become glaring. We see waste and
inefficiency in the use of resources. We see inactivity where activity is
called for on the organizational charts. We see parties time and again
angering their supporters, driving many away just when they seem to
need them. Party organization seems incomprehensible.

Because we fail to grasp the character of party organization, whenever

parties do appear effective we often seek explanations from nonparty
perspectives. Thus the corporate model leads us to explain effective
party organization by hidden hierarchies of bosses, machines, behind-the-
scenes operatives stalking the corridors of power. Since such phenomena
have been hard to find at the national level, a commonplace of political
science has been that our parties have no national organizations, except
perhaps for a brief period every four years when a president is nominated.
Under the intense gaze of modern journalism and social science today's
parties appear weak and debilitated; we look back nostalgically to a time
when these parties must have been better organized and surely better
loved by the citizenry. Otherwise, how could they ever have come into
being, prospered, and accomplished anything?
To understand parties, we must recognize that they do not perform and
adapt as do businesses, bureaus, or interest groups; nor can they be ex-
pected to do so, given their peculiar combination of organizational prop-
erties. Parties are perhaps best described as forms of organized trial and
error. Thousands of individuals and interests seek to control the party's
decisions. They push candidates, frame issues, recruit workers, make
alliances, and devise campaigns. Among these competing forces choices
are made, choices whose correctness is ultimately determined not by the
party but by the electorate. Nevertheless, it is the party organization
which assures that the right choices, i.e., those which win elections, are
retained and the wrong ones are rejected.

The right choices are not preordained but rest upon the peculiar com-
bination of properties delineated in our model. The market base and the
characteristics which flow from it (see fig. 1) assure successful electoral
candidates and their policies positions of strength within the party. As
well as collective benefits, they receive and control whatever private
benefits the party has to bestow. Those who support demonstrably losing
candidates and policies find no base in the party from which to exert in-
fluence. They may choose to retain their collective goals, but these will
be clearly insufficient to retain active participants.
In organizing through trial and error a party makes none of the usual
distinctions we find important in other organizations. It does not favor

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 391

UQ

z ~~~~~~~~~~~~~~~~~~C
0 zo~~~~~~~~~~~~~~
U kS 2
F-~~~~~0 <00c

~~' 'p ~~~~ w oet

ci a. c% tir
CO Q.~~~~~~~~~~tc

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
392 THE JOURNAL OF POLITICS, VOL. 46, 1984

old over new faces, professionals over amateurs, the loyal over the
disloyal, nor even leaders over followers. Trial by elections is a ruthless
test. Those who would maintain the appearance of power within a party
are thus well advised to follow the old Chicago maxim, "Don't back no
losers" (Rakove, 1975). Furthermore, such concepts as "effectiveness"
and organizational "strength" have no meaning in the party beyond the
electoral test.

ABERRATIONS FROM THE MODEL

The party we have outlined is, of course, a model or ideal type against
which we can place the variety of real world parties. The model
delineates the basic forces which operate within their organizations and it
is, of course, designed to encompass the loose structure of American
political parties. Yet the same forces operate within parties seeking office
in any democratic electoral system. It is not difficult to place American
parties in this framework because our parties have adopted so few of the
devices which might pull them away from the specifications of the model.
But even American parties are constantly being pulled in some manner to
deviate from this model. One way to test the strength of this model,
then, is to note some of these counterforces and examine how the parties
are constantly being pulled back by the forces implicit in our basic
variables. What happens if we alter each of the assumptions in the
model?

Party as producer of private goods. If we were to relax the assumption

that the parties' output consists of collective goods, then they would begin
to look more like businesses. They would produce policies designed to
help specific individuals and trade these for income. Since parties in-
evitably produce public policies, their ability to become purely private
goods-producing organizations is limited. After all, the national defense
cannot readily be distributed privately. Some aspects of it can, however,
be distributed privately in the form of defense contracts to specific com-
panies. Such contracts can be traded for campaign contributions.
Similarly ambassadorial appointments can be sold to the highest bidder.
Again, the line between private and collective benefits is not always
sharp. Thus a party may raise milk price supports, to take a hypothetical
example, in return for campaign contributions from the milk producers'
association. Indeed much of the picture we have of the old-fashioned
urban political machine makes it look very businesslike as it trades policies
for contributions.
From the party's standpoint, unfortunately, such exchanges are
generally perceived as corrupt. A political party which is widely seen as

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
 THEORY OF PARTY ORGANIZATION 393

selling policies as private benefits runs the serious danger of losing elec-
tions. Charges of bribery, corruption, and unethical behavior histori-
cally have been one of the most effective campaign appeals employed by
opposition parties. Thus the competitive electoral process is the major
constraint against this particular relaxation of the assumptions in the
model.

Party as a nonmarket-based organization. If we relax the assumption

that parties gain their key resources through market exchange, then,
depending on what sort of nonmarket mechanism it uses, we would ex-
pect the party to move in the direction of either the interest group or the
bureau. If, for example, a party were to emphasize consistency in policy
positions over winning elections, it would move away from the test of the
electoral market and thus need to develop other means of assuring con-
tinuity. In line with Olson's theory, the collective purposes of the group
would still be inadequate by themselves to maintain the party. It would
still need side payments of a material or solidary kind. Nevertheless, the
party would have to be far more concerned that it set its purposes in such
a way as not to repel its participants.

Much of the research on American parties over the past decade implies
that they are increasingly dominated by amateurs or purists, i.e., activists
less concerned with winning elections than with assuring that the parties
take the proper position on issues.24 Should such attitudes dominate the
decisions of the parties, then certainly we could say that the parties have
moved more in the direction of interest groups in the model. However,
the very looseness of the American party organization gives little leverage
to such individuals. The persistent complaints from right-wing conserv-
atives about the failures of the Reagan government are simple testimony
to the difficulties parties have in meeting such pressures.25 Indeed, the
significant changes in American political rules over the past decade have,
if anything, given policy-oriented activists less rather than more control.
For example, the expansion of the direct primary in the presidential
nominating process and the channeling of public financial support
through candidates have further enhanced the inherent candidate orien-
tation of American parties. Little in the American structure restrains
candidate, and thus party, maneuverability in the search for votes.
Another way to relax the assumption that parties are market based
would be to assure their support through some form of coercion, such as
taxation in the manner of a bureau. This is implicit in the public financ-

24 Examples of such research are Abramowitz, McGlennon, and Rapoport (1982), Roback
(1980), Soule and McGrath (1975), and Wildavsky (1965).
25 Note, for example, the article "Reagan 'Shift' Upsets Conservatives," New York Times,
26 June 1983.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
394 THE JOURNAL OF POLITICS, VOL. 46, 1984

ing of electoral campaigns. To a degree this has, of course, been ac-

complished with respect to the presidency and some state offices. If par-
ties were completely financed by the public budget, then our model
would lead us to expect more bureaucratic behavior from the group. This
would be particularly true if the budgetary allotments were unrelated to
the ability of parties to gain votes. As far as maintenance is concerned,
they would thus be freed from the electoral market. Political parties
totally supported by the state and enjoined from gaining additional sup-
port, including that of volunteers, would in the manner of bureaus be
more concerned with assuring continued budgetary support than in con-
testing elections. They would, in accordance with this model, be far less
responsive to the electorate.
Here again, we must note how resistant the American political system
has been to such developments. The greater portion of public support of
campaign finances has been not to the general party apparatus but rather
to individual candidates. Individual candidate organizations, tied to a
specific person, are not likely to generate bureaucratic structures.

Party as income provider. If we relax the assumption of our third

variable - that parties pay no one directly - then we move the party closer
to either the bureau or business. In American politics the picture painted
of the classic urban machine was that of workers who were in effect
employed by the party. They worked for the city, but these jobs were en-
tirely at the disposal of the machine and its boss. Of course the machine
also had to control elections. As a political monopoly, therefore, it was
not subject to market controls. Thus the bureaucratic image rather than
that of the business would seem to be more suitable. The evidence we
have of the ways patronage was distributed in Chicago (Wilson, 1961)
supports this contention.
Today, of course, vast amounts of money are spent by candidates and
parties. Does this challenge our assumption? Unpaid work still
dominates. Nevertheless there has emerged a large class of professionals
who make their livings from the electoral process. According to our
model, if these professionals were to become employees of the parties,
hired and fired by parties, promoted and rewarded by them, then a major
alteration in the organizations would be produced. It would inject into
partisan considerations a dimension which is now minor - the staffs' needs
and concerns for status, security, and even their attitudes towards
policies. The size of the party's budget and assurance of its continuity
and stability would take on a much higher priority.
If we examine how these large sums have been spent in American
politics we can see how resistant the party structure is to the development
of this type of professionalization. In part the accumulation of a staff in-

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 395

terest has been prevented by the fact that professionals have been hired by
individual candidates and officeholders. Of course, where officeholders
do have access to assured funding -as in the case of the Congress and in
some state legislatures - we have seen a sizable growth in paid staffing.
This is not, however, a unified staff; it is, rather, dependent on individual
officeholders. Along with this basic candidate orientation, the growth of
the campaign and polling professional has seen the establishment of a host
of private or nonparty campaign organizations. These employ their own
staffs. They operate nationally or regionally and contract their services
to assorted units of the parties. The floating contract workers, therefore,
do not alter the structure of the parties. Parties for the most part, then,
continue to operate with personnel who are indirectly compensated. This,
in turn, helps to perpetuate loose party organization.
The party organization of our model is thus highly resilient. It resists
the inevitable pressures to make it behave more like an interest group, a
business, or a bureau. The key to the party's resilience lies in its unique
combination of a market base with an output consisting of collective
benefits. This combination reverses the usual interpretation of how each
of these factors works. The officeseeking force within parties is generally
perceived as the flexible, compromising aspect; collective goods, or policy
goals, are usually seen as the less flexible, more ideological aspect. In this
model, however, it is the goal of attaining office which the party
organization is unable to alter. The electoral market inexorably tethers
the party to this objective. On the other hand, the party's collective pur-
poses provide little effective resistance because they do not provide the
long-run basis for attachment to the party.
As we noted at the outset, this model is embedded in much of our com-
mon understanding of how parties operate. Our central assumption is
that politicians will respond to their electoral demands and that the
organization of parties aids and abets them. This assumption is implicit
in the commonsense explanation of the recurrent conflicts between the
presidential and congressional wings of the parties, and in the differences
in behavior of the Senate and House. After all, different constituencies
produce different behavior. Our arguments about the virtues and defects
of the electoral college assume that presidential candidates and their par-
ties respond principally to the rules of election. Again we are not sur-
prised to find such one-time segregationist party leaders as George
Wallace and Strom Thurmond appealing to black voters, for the voting
rights act has altered the electoral market. Similarly, whatever merit the
single-member district explanation of the two-party system has rests on
the assumption that no one will persist in organizing and working for a
party which has no realistic chance of electoral success. When we assert
that the major parties undercut minor parties by adopting their popular

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
396 THE JOURNAL OF POLITICS, VOL. 46, 1984

issues we again assert the dominance of the office goal and the parties'
flexibility in adjusting their policies to it. These are all mundane ex-
amples, and all assume that parties behave precisely as called for in this
model.
This model of organization is central to a more general theory of
democratic parties, that which rests on the assumption that parties are
dominated by their office drives. This model demonstrates why that
assumption is correct. Based on general properties of organizations, the
model spells out none of the actual content or structure found in parties.
We can infer from it, however, that the real shape or content of parties
will reflect what candidates and officeholders see as useful to their elec-
tion and reelection. Thus the competitive conditions surrounding an of-
fice define how much actual organizational activity is worth the effort. It
explains why, for example, there was so little content to either
Democratic or Republican organization in the old, one-party South (Key,
1949) and why this has changed as the states have become more com-
petitive (Gibson et al., 1983). The actual structure of parties, or the flow
of contributions and acts of cooperation among a party's nuclei will
reflect the ways in which the opportunities for political advancement are
themselves structured in a country. This environment (Harmel and
Janda, 1982) or structure of political opportunities (Schlesinger, 1966)
varies from system to system, and indeed can change over time, as it has
been doing in the United States for the past thirty years.
Where, then, does this leave us with respect to an assessment of the con-
dition of American parties today? In the model there is but one standard,
a party's ability to win office. By that standard we must assert that the
two major parties have never been healthier. The weakening of regional
one-party bases due to the willingness of today's voters to give up party
identification and to consider the current activities of both parties-i.e.,
their candidates and perhaps even their policies-has given both parties
realistic expectations of winning offices unobtainable twenty years ago.
This has attracted to each party more candidates, more active workers,
and more financial support. It has sparked a greater concern among the
levels of party and has created more truly national party structures than
we have ever had. That it has unsettled politicians raised under the older
system is also clear. The weakening within Congress of the power of
seniority, that most antipartisan means of distributing power, has meant
that party leaders have had to learn other means of coordinating their
followers. Rather than seeing the first years of partisan influence under
the Reagan administration as an aberration, I would hazard that it is a
precursor, as presidents and party leaders learn the uses of this re-
juvenated force.
There are, of course, standards for judging parties other than their

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 397

ability to win elections. Nothing in the model ascertains how well the
parties represent the variety of interests in the society, how well they ar-
ticulate or moderate conflict, how well they are perceived by the elec-
torate, how successful they are in inducing popular participation, how
well they provide effective government, nor how responsible they are in
providing realistic alternatives. Just as winning office is simply in-
strumental and thus neutral to other political values, this model is also
neutral. If, however, it is a realistic description of how parties in
democracies work, then anyone who would alter these parties to make
them more desirable by some other standard must work from this model.

REFERENCES

Abramowitz, Alan, John McGlennon, and Ronald Rapoport (1982). "Incentives for Ac-
tivism in the 1980 Presidential Nominating Campaign." Paper prepared for the Annual
Meeting of the American Political Science Association, Denver.
Abramson, Paul R. (1983). Political Attitudes in America. San Francisco: W. H. Freeman.
Abramson, Paul R., John H. Aldrich, and David W. Rohde (1982). Change and Continuity
in the 1980 Elections. Washington, D.C.: Congressional Quarterly.
Agranoff, Robert, and Edward F. Cooke (1964). "Political Profile of State Party Chair-
men." Paper prepared for the Midwest Conference of Political Scientists, Madison, WI.
Aldrich, John H. (1980). "A Dynamic Model of Presidential Nomination Campaigns."
American Political Science Review 74: 651-69.
Aranson, Peter, and Peter Ordeshook (1972). "Spatial Strategies for Sequential Elections."
In Richard G. Niemi and Herbert F. Weisberg (eds.), Probability Models of Collective
Decision-Making. Columbus, OH: Charles E. Merrill.
Barnard, Chester I. (1938). The Functions of the Executive. Cambridge, MA: Harvard
University Press.
Blank, Robert H. (1980). Political Parties, an Introduction. Englewood Cliffs, NJ: Pren-
tice-Hall.
Broder, David (1971). The Party's Over. New York: Harper and Row.
Burnham, Walter D. (1970). Critical Elections and the Mainsprings of American Politics.
New York: W. W. Norton.
Campbell, Angus, Philip E. Converse, Warren R. Miller, and Donald E. Stokes (1960). The
American Voter. New York: John Wiley.
Clark, Peter B., and James Q. Wilson (1961). "Incentive Systems: A Theory of Organiza-
tions." Administrative Science Quarterly 6: 129-66.
Coleman, James S. (1971). "The Positions of Political Parties in Elections." In Richard G.
Niemi and Herbert F. Weisberg (eds.), Probability Models of Collective Decision-Making.
Columbus, OH: Charles E. Merrill.
Cotter, Cornelius, James Gibson, John Bibby, and Robert J. Huckshorn (1980). "State
Party Organizations and the Thesis of Party Decline." Paper prepared for the Annual
Meeting of the American Political Science Association, Washington, D.C.
Crotty, William J., and Gary C. Jacobson (1980). American Parties in Decline. Boston:
Little, Brown.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
398 THE JOURNAL OF POLITICS, VOL. 46, 1984

Cyert, Richard, and James G. March (1963). A Behavioral Theory of the Firm. Engle-
wood Cliffs, NJ: Prentice-Hall.
Dennis, Jack (1975). "Trends in Public Support for the American Party System." British
Journal of Political Science 5: 187-230.
Downs, Anthony (1957). An Economic Theory of Democracy. New York: Harper and
Brothers.
(1967). Inside Bureaucracy. Boston: Little, Brown.
Ehrenhalt, Alan (1983). "Campaign Committees, Focus of Party Revival." Congressional
Quarterly Weekly Report 41: 1345.
Eldersveld, Samuel J. (1964). Political Parties: A Behavioral Analysis. Chicago: Rand
McNally.
(1982). Political Parties in American Society. New York: Basic Books.
Everson, David H. (1982). "The Decline of Political Parties." In Gerald Benjamin (ed.),
The Communication Revolution in Politics. New York: Academy of Political Science.
Fiorina, Morris (1980). "The Decline of Collective Responsibility in American Politics."
Daedalus 109: 25-45.
Fishel, Jeff, ed. (1978). Parties and Elections in an Anti-Party Age. Bloomington: Indiana
University Press.
Gelb, Joyce, and Marian Lief Palley (1975). Tradition and Change in American Party
Politics. New York: Thomas Y. Crowell.
Gibson, James L., Cornelius P. Cotter, and John F. Bibby (1983). "Assessing Party Or-
ganizational Strength." American Journal of Political Science 27: 193-222.
Goodman, William (1980). The Party System in America. Englewood Cliffs, NJ: Prentice-
Hall.
Harmel, Robert, and Kenneth Janda (1982). Parties and Their Environments. New York:
Longmans.
Henderson, Gordon (1976). An Introduction to Political Parties. New York: Harper and
Row.
Huckshorn, Robert J. (1976). Party Leadership in the States. Amherst: University of
Massachusetts Press.
Huckshorn, Robert J., and John F. Bibby (1982). "State Parties in an Era of Political
Change." In Joel Fleishman (ed.), The Future of American Political Parties. Engle-
wood Cliffs, NJ: Prentice-Hall.
Janda, Kenneth (1980). Political Parties, A Cross-National Survey. New York: The Free
Press.
Jewell, Malcolm E., and David M. Olson (1982). American State Political Parties and Elec-
tions. Rev. ed. Homewood, IL: Dorsey.
Key, V. O., Jr. (1949). Southern Politics. New York: Alfred A. Knopf.
(1960). "The Politically Relevant in Surveys." Public Opinion Quarterly 24:
54-61.
(1964). Politics, Parties, and Pressure Groups. 5th ed. New York: Thomas Y.
Crowell.
Kirkpatrick, Evron M. (1971). "Toward a More Responsible Party System: Political Sci-
ence, Policy Science, or Pseudo-Science?" American Political Science Review 65: 965-90.
Kornberg, Allan, Joel Smith, and Harold D. Clarke (1970). Semi-Careers in Political Work:
The Dilemma of Party Organizations. Beverly Hills, CA: Sage Publishers.
(1979). Citizen-Politicians -Canada. Durham, NC: Carolina Academic Press.
Lawson, Kay (1976). The Comparative Study of Political Parties. New York: St. Martin's.
Lindblom, Charles E. (1957). "In Praise of Political Science." World Politics 9: 240-53.
(1982). "The Market as Prison." Journal of Politics 44: 324-36.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
THEORY OF PARTY ORGANIZATION 399

Lipset, Seymour M., ed. (1978). Emerging Coalitions in American Politics. San Francisco:
Institute for Contemporary Studies.
March, James G., and J. Olsen (1976). Ambiguity and Choice in Organizations. Bergen,
Norway: Universitetsforlaget.
Moe, Terry M. (1980). The Organization of Interests. Chicago: University of Chicago
Press.
Nie, N., S. Verba, and J. Petrocik (1976). The Changing American Voter. Cambridge,
MA: Harvard University Press.
Olson, Mancur (1965). The Logic of Collective Action. Cambridge, MA: Harvard Univer-
sity Press.
(1982). The Rise and Decline of Nations. New Haven, CT: Yale University
Press.
Ornstein, Norman J., and David W. Rohde (1978). "Political Parties and Congressional
Reform." In J. Fishel (ed.), Parties and Elections in an Anti-Party Age. Bloomington:
Indiana University Press.
Petrocik, John R. (1981). Party Coalitions: Realignment and the Decline of the New Deal
Party System. Chicago: University of Chicago Press.
Pomper, Gerald M. (1977). "The Decline of Party in American Elections." Political Sci-
ence Quarterly 92: 21-41.
(1978-79). "The Impact of The American Voter on Political Science." Political
Science Quarterly 93: 617-28.
, ed. (1981). Party Renewal in America. New York: Praeger.
Rakove, Milton (1975). Don't Make No Waves . .. Don't Back No Losers. Bloomington:
Indiana University Press.
Roback, Thomas H. (1980). "Motivation for Activism among Republican National Con-
vention Delegates: Continuity and Change, 1972-1976." Journal of Politics 42: 182-201.
Rohde, David, and Kenneth Shepsle (1978). "Taking Stock of Congressional Research: The
New Institutionalism." Paper prepared for the Meetings of the Midwest Political Science
Association, Chicago.
Salisbury, Robert H. (1969). "An Exchange Theory of Interest Groups." Midwest Journal
of Political Science 13: 1-32.
Schattschneider, E. E. (1942). Party Government. New York: Rinehart.
(1960). The Semi-Sovereign People. New York: Holt, Rinehart and Winston.
Schwartz, Mildred A., and Stephanie Rusnak (1982). "Competing for Power in the Repub-
lican Party of Illinois." Paper prepared for the Annual Meeting of the American Political
Science Association, Denver.
Schlesinger, Joseph A. (1964). "Political Party Organization." In James G. March (ed.),
Handbook of Organizations. Chicago: Rand McNally.
(1966). Ambition and Politics, Political Careers in the United States. Chicago:
Rand McNally.
(1975). "The Primary Goals of Political Parties: A Clarification of Positive
Theory." American Political Science Review 69: 840-49.
Scott, Ruth K., and Ronald J. Hrebenar (1979). Parties in Crisis. New York: John Wiley.
Sills, David L. (1957). The Volunteers: Means and Ends in a National Organization. Glen-
coe, IL: The Free Press.
Soule, John W., and Wilma E. McGrath (1975). "A Comparative Study of Presidential
Nominating Conventions: The Democrats of 1968 and 1972." American Journal of Politi-
cal Science 19: 501-17.
Sorauf, Frank (1976). Party Politics in America. 3d ed. Boston: Little, Brown.
Sundquist, James L. (1982). "Party Decay and the Capacity to Govern." In Joel Fleish-

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms
400 THE JOURNAL OF POLITICS, VOL. 46, 1984

man (ed.), The Future of American Political Parties. Englewood Cliffs, NJ: Prentice-
Hall.
Truman, David B. (1951). The Governmental Process. New York: Alfred A. Knopf.
Verba, Sidney, and Norman Nie (1972). Participation in America: Political Democracy
and Social Equality. New York: Harper and Row.
Walker, Jack L. (1983). "The Origins and Maintenance of Interest Groups in America."
The American Political Science Review 77: 390-406.
Ware, Alan (1979). The Logic of Party Democracy. New York: St. Martin's.
Wiggins, Charles W., and William L. Turk (1970). "State Party Chairmen: A Profile."
Western Political Quarterly 23: 321-32.
Wildavsky, Aaron (1965). "The Goldwater Phenomenon: Purists, Politicians, and the Two-
Party System." Review of Politics 27: 386-413.
Wilson, James Q. (1961). "The Economy of Patronage." Journal of Political Economy 69:
369-80.
(1973). Political Organizations. New York: Basic Books.
Wittman, Donald A. (1973). "Parties as Utility Maximizers." American Political Science
Review 67: 490-98.
(1983). "Candidate Motivation." American Political Science Review 77: 142-
57.
Wolf, Charles, Jr. (1979). "A Theory of Non-Market Failures." Public Interest 55: 114-33.
Wright, William E. (1971). "Comparative Party Models: Rational-Efficient and Party
Democracy." In William E. Wright (ed.), A Comparative Study of Party Organization.
Columbus, OH: Charles E. Merrill.

This content downloaded from

202.120.224.68 on Fri, 20 Jan 2023 14:24:13 UTC
All use subject to https://about.jstor.org/terms