Paper stones : a history of electoral socialism 0226684970, 9780226684970

Citation preview

A History of Electoral Socialism Adam Przeworski and John Sprague

ISBND-E2b-bflM^7-D

^B\PERr^

StoneS A

Adam

History ot Electoral Socialism

Przeworski and John Sprague

At the end of the nineteenth century, several

political

Europe entered into electoral politics with the goal of winning an overwhelming majority of votes for parties in

socialism. Elections were to open the “parliamentary

road” to socialism; they were to bring a “peaceful revolution.” Barricades

were no longer needed when work-

ers could cast ballots: votes political party has ever

program offering

were “paper stones.” Yet no

won an electoral

majority on a

a socialist transformation of society.

Adam Przeworski and John Sprague explain why. (Continued on back

flap)

Jacket illustration:

This allegory of the

famous Swiss

Battle of

Sempach was used on an electoral broadside by the

Swiss Democratic Social Party in the 1890’s.

Printed in U.S. A.

xsisz OMNIV/^

CIVIVAl ^aseecLO!

BOSTON PUBLIC library

Digitized by the Internet Archive in

2017 with funding from

China-America

Digital

Academic

Library

(CADAL)

https://archive.org/details/paperstoneshistoOOprze

„PAPER_

Stones

A History of Electoral Socialism

Adam Przeworski and John Sprague

The University of Chicago Press Chicago and London

Adam Przeworski

is

Martin A. Ryerson Distinguished Service Professor of

Political Science, University

of Chicago, and the author of Capitalism and Social

Democracy. John Sprague

is

University

Court and

in St. is

Louis.

He

co-author with

is

professor of political science, Washington the author of Voting Patterns of the U.S.

Adam

Supreme

Przeworski of Systems Analysis for Social

Scientists.

The University of Chicago

Press,

The University of Chicago

Press, Ltd.,

©

Chicago 60637

London

1986 by The University of Chicago

All rights reserved. Published 1986

Printed in the United States of

America

95 94 93 92 91 90 89 88 87 86 5 4 3 2

This project was supported

Foundation, to

SOC75- 17906

in part

1

by grants from the National Science

to the University

of Chicago and

Washington University.

Library of Congress Cataloging-in-Publication Data Przeworski,

Adam.

Paper stones.

Bibliography: p. Includes index. 1.

Elections

—

History.

3.

Socialist parties

11.

Title.

—

2.

Voting

History.

1.

—

History.

Sprague, John.

HI. Title: Electoral socialism.

JF1027.P79

1986

ISBN 0-226-68497-0

324.9

86-6984

SOC75-I7456

Contents

Prologue /

2

j

1

Electoral Participation and The Decision to Participate

Its

Consequences

13 13

Electoral Participation and Class Organization

18

Socialism and Electoralism

22

Great Expectations

25

The Dilemma of Electoral Socialism

29

The Magic Workers as

Barrier

29

a Minority

31

Allies

40

Party Strategy and Class Ideology

45

Party Strategy and Class Organization

52

The Dilemma of

55

Electoral Socialism

The Stagnation of the Socialist Vote

57

Electoral Trade-Off

57

Party Strategies and the Vote

79

Causes of Stagnation

88

Appendix

96

The

Contents

vi '

4

\

Are Socialist Leaders Vote Maximizers?

101

Choice and Necessity

101

The Realm of Choice Vote Maximizing Strategies

103

Was

109

the Stagnation of the Socialist

Vote Inevitable?

113

Are Socialist Leaders Vote Maximizers?

119

Appendix

126

Pure Strategies

Time Paths of

the Vote

127

Under Pure

Strategies

128

Carrying Capacities Associated with Pure 131

Strategies

Would

Different Assumptions

Make Much

Difference?

133

Vote Maximizing Strategies

j

The Voting Behavior of Individuals Conflicts

Over Class and

138

143

the Voting Behavior

of Individuals

143

Electoral History and the Act of Voting

147

Historical Patterns of Class Voting

158

Clinching Evidence

167

Epilogue

181

Appendix The Calculations The Data

187

Bibliography

203

Name Index

217

Subject Index

221

187

192

Prologue One should

modem

stress the

importance and significance which,

world, political parties have

in the

in the

elaboration and dif-

fusion of a conception of the world, because essentially what

they do

is

to these

work out

to

the ethics

and the

conceptions and act, as

it

politics

corresponding

were, as their historical

‘laboratory’.

Antonio Gramsci

No

political party

ever

won

an electoral majority on a program

offering a socialist transformation of society. At the end of the

nineteenth century, as socialist parties entered into the competi-

saw in universal suffrage an institution that working class to proceed from “political to so-

tion for votes, they

would allow the cial

emancipation,” as Marx had put

tions

were

were

to bring

to

it

fifty

years earlier. Elec-

open the “parliamentary road”

to socialism; they

about a “peaceful revolution” from a society based

would provide conditions for universal liberation. Barricades were no longer needed when workers could cast ballots: votes were “paper stones.” The syllogism was simple and persuasive. Since most people suffer from poverty and oppression inherent in the capitalist organization of society, and since elections are decided by numbers, socialism would become the electoral expression of an immense majority. Great masses would provide the mandate for legislating society into socialism. Socialism was the telos, universal suffrage on the exploitation of workers to one

was

to

be the instrument, and yet

it

that

has never happened,

at least

not thus far.

Involvement ties

were

in electoral politics

to establish roots

symbolic.

As long

among

was

inevitable

if socialist

workers. Nor could

it

par-

be merely

as participation in electoral competition

is

in-

Prologue

2

strumental for improving the conditions of workers in the short

mass following must avail itself of this opportunity. In spite of the distrust which the plunge into electoral politics often evoked among socialists, abstention was never a feasible option. Workers did not become organized as a political party everywhere; but wherever they can, working-

run, any political party that seeks a

class parties

do

participate in electoral competition.

Observers often saw dilemmas in

any of the institutions of

“Whatever seeks

in the

involvement of workers

capitalist society, including elections.

to

extend

—

this

itself

under domination runs the

risk

would be the fate of working-class organizations under capitalism (Horkheimer 1973:5). In order to real-

of reproducing it”

ize the revolutionary goal

class parties

must

avail themselves of the opportunities that exist

in capitalist society.

tion

is

of bringing about socialism, working-

But to the extent to which electoral participa-

instrumental in improving the condition of workers within

the confines of capitalism, socialism sary. Socialism

is

no longer urgently neces-

cannot be achieved without participation

cratic institutions, but participation

in

demo-

erodes the will for socialism:

bemoaned dilemma of democratic socialism. Yet the protagonists socialist leaders who marched their parties onto electoral battlefields never saw their choice as a dilemma. They entered electoral politics with the goal of winning an overwhelming popular mandate for socialism, even if they were comthis is the frequently

—

—

pelled to enter by the need to improve the immediate conditions of

knew

workers. They entered with ultimate goals and they

they

would win. This divergence between cause and purpose of rationalization. Yet the crucial question

is

is

perhaps a sign

not about the moti-

vations of party leaders but about the effect of electoral participation

on the movement for socialism.

It

is

a

question about

conditions which are independent of anyone’s will, about the structure of the situation in in a

which a

socialist

movement

finds itself

democratic capitalist society.

The decision

to participate

is

but a prologue to the history of

socialism, but prologues delimit the entire play. socialist parties

electoral

elections

Once

leaders of

decided to enter into electoral competition, the

system structured their future choices. To be effective

—

for

whatever goals

—

a

party must win

votes,

in

and

Prologue

3

votes are measured in numbers.

Hence

the parties that organize workers

suoport elsewhere

the perpetual issue facing

whether or not

is

to seek elec-

in the society.

Leaders of socialist parties

must repeatedly decide whether or not

to seek electoral success at

toral

the cost, or at least the risk, of diluting class lines

and conse-

quently diminishing the salience of class as a motive for the politi-

behavior of workers themselves.

cal

Contrary to the enthusiastic expectations with which socialist leaders initially joined the

who “had

letarians

game of

nothing to

elections, workers

sell

“nothing to lose but their chains”

—

the pro-

power” and

but their labor

— never became

a numerical

majority in any society. Hence the electoral mandate for socialism

could not be obtained from workers alone. Democratic institm tions played a perverse trick with socialist intentions

—

the

eman-

cipation of the working classes could not be the task of workers

themselves

if this

emancipation was to be accomplished through

elections.

Given the minority

status of workers, leaders of class-based

must choose between a party homogeneous

parties

in

its

class

appeal but sentenced to perpetual electoral defeats or a party that struggles for electoral success at the cost of diluting entation. This labor,

ocratic,

is

its

class ori-

the alternative presented to socialist, social

communist, and other

combination of class structure and

parties

dem-

by the particular

political institutions

found

in

democratic capitalist societies. This choice priori

reason

dence

to

is

not between revolution and reform. There

— pace

Lukacs (1971:60)

— and

no

is

no

a

historical evi-

suppose that a minority, class-pure, electoral party of

workers would be any more revolutionary than a majority party heterogeneous ties

in its class basis.

Indeed, class-pure, electoral par-

of workers, of which the Social Democratic Party of Germany

(SPD) during the Weimar period

is

perhaps the prime example,

can be committed exclusively to the defense of corporate interests of workers within the confines of capitalism.

A

pure party of

workers constituting a majority of the electorate could have per-

commitments without compromise, as sothey would when they saw the working class as

haps maintained cialists

said

its

majoritarian. But to continue as a minority party dedicated ex-

clusively to ultimate goals in a system which requires a major-

4

ity

Prologue

— more,

these goals

produces

an overwhelming mandate

—

to

would have been absurd. Keeping

To

a party class pure

of guardians of the eternal flame (Schum-

at best a sect

peter 1942).

attempt to realize

gain electoral influence for whatever aims, from

the ultimate to the

most immediate, working class parties had

to

seek support from other people, to enter into alliances, and to

make compromises. The

decision to participate in elections inel-

uctably altered the logic of revolutionary transformations.

A

ma-

composed of workers could not provide the mandate for socialism because workers never became a majority. The only question was whether a majority for socialism, and indeed for any more proximate goals, could be obtained by seeking electoral sup-

jority

port

among people

other than workers.

With the support of workers alone or of the people electoral majorities turned out to be an elusive goal.

in general,

No party won

an electoral majority on a program offering a socialist transformation of society, but very

few won majorities with any program.

Elections rarely result in majorities: about one election in fifteen yields the majority of votes cast to a single party.

And

there

is

no

magic number of 50 percent. Why has no party ever obtained an overwhelming mandate in a free election, reason to stop

for anything?

at the

How

does

happen

it

that

no party has ever obtained

the support of one-half of those entitled to vote?

When,

at the

beginning of the century, socialist leaders witnessed the spectacular

growth of

ing majority elections.

seemed

And

winning an overwhelm-

their electoral strength,

yet

just

all

a

matter of time, indeed of a few

extrapolations remained frustrated

down

—

so-

some share of the vote typically inferior to 50 percent. All growth was arrested as it approached 50 percent, almost as if electoral institutions were designed in a way cialist parties settled

that

would prevent any

to

political force

from obtaining overwhelm-

ing support for any social transformation.

What kind of system is it that socialists got themselves into? Can parties that appeal to workers win an overwhelming majority of votes given the class structure of industrialized capitalist societies? Is there an inherent

dilemma

that

date for socialism impossible to obtain?

makes an overwhelming victory of any gram, impossible

?

What

is

makes an Is this a

mandilemma that

electoral

party, behind any pro-

the range of choice available to so-

Prologue

5

Have they exploited them?

cialist parties?

available to

To answer the way they

the opportunities historically

we need do know much

these questions

do.

We

know why people

to

vote

about voting; few social

phenomena have been studied as extensively as the voting behavior of individuals. With the aid of surveys, social scientists have accumulated detailed information about the voting patterns of persons located variously in the social structure and characterized by

and preferences. While

differing opinions, attitudes,

France small farmers are more likely to vote Left, agricultural laborers are

worker

more prone

he works

if

his father

in a large factory,

would be surprised Catholic

do

so.

A

southern

in Italy landless

Swedish manual

almost certain to vote for the Social Democrats

is

graduated from high school, er, if

to

in

widow

and

is

was

anywhere but

to find

voting socialist.

empirical knowledge, the reasons

in

And yet, why these

he

manual work-

also a

a union

if

member. But we

Austria an elderly

in spite

of this rich

people vote the

way

they do remain unclear.

The

first

surveys of voting were conducted with the hope that a

few factors would turn out rection of voting, that their

to

have a dominant impact on the

list

would be

short,

di-

and the explana-

Appendix A to Voting (Berelson et al. 1954) to see what happened to this hope over two hundred factors had some bearing on the vote. It seems as if everything had some relation to the way people vote and with tion almost complete.

Yet

it is

sufficient to look at

nothing being truly decisive.

Most hopes were pinned on class. Class is indeed one factor for which there are good reasons to expect that it should shape patterns of voting. “A relation between class position and voting behavior,” a typical argument ran, “is a natural and expected association in the Western democracies for a

number of

reasons:

the existence of class interests, the representation of these interests

by

political parties,

and the regular association of certain

parties with certain interests. tion order

and the way

different class interests,

Given the character of

the stratifica-

political parties act as representatives of it

would be remarkable

if

such a relation

were not found” (Alford 1967:68-69; also Alford 1963). And

was often found,

for

example by

Lipset:

“The

it

simplest explana-

Prologue

6

widespread pattern

tion of this

The

leftist parties

change

order to

in

of equality; the lower-income groups sup-

become economically

higher-income groups oppose them

nomic advantages. The

in

better off, while the

order to maintain their eco-

statistical fact

dence of the importance of class If societies

simple economic self-interest.

represent themselves as instruments of social

in the direction

them

port

is

can then be taken as evi-

in political

behavior”

are differentiated in terms of class,

if

located in the class structure share interests, and

(

1963:29).

persons similarly if

parties differ in

representing these interests, then indeed people can be expected to vote on the basis of their class position.

Many

people do. Survey studies indicate that between 60 and

80 percent of the voters

in

Western Europe and the Anglo-Saxon

countries cast their ballots in a tion in the class structure.

whelming. tives,

A

manner consistent with

But these patterns are

far

their loca-

from over-

large part of British workers choose the Conserva-

supposedly out of deference for their inherited

skills in

han-

complex matters of governance. Women generally tend to vote for conservative parties more frequently than men of the same class. Religious persons often express their confessional dling the

identification at the polls. Ethnic, linguistic, racial, or regional identification leads to voting that cuts across class divisions. In-

deed, class

some observers have concluded is

the

that religion rather than

most important determinant of voting (Rose and Urwin

1969). Several students of voting behavior maintain that tification

is

iden-

with parties rather than with any social groups that moti-

vates voting decisions, arguing that there are

reasons

it

why

good psychological

class should not matter (Converse 1958). Still others

see voting not as a matter of expressing identifies or prior

commit-

ments but as calculated decisions made independently by

indi-

viduals for each election.

Most of these explanations

—

in

terms of class, age, sex, edu-

cation, religion, ethnicity, party identification, or region the

same

logic.

They

cite individual attributes to explain indi-

vidual acts. In this logic tified as

— have

some

social distinctions are first objec-

cleavages. Places in these cleavages are attributed to

individuals, and after this reduction the locations in the structure

of society appear as individual

woman. These

traits are

traits,

such as worker. Catholic, or

thought to determine acts because they

Prologue

are

7

viewed as representing

and the

logical attachments,

Through

this

like.

reduction the locus of causality

each individual. Social relations are reduced to tions of individual traits.

havior

mode

in this

A

class

is

more

is

placed within

statistical distribu-

successful explanation of voting be-

of analysis consists of finding a

covariation between individual traits and acts. sufficient,

norms, psycho-

interests, internalized

If

one

statistical trait is

not

are brought to bear the weight of explanation. If

not sufficient,

it

must be because religion also matters:

“Part of the explanation of these deviations has already been pointed out; other characteristics and group affiliations such as

more salient in particular situations than high and economic position” (Lipset 1963:240).

religious belief are

or low social

But ior in

why

is

some

class important in

molding individual voting behav-

societies but not in others, during

Why

not during others?

some periods

do individuals sometimes vote on the

sometimes on the basis of confessional attach-

basis of class,

ments, sometimes because of loyalties to parties, and

at

because of the overwhelming charm of someone’s grin?

Why

Norwegians more French?

Why

but

are

likely to vote

times are

on the basis of class than the

Swedish workers more prone

to vote for the

Social Democratic Party today than they were sixty years ago?

These are not questions about individuals. For even vidual acts tend to coincide with individual traits, lar traits

become causes of

why do

if

indi-

particu-

individual acts? In nature, causes are

causes and no one can do anything to alter them. The causal struc-

ety

is

people

not.

world

is

given, but the causal structure of soci-

The causes of

individual behavior are produced by

ture of the natural

in interaction

with one another.

Reduction does not suffice as an explanation because the causal path from individual traits to individual acts passes through the totality of social relations.

Gramsci put

it,

“is the final

(1971:193). This

is

“The counting of

ceremony of

'

‘votes’,” as

a long process.

...” j

images of society, of

a process of creating

j

forging collective identities, of mobilizing

commitments

to partic-

j

ular projects for the future. Class, religion, ethnicity, race, or |

nation

do not happen spontaneously

as reflections of objective |

conditions in

the psyches of individuals. Collective identity,

group solidarity, and

political

commitment

are continually trans-

j

^

Prologue

8

formed

— shaped,

destroyed, and molded

conflicts in the course of

which

anew

—

as a result of

political parties, schools, unions,

churches, newspapers, armies, and corporations strive to impose a particular relation

form of organization upon the

of society. The

life

between places occupied by individuals

their acts

is

a contingent historical

front interests

and images,

outcome of struggles

that involve preferences

and defeats. The

that bring victories

political

and

in society

that con-

and strategies,

behavior of indi-

viduals can be understood only in concrete historical articulation

with these conflicts acts

when

they are

—

particular traits

become causes of individual

embedded within

a definite structure that has

been imposed upon

political

relations

at

a given

moment

in

history.

The organization of There

is

politics in terms

nothing inherent

history that

would make

in

of class

is

not inevitable.

capitalism and nothing in the logic of

the

emergence of classes

as collective

subjects inexorable. Class position structures daily experience,

generates a certain kind of knowledge, terests,

endows people with

and may even evoke an understanding of a shared

feeling of solidarity. But this experience

in-

lot, a

need not become collec-

As Marx said and Gramsci was fond of repeating, individuals become conscious of social relations in the realm of ideology, people become aware of conflicts of interests tivized as

one of

at the level

class.

of ideology. “It

is

not the simple existence of op-

pressive conditions,” Michels points out, “but tion

it

of these conditions by the oppressed, which

is

the recogni-

in the

course of

history constituted the prime factor of class struggle” (1962:228).

Class relations are not transparent at the level of the “immediate”

(Gramsci), the “lived” (Althusser 1971) experience

ence which

ence

may

is

simply a reflection on everyday

life.

—

the experi-

This experi-

be one of poverty, of compulsion, of inequality, of

oppression.

It

may

be one of similarity. But

is

not an experience

solidarity, as

Olson (1971) has

it

of class.

Even

similarity

need not breed

shown. Indeed, Marx and Engels noted ifesto that

in the

Communist Man-

“the organization of proletariat into a class, and conse-

quently into a political party,

is

continually being upset by the

among workers themselves” similar situation and who have

who

competition

(1967: 144). People

are in a

identical interests often

find themselves in competition with

one another.

Prologue

9

Thus

the division of society into classes does not necessarily

result in the organization of politics in

terms of class. The “sys-

tem of interests” need not become the “system of solidarities” (Pizzomo 1966). “The simple objective conditions of producers,” Sartre emphasized,

“defines the concrete

man

—

his

needs, his vital problems, the orientation of his thought, the nature of his relationships with others:

belonging to a class” (1958:96). Nor as Sartre supposed, since

it

is

does not determine his

it

is this

experience as simple

not the only objective experience.

“objective” means experience that

is

inherited

by individuals and

independent of their will, then being a Catholic jective experience, as

woman

is

being a black

in the

If

in Italy is

an ob-

United States, or a

The people who perpetuate their existence by selling their capacity to work are also men or women. Catholics or Protestants, Northerners or Southerners. They are consumers, taxpayers, parents, and city dwellers. They may be mobilized into politics as workers, but they may also become Catholic in Switzerland.

workers. Catholics, or Bavarian Catholics. Hence,

we have

wealth of information

Our

way

of the

today about patterns of individual

voting, the explanatory question remains open:

vote the

in spite

why do

people

they do?

central thesis

is

that the voting behavior of individuals

More

an effect of the activities of political parties.

is

precisely, the

relative salience of class as a determinant of individual voting

behavior

is

a cumulative consequence of the strategies pursued by

political parties

of the Left. Hence,

not of voters, that

nomic and

is

this is a

study of voting, but

predicated on a theory of the system of eco-

political organization

under which

political parties de-

velop their strategies and individuals cast their votes.

The organization of

politics in terms of class

tingent result of conflicts in the course of

is

always a con-

which multiple

political

forces strive to maintain or to alter existing social relations.

Along

with other organizations, political parties forge collective identities, instill

commitments, define the

collective actions

and deny them.

Is

interests

on behalf of which

become possible, offer choices to individuals society composed of classes or individuals with

harmonious interests? Are class relations the fundamental social conflicts or are they subordinate to

What

are the classes?

Are

some

line

of

other divisions?

class interests antagonistic or

do they

encourage cooperation? Which classes represent interests more

l4.

/

I Electoral Participation

and

Its

Consequences we do pretender’s sham

Republicans before everything,

not indulge the crazy

idea of appealing to a

prestige or a dictator’s

sword

to secure the

universal suffrage.

triumph of our doctrine. It

is

the voter

We

whom we

appeal only to

want

to set eco-

nomically and politically free.

Alexandre Millerand,

in

1902

The Decision to Participate The

crucial choice

was whether or not

to participate.

Earlier

events resulted in establishing the principle of democracy in the political realm.

Yet

political rights

were merely formal when ac-

companied by the compulsion and inequality that reigned in the social realm. As it emerged around 1850, socialism was thus a

movement

would complete the revolution started by the bourgeoisie by wresting from it social power just as the bourgeoisie had conquered political power. The recurrent theme of the socialist

that

movement ever

since has been this notion of extending the

democratic principle from the political to the social,

in effect pri-

marily economic, realm.

Yet precisely because the principle of democracy was already

means by which socialism choice. The project of the early,

present in the political institutions, the

could be achieved appeared as a

communitarian a

socialists

was

to build a society within the society,

community of immediate producers associated

in

workshops

and manufactures, cooperating as consumers and administering their

own

affairs.

This society of associated producers was to be

Portions of this chapter

1980)

in a

somewhat

first

appeared

different form.

in the

New

Left Review, no. 122 (July

Electoral Participation

16

He wrote

one.

who

Its

Consequences

1879 that “a historical development can only

in

remain ‘peaceful’ so long as those

and

wield power

it

in

is

not opposed by the violence of

society

time” (McLellan

that

at

1971:201).

Thus,

the

essential

question

facing

whether, as Hjalmar Branting posed

it

its

it

was

parties

1886, ‘‘the upper class

in

[would] respect popular will even when

of

socialist

demanded

the abolition

privileges'" (Tingsten 1973:361). Sterky, the leader of the

wing of the Swedish Social Democrats, was among those who took a clearly negative view. ‘‘Suppose that ... the working left

class could send a majority to the legislature; not

even by doing

would it obtain power. One can be sure that the capitalist class would then take care not to continue along a parliamentary this

No

course but instead resort to bayonets” (Tingsten 1973:361).

one could be completely certain. Austrian Socialists, for example,

promised

in their

Linz program of 1926 to ‘‘govern

in strict accor-

dance with the rules of the democratic state,” but they pelled

warn

to

‘‘should

that

the

bourgeoisie

felt

com-

by boycotting

revolutionary forces attempt to obstruct the social change which the labour

movement

in

assuming power

then social democracy will be forced to to

pledged to carry out,

is

employ

means

dictatorial

break such resistance” (Leser 1976:145). The main doubt

about electoral participation was whether revolution would not be necessary anyway, as August Bebel put

it

in

1905, ‘‘as a purely

defensive measure, designed to safeguard the exercise of power legitimately acquired through the ballot” (Schorske 1972:43).

Under these conditions the attitude toward electoral participation was understandably cautious. Socialists entered electoral politics gingerly, ‘‘only to utilize them for propaganda purposes,” and vowed ‘‘not to enter any alliances with other parties or accept any

of the

SPD

compromises” (Resolution of in

1870). At best,

was one instrument among

many

the Eisenach Congress

thought, universal suffrage

others, albeit

one

that

had

‘‘the

incom-

parably higher merit of unchaining the class struggle,” as put

it

in

to

1850 (1952:47). Elections were

to

Marx

be used only as a

ready-made forum for organization, agitation, and propaganda.

The

typical posture

is

well illustrated by this motion offered in

1889: ‘‘Since Sweden’s Social Democratic Workers’ Party

propaganda party,

i.e.,

it

considers

its

main objective

to

is

a

be the

dissemination of information about Social Democracy, and since

The Decision

participation in elections

recommends

gress

17

to Participate

were also useful

is

a

good vehicle

for agitation, the

participation” (Tingsten 1973:357). Elections

providing party leadership with a reading of

in

the revolutionary fervor of the masses. But this to

promise

The

last

the State in

at the

moment when

edition of

socialists

is all

they seemed

decided to participate.

The Origins of Private Property, Family, and

which appeared during Engels’ lifetime

1891 the assertion that universal suffrage

of the maturity of the working class.

anything more

Con-

in the

is

still

contained

merely “the gauge

cannot and never will be

It

present-day state”

(

1942: 158) and continued

somewhat ambivalently, “but that is sufficient.” Each step toward full participation rekindled controversies. The German Social Democratic Party argued whether to allow one of its members to become the Deputy Speaker of the Reichstag, whether to vote on the budget, even whether to trade votes in the second round of elections (Schorske 1972).

wegian Labor Party refused round

even

though

in

1906

absolutely

The Nor-

to trade votes in the

second

no compromise was implied

(Lafferty 1971:127). In 1898, a survey of the opinions of promi-

nent leaders of the Second International

showed

that while inter-

ventions into bourgeois politics were thought to be

at

times

advisable, six of the respondents voted 'famais'' with regard to participating in a government, eleven admitted

it

was possible

only "dres exceptionnellement,'’'' and a minority of twelve thought that

such participation

in the

is

either always desirable or at least

case of Millerand (Fiechtier 1965:69-75).

Of the

it

was

sixty-nine

Swedish Social Democrats polled by telegram whether the party should join the Liberal government against participating (Tingsten

“suspended”

class struggle

ject of intense

first

1973:418). While

and entered

ments before the end of World decision to form the

1911, sixty-three responded

in

War

I,

parties

into coalition govern-

even

Labour government

in

Great Britain the

in

1924 was a sub-

polemics and was rationalized as an opportunity to

acquire experience necessary for the socialist era

Opponents of participation seem the political spectrum.

ward

some

full participation,

tion according to

to hold a

(Lyman

1957).

permanent place

in

As established parties take each step tonew voices emerge to continue the tradi-

which the belief

in the

parliamentary battles

“between frogs and mice” (Luxemburg 1967:37) is simply a manifestation of what Marx called under very special circum-

Electoral Participation

18

and

Its

Consequences

\

*

stances “parliamentary cretinism” (1952:77). “Integration

is

the

price,” Horkheimer repeated in 1940 in the anarchist memento,

“which individuals and groups must pay

in

order to flourish un-

der capitalism” (1973:5). “Elections, a trap for fools,” was the title

of an article by Sartre on the eve of the 1973 French parlia-

mentary elections. '"Voter, c'est abdiquer" shouted the walls of Paris in 1968.

Electoral Participation and Class Organization The

rea

^ why

involvement

in representative politics

geois society has never ceased to evoke controversy

is

of bo urthat th e

very'JELoOaEngpart in this particular sy;stem_shapes jhe„iT)LQvement for socialism and its relation to workers as a class. The recurrent qu estion is whpjhpr invplvpmpnf in elect oral politics c an result in socialism or

italist,~'^iaL-or([er find a passage

m ust

Is

if

strengthen the existing, that

is,

cap-

possible for the socialist movernent to

between the two reefs charted by Rosa Luxemburg:

“abandonment of the mass character or abandonment of the final goals?” (Howard 1973-74:93). Participation in electoral politics is necessary if the movement for socialism is to find mass support

among workers,

yet this participation appears to obstruct the at-

tainment of final goals. Working for today and working toward

tomorrow appear argue, as E. H.

as horns of a

Tawney

dilemma

to

many

observers

who

did about Labour, that the “Party can be

either a political agent, pressing in the Parliament the claims of

wage earners or it can be an instrument for the establishment of a socialist commonwealth. What it cannot be is both at the same time and in the same measure” (1923). particular groups of

Participation imprints a particular structure on the organizat ion

of worLei:s.asL.a .class. TRese effects class relations have

is

internal

economic struggle and

but an artificial product, even

standable, of the parliamentary period.

peaceful development,

economic struggle

upon

been best analyzed by Luxemburg: “the divi-

sion between political struggle and

separation

participation

is

On

if

their

historically under-

the

one hand,

in the

‘normal’ for the bourgeois society, the

fractionalized, disaggregated into a multi-

tude of partial struggles limited to each firm, to each branch of

production.

On

the other hand, the political struggle

is

conducted

not by the masses through direct action, but in conformity with

Electoral Participation and Class Organization

19

the structure of the bourgeois state, in the representative fashion,

by the pressure exercised upon the legislative body” (1970:202).

The first effect of the structure of the bourgeois state is thus that wage earners are formed as a class in a number of independent and often competitive organizations, most frequently as trade

unions and political parties, but also as cooperatives, neigh-

borhood associations, clubs, capitalist

democracy

is

the realms of politics

and ideology (Lukacs 1971:65-66; Poul-

the system of production

are capitalists or

all

appear

Hence, even

individuals-citizens.

characteristic feature of

the individualization of class relations in

who

antzas 1973). People

One

etc.

if

wage

earners within

in politics as undifferentiated

a political party succeeds in

organizing a class on the terrain of political institutions, economic

and

need not coincide. Multiple unions and

political organization

riini.M

'

parties_Qften represent different interests and compete with each

Moreover,

other.

while the

cl a ss

base of unions

is

confined to

more or less permanently emp InypH wage e arners, pol itical parties which or ganize workers can als mohi li7p..pmp]e who r^^nnot be member s of union s. HenceJh^

certain groups of people

is

jL perrnaiiept trni&ioxiJj^tweealhe

narrow er int£r^ts of unions

and th^ htundar infprpll/~i

od O'

u o

q d o

q "3

A

o

3 3 C 3

3 c «

m

3 c 3

Urn

**

ri on

c

"0

q in

r

3-

o

O'

q m nC

U4

U

u c

eg

u

r^’

-o 9t

u
/^

•3'

q

on

on oo

3

-

a>

rj

O' rn

q d n

q d m

q ri rj

>

c •o

s/i

1/

C o 'c

u W

T3

o OJ

N s

_N "c _o

c

O'

O

c/3

U

c

O

C'J

ri

r^

Tf

r^i

q od

(/3

0/

rn

0^

o q rd d r^ o —

V-)

od O'

u 0

oo on 00

o

A.

t-

2 > c o

x:

u

q d — oo

(N

oi

O'

—

3 3 C 3

00 od m

3 3 C 3 ••—1

q q q —L d d 00 o O'

00 ri oo rn

d

u

—

—

rn

ri

q q 3-

d

rod

q 3"

—

oa

7Z Oil k.

o ^5 ^/5

J2

u o u

o o O'

q q (N d oo o d

oo nO r

00

o

c 3 >N 3

E 3 C

c/5

k«

.

,

O' C P 3 0 P ’3 x: P

di

3 c P k.

>>

P

X 0 O'

rn mi a^ o^ O'

0

0

sO

X

0

k.

3

)

c/3

C P 0

c/3

c/3

00 r-

X X E 00 VM

-

c

CO :d

c

:£>

.B-

3 E

Cl c/3

> Ck> P (U 0 > m > 0

P X p c E 3 p U p r-C E

0

o c o -3

P > o

"Jo

>>

c/5

^3

c/5

p p P 'C

3

3 3

'w k.

c/5

T3

C

_3

c p p

P

-o

o

X X>3 Q Q c m cp 2 > T3 3 c c 00 O m o

*C

_3

cu

O'

CQ

CU

a

p

'C Cu od •3

'3

o C

c/5

o c

p p

CO'

k>

C 3

X

d)

Xa. u>

c x'

c 3

O p 3 3 O P

P E

>1 X!

0 oc

•3

>

ON

that k

—

would

like to

maintain

composition they seek to achieve.

that the party

N(t) and kW(t)

indicates the class

>

N(t)

is

If

supported by a sufficient

N(t)/W(t), and as a consequence

>

0, then leaders will be free to

pursue a supraclass strategy and recruit more non workers. The

and thus

it

can afford to

dedicate itself to the middle-class electorate. But

if

there are not

party

is

sufficiently working-class based

enough workers or too many so that k


oi

oi

kO

UJ

o f3

c.

u

y) •V

OfJ

OC

c

^ —

—

k c 70 r3 3 '3: O c o .2 2 'o 3 c o C a. y)

Class-only

;/3

o

34 34

3-

k.

o

00

o.

with

OO

>3>

o

(3 3" 3) 34 3" 13) 0 O' 00 »3) O' f3

—

O —

»3)

OO 3" 34 '0 3’

—

y)

3 — o K 3 c o O O O x: (u o y)

(U

•S y)

•-

Comparison

Cl, 045

y)

i2

O o

3.1E-S S S

u .X

4.2

•2 Table

^ 1>

OQ

C >> c 3 o 3 o (U c o •o B _3 c

1.

This

not a surprise

is

—

for

each additional middle-class voter, the party loses {dip) workingclass voters in equilibrium.

Thus

if

dip

>

the party loses net

1,

votes by turning for support to the non workers. But off




struc-

as the ratio of

to allies in the electorate is larger than the trade-off,

dip, the party

trade-off

is

is

best off pursuing a class-only strategy

steeper than

the trade-off

and

if

dle-class voters

even

if

there are

party strategies

recruiting allies, the party if

if

the

and best off with a supraclass strategy

milder. But

is

the electorate

1

XIL

may

more allies than workers are somewhat effective

if

in in

be electorally best off to seek mid-

the trade-off

is

very steep.

Table 4.2 shows carrying capacities associated with the

strat-

egies that best reconstruct the actual experience of socialist voting

compared with

would have resulted from the pure supraclass and the pure class-only strategies. These numbers indicate what proportions of the electorate each major socialist party or the left-wing parties together would be able to recruit and hold if the class structure were frozen at its 1970 levels. The column associated with the “full supraclass strategy” indicates whether the party would maximize its vote share if it pursued a strategy of mobilization of nonworkers independently of support from workers. Since such a strategy would alter the the carrying capacities that

logic of socialist electoral competition, cific

numbers

we

to identify the potential pool

hesitate to provide spe-

of nonworker socialist

support. Orders of magnitude are sufficient, however, to this strategy

with the supraclass posture

allies is limited

ever a

egy

“max”

that

in

compare

which mobilization of

by the success of the party among workers. Wherappears in the “full supraclass” column, the

would involve a complete abandonment of

strat-

class

is

Was

the Stagnation of the Socialist

Vote Inevitable?

1

electorally superior both to the limited supraclass strategy

and

13

to

The remaining columns in this table indicate composition that would characterize the partic-

class-only strategy. the resulting class ular strategies

—

the actual strategy and the pure ones. Instead of

presenting the values of

k,

however, we have translated these

val-

ues into proportions indicating

how many workers would

found among party supporters

the party adopted the particular

if

be

strategy.

The Finnish, French, and

the pre-1933

parties faced an equilibrium trade-off ty, as

German major

socialist

which was steeper than

uni-

did the Left as a whole in Belgium, France, and Germany.

The Finnish Left faced

a milder trade-off, but workers constituted

only about 19 percent of the Finnish electorate in 1970. In these cases the unlimited supraclass strategy

is

all

best electorally. In

where the trade-off is steep, the pure class-only strategy is second best and the supraclass strategy which limits the recruitment of allies on the success among workers is the worst. These

the cases

parties thus faced a bifurcation in their road to

abandon

all

was the worst from the one

in

to

to decide

concerns with class or alternatively to struggle to

keep the party class pure. whether

— they had

A

half-hearted middle-class orientation

They had to decide would be no workers or

electoral point of view.

have a party

in

which there

which there would be only workers.

Denmark, Norway, and Sweden the limited supraclass strategy would maximize the vote. The Left as a whole and the major socialist parties in these countries would get most votes by turning In

for support to

nonworkers and they could get

retain a sizable support

among workers. They

or-nothing choice as far as class

is

support and

still

did not face an

all-

this

concerned. They would have to

which workers were a minority, but they could choose continuously between having more workers and fewer votes or fewer workers and more votes. In all these cases the pool of the potential middle-class supporters was, by 1970, tolerate electorates in

larger than the proportion of

workers

in the electorate.

Thus, the

pure class-only strategy was electorally the worst.

Was the Stagnation of the

Socialist

Vote Inevitable?

Within the realm of available choice, leaders of left-wing parties disappoint those

who

believe that their exclusive concern

should be, to maximize the vote.

Among

the parties

is,

or

which would

Are Socialist Leaders Vote Maximizers?

114

maximize their vote share by abandoning workers altogether, none has accepted this option. This is not surprising given all we know about the view of the future with which socialist leaders analyzed their choices eighty years ago. Most of them certainly

would require abandoning the altogether would eventually be the one that would win

did not expect that the strategy that proletariat

most votes. Bernstein had the correct

the

seen, there

was

in fact little

this strategy,

we have

empirical support for his arguments.

There were many more reasons have adopted

intuition but, as

why

and the

socialist leaders

could not

fact that they did not

not be taken as the evidence that they did not want to

should

maximize

their future votes.

But the evidence against pure vote maximizing goes further.

SPD

our numbers can be believed, only the pre-1933 the second best,

they could.

The

by keeping the party as pure leaders of the Finnish Social

in

If

leaders did

terms of class as

Democrats had

little

choice and placed themselves somewhat closer to their better strategy, again the class-pure one.

der consideration, that to

is,

until

At

least

during the period un-

1970, the French Socialists seem

have chosen almost the worst electoral strategy. They

tried to

conquer middle-class voters without losing workers and did not succeed

in either.

SPD, placed strategy,

egy

is

Left,

which includes the post-war

again exactly halfway between the second best

which would have been class pure and the worst strategy

which turned

The

itself

The German

in a

halfhearted

way toward

three Scandinavian parties for

electorally superior,

by quite

a

the middle classes.

which the supraclass

wide margin

in

strat-

Norway and

Denmark, nevertheless have opted to pursue strategies that have kept them closer to their working-class base. To maximize votes, the Danish Social Democrats would have to tolerate an electoral support in which workers would have constituted 41 .4 percent. In the light of our calculations, they have opted for a class

composi-

—

which workers would make up 72.5 percent a strategic decision which placed them closer to the electorally inferior classtion in

only strategy. The Norwegians would have maximized their share of the vote with an electoral base in which workers would constitute

29.4 percent; they have opted for one

in

which workers

would make up 47.3 percent. Finally, the Swedes would have been electorally best off by bringing the workers share in their electoral support to 20.6 percent. Instead, they chose to compete

Was

the Stagnation of the Socialist

Vote Inevitable?

1

15

with a party of which 77.6 percent of the electoral support would

be from workers. All these parties did

embark on

strategies designed to recruit the

middle classes. Danes turned for support of peasants as early as 1888 and always kept the party open to nonworkers. Swedes made their first deliberate

working class

to

extend the electorate beyond the

1911 and formalized this

in

wegians began

move

same way

in the

move

1921. Nor-

in

as the Danes, then adopted a

working-class posture, but by 1933 joined the other two parties turning to “the people” (Martin 1972:48).

decided to turn to office employees in the

wage

embarked on

the

indicates that

none of these

in

The Swedish party late 1950s, when it

earners’ solidarity policy. Yet our analysis parties has

gone even half of the way

Even the Norwegian Labour Party, which was willing to tolerate more middle- than working-class supporters in its electorate, could have gained by becoming more supraclass oriented. they could have to recruit the middle classes.

At

this

time

we must

return to the causes of the stagnation of

we drew from the analysis one hand, we argued that the

the socialist vote, for the conclusions

of trade-offs are implausible.

On

the

social democratic parties faced a relatively mild trade-off in the

three Scandinavian countries because the trade unions played an

we

important role in organizing workers as a class. But over, on the other hand,

showing

might have been seen as

that these parties are quite free

from any attachment strategies.

values of the criterion of class composition,

ties

k,

The

to

actu-

with which par-

chose strategies, were higher for the Danish and Norwegian

Left and for the Social it

that

working class when they choose electoral

the al

numbers

passed

Democrats than

might have seemed as

their

if

the

Thus Scandinavian Left benefited from for the other parties.

symbiotic relation with the unions without being

strained

by them. This would have been the best of

at all

all

con-

possible

worlds.

We

see

that social

now, having analyzed democratic leaders

which were

in fact far

point of view. In

the alternatives that

in these countries

were open,

pursued strategies

from optimal from the purely electoral

Denmark, Norway, and Sweden,

the Social

Democrats could have permanently conquered a majority of the electorate

won

—

all

those

who had

elections with an

the right to vote

overwhelming majority

— and repeatedly if

they had only

1

Are Socialist Leaders Vote Maximizers?

16

embarked from

the beginning and persevered unswervingly in a

would distance them permanently from workers as a class. Just examine figure 4.2 which compares the share of the vote the Danish and the Swedish Social Democrats could have obtained with the actual historical record of strategy of electoral recruitment that

their electoral

performance.

the Danes, nor the

And

yet neither the Norwegians, nor

Swedes did pursue

would have been most successful

in

the course of action that

winning elections

at the cost

of abandoning the working class.

The reason why these parties could have made this much of a difference by following supraclass strategies is that they faced relatively mild trade-offs

emphasis on class

ior

would have had

they diluted the

organization and propaganda. But the

in their

reason, at least in part, parties

among workers when

why

the

relatively

abandonment of class by these little effect on the voting behav-

of workers was that workers were organized as a class by

large, concentrated,

and centralized unions.

closes rather viciously, for

it is

pendence with the unions

Democrats

that Social

all

the unions. Indeed, before

here the circle

precisely because of their interde-

Norway, and Sweden could not adopt maximized their vote. Since their foundation,

And

strategies that

three parties

Denmark, would have

in

were closely related

1900 the two organizations were

most indistinguishable, particularly

at the local level.

al-

While the

Danish party had already abolished collective membership

Norway and particularly in Sweden the of union members in the party continued

to

in the

mem-

1870s, in

collective

bership

to be important.

membership of the Norwegian Labour Party and 73 percent of the Swedish SAP consisted of union members affiliated collectively. Unions provided, however, not only members but also finances for electoral campaigns. The leadership of the unions and parties often overlapped, and in Denmark several union leaders became Social Democratic prime min-

As of 1974, 40 percent of

isters (see

the

Elvander 1979; Esping-Anderson 1985).

Unions are necessarily class organizations, this term.

Their potential membership

terms; unions are not and cannot

is

in

two senses of

defined only

in class

become organizations of Catho-

women, or citizens; they can associate only workers or at most wage earners. Hence their constituency is necessarily more lics,

restricted than that of electoral parties for

whom

the ultimate pool

8Z61 ez6i 0Z61 8961 t^96l

0961 8961 9961 2961 8t^6l

pm 0r6l 9861 2C61 8261


-

o »— u

Sweden

and

1^261 Denmark

1261

0261 Zl6l

in

^161 t^l6l I

Parties

161

Democratic

Social

of

Shares

Vote

Actual

O'

and

< Possible

o hu

4.2

Figure

Are Socialist Leaders Vote Maximizers?

118

of support consists simply of citizens, whoever they

may

be other-

wise.

But perhaps more importantly unions are organizations

whose

principal task

is

to protect their

members from competition

with one another; to prevent workers from bidding their wages

down

in

exchange

for

employment. And

this

means

union

that the

organization constitutes a collective good only for those people

who may conceivably

find themselves in competition for the

same

kinds of employment, not for everyone. Thus unions represent interests

narrower than those that can be organized by

political

parties.

The

trade unions in Scandinavia often resisted the adoption of

supraclass strategies (Hentila 1978:331-32; Scase 1977:323; El-

vander 1979:18). Heidar reported (1977:300) that there

were “strong counterforces

the ‘organic’ link the trade union

to this process of

between party and

movement which

vote.’’

breakdown

in is

a potent force in resisting

mandate

for the working-class

Martin (1972:15, 168ff) showed that most leaders of the

Norwegian Labour Party had thought its

were

to

would have had

to

that if the party

actively pursue the white-collar electorate

weaken

Norway

class. In the forefront here

is still

the trend towards a purer political

in

it

organizational ties to the trade unions, specifically, to

abolish collective affiliation.

Belgium, Finland, France, and

at least the

pre-1933 Germany,

on the one hand, and Denmark, Norway, and Sweden, on the other hand, constitute two different types of situations. In the

first

group

of countries unions did not remove the burden of organizing workers as a class

from the shoulders of

political parties.

They did

not

play this role because they were apolitical, or because they were politically fractionalized, or

because they were organizationally

decentralized, but they could not and did not. class ideology

The

result

was

that

and class organization were highly vulnerable

socialist electoral strategies.

The

electoral trade-off the left

to

wing

was so steep that any major turn for support to the middle classes would have been disastrous for their vote-getting power. The best the Left could have done from the purely electoral point of view was indeed to abandon the working-class electorate altogether, but that option was not feasible for reasons of ideology as well as for fear that the union movement would become completely independent and apolitical parties encountered in these countries

(Schorske 1972;

Gay 1970:136-37;

Fiechtier 1965:148-52). In-

Are Socialist Leaders Vote Maximizers?

deed

in

1925 when the

SPD

1

19

drafted a program which asserted that

“the struggle of the working class against capitalist exploitation

is

necessarily a political struggle,” party leaders were soon taught by the

ADGB

that

it

is

“not only an economic one, but also neces-

sarily a political struggle”

best option

—

(Hunt 1970:175-76).

a pristinely ouvrierist stance

— was

win majorities and mandates because the

to

“great industry” turned out to be socialists

much

And

the second

good enough development of the not

less labor-intensive than

had expected and workers never became the majority.

The Danes, Norwegians, and Swedes were in a different predicament. The trade unions in these countries grew rapidly and uni-

became institutionalized by a bargaining system. As a result. Social Dem-

fied at an early stage. Class relations

nationwide collective

ocrats in these countries could pursue middle-class oriented strat-

egies at a tolerable or even negligible cost. But that very

same

partner which took from the parties most of the burden of organizing workers as a class

imposed constraints on

the degree to

which

these parties could freely pursue their electoral opportunities.

Are Socialist Leaders Vote Maximizers? The interdependence between socialist parties and trade unions is not the sole constraint which binds strategic alternatives. Internal relations within each party are likely to provide another con-

members may have preferences distinct from party leaders and leaders may be unable to wage campaigns which do not have the support of members. Many people become party acstraint.

tivists

tants

Party

because they have strong beliefs about issues and such mili-

may

be unwilling to follow the leaders

in

an opportunistic

pursuit of votes. Unfortunately, while Schorske’s (1972) study of

SPD between

the

1905 and 1917 highlighted the tension between

party leaders and the rank and

file,

we could

not find direct evi-

dence of members’ resistance to leaders’ electoral strategies.

We

remain persuaded, nevertheless, that internal relations within the party

must

restrict the strategic

Thus the issue

is

freedom of the leadership.

not whether leaders of electoral parties seek to

maximize the vote or follow ideologies. Leaders of

electoral par-

win more votes rather than fewer: this we do not doubt. The question is under what constraints they maximize votes, and the issue may be at most whether there is any room for ties

prefer to

Are Socialist Leaders Vote Maximizers?

120

choice

left after all

constraints, to

the constraints are considered.

assume

To

ignore such

can pick any strategy,

that party leaders

address themselves to any group with any program, reduces the

To conclude

study of parties and elections to empty formalisms. that party leaders

who do

maximize must be victims of

not

the

how many obstacles stand in the way of those who would very much like to win elections. One of these obstacles is intertemporal. Once we look at party strategies in historical perspective, as a process, we cannot view irrationalism of ideology

is

to forget

each election as an event that had no antecedents and would have

no subsequent consequences.

We

must understand how the past

choices have shaped the current alternatives, and

how

far into the future party leaders look

when

we must

ask

they examine the

eventual consequences of their strategic choices.

The

analysis of long-term consequences placed an unfair bur-

den on party leaders. The question

we asked above was whether

leaders of left-wing parties sought to

long run, trying to get as

from now

elections,

many

ologues were able to think

in

the vote in the

votes as possible in

all

future

But while the Catholic Church

to eternity.

perhaps able to see the future

maximize

in

millennia and

centuries,

it

is

Communist

is

ide-

unreasonable to ex-

much attention to anything who maximize for the long

pect leaders of electoral parties to pay but the proximate future. Politicians

run end up writing

We

politicians. to

win

all

memoirs within

a very short one, even socialist

should ask, therefore, whether party leaders tried

the votes they could in the proximate elections, re-

gardless of the consequences the particular strategy might have for the longer run.

Two

parties are particularly interesting since they represent the

clearest contrast, although each

same

situation.

The Swedish

is

representative of others in the

Social Democrats, as well as the

Danish Social Democrats and the Norwegian Labour Party, could have improved voting performance by adopting a strategy that

The French Socialists, as well as the Social Democrats in pre-1933 Germany, were in a situation in which their best strategy was to break all ties with the working class. Their second best strategy, however, was would distance them further from the working

to

maintain the electorate pristinely pure

SPD

did

all

it

could

in this direction.

in

class.

terms of class, and the

But the French Socialist

Party adopted a strategy of mixing appeals to workers with the

Arc Socialist Leaders Vote Maximizers?

121

organization of the middle classes and the effect was quite suboptimal electorally.

These were the conclusions of the analysis of long-run maximizing. But the consequences of the different conditions these

two

parties faced during their electoral histories manifest

them-

shows what would have been the consequences if the French Socialists and the Swedish Social Democrats chose, at any moment, either of the pure stratselves also in the short run. Figure 4.3

egies and followed this strategy during just tions.

The

point of origin

—

the bigger dots

two consecutive

—

at

elec-

each election time

represents the vote share which the party had actually obtained.

The two

which originate from each point represent the vote party would have obtained had it adopted at the time the

lines

shares the

pure supraclass and the pure class-only strategy. For example, the

French Socialist Party

won

10.1 percent of the electorate in the

1962 elections. Had the Socialists adopted

at this

moment

the

pure supraclass strategy, they would have gained 15.5 percent in the 1967 elections.

Had

they persevered with the

1968 elections, they would have

for the

percent of the electorate. In turn,

if

won

same

strategy

the votes of 8.9

the Socialists had opted for the

pure class-only strategy, they would have obtained 13.3 percent in

1967 and 13.5

The

in

1968.

analysis of alternatives available in the short run to the

leaders of the

Swedish Social Democratic Party confirms the con-

clusions derived from the long-run analysis. At each history. Social

Democrats would have improved

moment of

their electoral

performance by embracing a pure supraclass strategy. Until the

male suffrage, about 1905, the trade unions any opening on the part of the Social Democrats (Hentilla

battle for universal

resisted

1978:331-32). But having learned about the importance of liances, the party gingerly

changed the clause by which class to “all the first

gone the

was

full

would have won

it

its

ranks in

identified itself

when

1911,

far

way

in the

from a pure supraclass in

its

it

from the working

oppressed” (Tingsten 1973). Nevertheless,

strategic turn

party

opened

al-

strategy.

this

Had

the

appeal to the middle classes,

two successive elections of 1914

first

it

25.3

and then 32.9 percent of the electorate, instead of the 20.9 and 23.9 percent

it

did win. Since the party would have obtained 19.4

and 20.5 percent with the pure class-only strategy, the 1911 turn to the

middle classes seems halfhearted indeed. The quest for

30

I

I

I

I

I

I

I

I

I

I

I

I

1

I

1

I

I

I

I

»

CNOq— Ckt;J00(ni-O'O'O'O' Ooo(Nirv.aO'0'OOOOOOOO0'C>'O

ELECTIONS

ELECTIONS

Are Socialist Leaders Vote Maximizers?

middle-class voters

had suffered their

was extended

in

123

1921 after Social Democrats

and formed the

electoral reversal

first

nority government. But again the strategy adopted in 1921

from the pure supraclass orientation

— unreserved

mi-

first

was

far

pursuit of the

middle-class electorate would have resulted in a share of 26.8 percent in 1924 and 32.7 percent in 1928, a pure working-class

would have brought 21.2 percent in 924 and 23.1 in 928 while in fact that party won 21.7 and 24.9 percent in these two elections. Even the wage-earner campaign of the late 1950s did stance

1

not constitute a vote maximizing strategy.

barked upon such a strategy

in

1958,

it

1

Had

the party

would have won 46.4

percent in 1960 and 51.9 percent in 1964. In fact, tively

em-

40.9 and 39.4 percent of the electorate,

it

still

won

respec-

closer to the

pure class-only and electorally inferior strategy which would have

brought 37.8 percent by 1960 and 36.8 percent by 1964. This analysis merely confirms, therefore, the conclusions derived from the examination of the strategies oriented toward long-

run objectives.

A

similar dissection of the behavior of the Danish

Democracy and the Norwegian Labour Party would have brought the same results. In these countries, as table 4.1 has alSocial

ready shown, there

which

is

is

no intertemporal trade-off between

that

best in the short and in the long run. Supraclass strategies

are electorally superior in the short run

more votes it would win. vote maximizing supraclass

perseveres the opt for the

and the longer a party If party leaders

strategies,

it

is

did not

because

they cared about the class orientation of the party, not only about votes.

The notion sake

is

that party leaders seek to

not only ad hoc, but

which everyone derives

utility

it

maximize votes

also conjures a strange world in

from consumption

become leaders of electoral parties Pizzomo 1983). Our evidence demonstrates

they

cialist parties

Figure 4.3

for votes’

until

and unless

(Hirschleifer that

1976;

even when so-

could indefinitely increase their vote shares by mov-

The Choices Facing French

The dots represent

Socialists

and Swedish Social Democrats

The lines originating from them at party would have obtained had it fol-

the estimated vote share.

each time represent the vote shares the

lowed pure strategies during the two next elections. Vote shares as a proportion of the electorate.

Are Socialist Leaders Vote Maximizers?

124

ing

away from

their working-class roots, socialist leaders

had

other reasons to care about the class orientation of their parties.

Hence our analysis joins

port the assumption that party leaders ity

rather than votes (see

Votes are worth only as

The French Note

socialists

Wittman 1983,

much

moving toward

moment of

effects if

it

lasted

Communist

its

mandates they embody. tighter spot.

became

in the

The next

have suffered two elections

elections. Until 1924, before

inferior after

later the loss

to the result they

Had

perverse

a serious contender, the limited su-

was only mildly

pure class-only strategy.

elections.

move would have had

two consecutive

compared

two

the middle-class electorate, but during

chosen the pure supraclass strategy

electorate as

much

in a

performance

electoral

history such a

Party

praclass strategy Socialists

the expected util-

for a recent review).

electorally inferior within the span of

election by

the

as the

always have been

party could always improve

every

maximize

4.3 that the limited supraclass strategy invariably

in figure

became

which sup-

the line of empirical findings

they

made

two

in

elections.

Had

1902, they would

of 1.5 percent of the

would have had with a

this

choice

1906, after

in

would have foresaken 2.3 percent two elecby 1914. But after the Tour Congress, the cost of the

the unification, they tions later, flirtation

with the middle class increased.

If Socialists

had opted

would have received by 1924 almost 4 percent less than would have resulted from a pristinely ouvrierist stance. These costs continued to be high during the interwar period, but the difference between the result of the two pure strategies decreased after 1945. Only in 1968 did the strategic choice made two elections earlier, in 1962, matter: the difference was 4.6 percent of the electorate. And in 1978 the decifor the pure supraclass strategy in 1914, they

sions

made

in

1968 could make a difference of 3.2 percent of the

electorate.

While the two world wars roughly mark the periods tegic choices of the

French

in the stra-

socialists, the intertemporal trade-off

was present during their entire electoral history. After two elections socialists would have always won more votes had they pursued working-class oriented strategies. But for the next election the supraclass strategy

was always

superior, and often

by a sizable

chunk of the electorate. In 1902 socialists could have looked forward to winning in the next election 5.5 percent more with the pure supraclass than with the pure class-only strategy; indeed they

Are Socialist Leaders Vote Maximizers?

could have

won

5 percent

125

more than they did

probably not the time, however, to

flirt

was

1906. This

in

with the middle classes

since the pressing issues concerned the unity of the socialists

themselves and their relation to the trade union movement. Decid-

ed on

SFIO by 1924

—

which

made

1919, the pure supraclass strategy would have

in

far the

its

most important force

in the Cartel

des Gauches of

share of the electorate would have been 31.7 percent,

translates to almost

one half of the votes

cast.

Chosen

in

1936, the supraclass strategy would have resulted in a significant gain by 1945.

Chosen

in

1967,

it

would have prevented

the beat-

ing socialists received in 1968.

Thus

was always superior for class-only strategy was superior elec-

the limited supraclass strategy

the next election while the torally for

one election

later.

French

socialists

were caught be-

tween short-run maximization, which tempted them the middle classes

around 1890 and again

reduced them to a party of teachers

in the

to

flirt

with

1930s and which

1950s, and the longer-

in the

run superiority of the ouvrierist stance. During the interwar period, under pressure from the communists, the

SFIO maintained

a

working-class oriented stance and even expelled the main advocate of the supraclass strategy,

After the war, the party

Marcel Deat (Touchard 1977).

moved toward

the middle-class constitu-

ency and the electoral results were rather pathetic.

newly reconstructed electoral performance after

significant that the

improved

its

math of the 1977

split

It is

therefore

Socialist Party sharply it

decided, in the after-

with the PCF, to wrest

hegemony over

the

working class from the communists (Daniel 1978).

Thus where

in the

case of the French socialists, as well as the

strategies

SPD,

which maximize the vote for the next election

were electorally inferior soon afterwards,

it

is

just impossible to

speak of vote maximization ahistorically. The fiction of elections as a series of

unconnected events

is

no longer tenable.

cast in each election are indeed a cumulative strategies, then party leaders

If the

votes

consequence of past

cannot every few years begin history

would place it best electorally in a known distribution of public opinion. The assumption that public opinion is exogeneous to party strategies is absurd, whether one thinks that party leaders maximize votes (Downs 1957), substantive goals (Wittman 1973), or a mixture of anew, picking for

their parties the strategy that

both (Schlesinger 1975). Parties mold the ‘’public opinion”: they

Are Socialist Leaders Vote Maximizers?

126

present the public with images of society, evoke collective identi-

commitments. Thus the distribution of

fications, instill political

preferences parties encounter

Gramsci emphasized, which

each electoral campaign

in

a result

is,

as

of past actions; the conditions

parties encounter at the present are a product of strategies

they chose in the past.

And many

of today’s objective conditions

when

are the errors of yesterday. Therefore, strategies for the next election they

tions they will encounter

when

party leaders choose

must worry about the condi-

they will be making their future

strategic decisions.

The reason

the socialist vote stagnated after an initial burst of

growth, the reason socialist parties did not win the mandate of an

overwhelming majority of the electorate, is not that party leaders were blinded by their ideology to existing electoral opportunities.

The reason strained

is

that their choices

by class-structure

majority.

—

were constrained. They were con-

the fact that workers never

They were constrained

in

became

a

Belgium, Finland, France,

and Germany by the electoral trade-off they encountered

among

workers as they sought to win the votes of the middle classes.

They were constrained their

in

Denmark, Norway, and Sweden by

symbiotic relation with the organizations of workers as a

class, the trade unions.

They probably made

strategic mistakes.

People do make strategic mistakes and socialists did find cult to

it

diffi-

throw off the blinders imposed by the “scientific so-

cialism’’ of the

Second

when placed within

International.

But these mistakes pale

the context of constraints. Ultimately,

ably mattered relatively

little

whether

it

prob-

socialist leaders did every-

thing they could to win elections. Their choices were limited.

Appendix

to

Chapter 4

This appendix contains formal definitions and explanations of

some terms used assertions,

in the analysis,

proofs of the more important

and a discussion of alternative assumptions. Most of

the material involves only elementary conceptual

operations and should be accessible to anyone

who

and algebraic wishes to

in-

more deeply into the analysis. The only section which is mathematically more difficult is the last one, concerning maximi-

quire

zation.

Appendix

127

The order of this appendix does

We

not follow the

body of

the text.

begin by defining the pure strategies and proceed immediately

to the analysis

of the short tun. Only then limits are taken and the

Two

discussion concerns the carrying capacities.

presented subsequently to

els are

maximizing

mod-

the robustness of the re-

contains the proof of the assertion

sults. Finally, the last section

that vote

show

alternative

depend on the

strategies

intensity of the

equilibrium trade-off.

As

model

a reference, the

^W{^)

=

p[Xit

^Nit)

=

q[L(t

r(r)

The

=

-

1)

+

+

Wit)

+

1)

given by:

is

-

-

W(/)) N{t)\

+

N* = r*

=

N{t)], (4A.2)

(4A.3)

A^(/).

Appendix

to chapter 3):

(4A.4)

idlp)N*,

qLliq VT*

-

c[A'(/)W(/)

steady-state values are (see the

W* = X —

(4A.1)

dN(t),

+

+ ckW/iq +

c)

(4A.5)

c),

+ N*.

(4A.6)

Pure Strategies

A

pure supraclass strategy

in the

recruitment of

all

is

would

a course of action that

result

the allies in the next round of voting

party strategies were perfectly effective, that

if

is,

parties could

always get what they wanted. Thus a pure supraclass strategy time

t

is

Another way

+

1)

=

to write the

to recruit the

AA(r)

A

at

defined by a number kHt) such that

Nit

want

if

same

remaining

=

Lit

1)

when

1)

criterion

allies,

pure class-only strategy

maximize

+

L(/

=

c is

\

(4A.7)

.

that the party

or

—

A^(/),

is

a course of action that

when

c

=

the ratio of workers to socialist voters.

egy can be defined by a number

would

k^it)

which

is

(4A.8)

1.

Hence

would

this strat-

the solution to the

problem given as

max Wit + \)/[Wit + k

1)

+ Nit +

1)].

(4A.9)

Are Socialist Leaders Vote Maximizers?

128

Since

our model the only state the party can modify during the

in

next election

is

N{t

-I-

maximizing the

1),

socialist.

pure class-only strategy

+

Nit

which

or,

is still

Nit

state

+

by

=

is

=

c

(4A.10)

1,

when

=

c

1

(4A.

.

1

1

taken to be the strategic variable the

accessible at

party can influence only Wit

+

when

0,

—Nit),

that since k

1) is

such that

k^(t)

=

1)

to define the

the same.

AN{t)

Note again

Thus, the alternative way

is

in the

minimizing the proportion of non-

socialist vote is equivalent to

workers voting

of workers

ratio

t

+

but the state Wit

+

1) is not.

The

and indirectly, by affecting Nit

2),

Short-term consequences of our analysis are obviously sen-

1).

sitive to this lag structure but

To

calculate k^it)

longer term consequences are not.

we proceed

as follows:

first,

substitute the

=

objective (4A.8) as the left-hand side of (4A.2), setting c

ThenL(r +

-

1)

=

Nit)

q[Lit

+

1)

- A(0] +

[k^it)Wit)

-

1.

Nit)].

Solving for k^t) yields (1

-

kHt)

q)Lit

+

+

1)

qNit)

(4A.12)

Wit)

subject to the constraint that W^it

+

which eventually depends

2),

on k^t), be nonnegative. Analogously, tive

(4A.

1

1) in

-Nit) and solve for

=

can be obtained by substituting the objec-

k^it)

(4A.2), and setting c q[Lit

+

-

yV(0]

1

to obtain

+ mOWit) -

Nit)],

k^it):

kHt)

Any

1)

=

Lit

+

value of k such that kSit)

WU) 1)


0; otherwise, (limited mobilization of allies)

(pld)X(t),

or

if

X(r)

-

T

(full

1),

idlp)Lit

+

1)




1)

0,

otherwise

=

D*(/)

[1

-

{dlp)][{pld)X{t)

- QUt +

1)],

(limited mobilization)

or

D*(0 =

[Lit

+

1)

-

(full

Since

(1

— Q) =

ing capacities

c/ [c

when

+

X(t)]

-

[1

-

id/p)]QL(t

Xit)IL(t

— +

c)], the difference 1)

>

1)

(4A.33)

mobilization).

qi\

+

between carry-

0 equals the

limit

of the

difference between vote shares as given by (4A.20), assuming is

L

constant.

Would Different Assumptions Make Much Difference? In order to first

be able to answer some questions more generally,

we

introduced modifications of the model. These modifications

Are

134

Socialist

Leaders Vote Maximizers?

are intended to demonstrate that the conclusions presented in the

even

text are valid

if

the

model

reasonable objections that ular assumptions.

suppose

First,

was defined

We

will

may

be raised with regard to the partic-

work only with

that there

allies

it

as that part of socialist voting

may

socialist.

due

allies,

can be made that

We

did not

term q\L{t

to the

was defined

vote

all

allies

+

1)

which

are of

no

inde-

estimated

interest to us here;

the theoretical analyses.

—

is

The

the socialist vote

N{t)\ in (4A.2). Since this

as independent of party strategy, an it

is

are neither workers nor

which we are referring now

protest voting to

of

who

Such votes

indeed, they have been ignored in

by

in the statistically

also turn out that people

would vote

the steady states.

no protest voting. Protest voting

is

pendent of party strategies. Note that

model

accommodate some

altered to

is

argument

should not enter into the analysis of strategies.

remove

it

main body of

in the

the text because the

absolute magnitudes of socialist vote shares are politically impor-

Two

tant.

parties

may have

a range of choice of

one may be getting 25 percent

30 percent but protest vote and the other no

in

protest vote whatsoever. Their alternatives will be electorally the

same but

politically quite different,

were included

Note

anyway

in the analysis.

and

this is

why

But we could remove

protest votes it.

many cases the parameter q turned out to be zero the German and Norwegian Left and the Danish,

that in

—

for

French, and Norwegian socialist parties. In several other cases this

parameter

is

almost negligible. Hence, removing protest vote

from the analysis would not have made much of an empirical difference in most countries. But ical

let

us

first

examine some analyt-

consequences.

If

we

let

X/L

dlp\ otherwise

{plc{)X in the case of limited mobilization,

= L The carrying

in the

case of

and (4A.39)

mobilization.

full

capacities will be different, however,

when

the party

pursues pure class-only strategies. With no protest vote, the carrying capacity

among

and hold

workers, so that

all

allies

becomes zero and

=

D* =

[1

[1

— —

idlp)]L,

if

capacities

X/L

{dlp)]{pld)X

(4A.40)

X.

The difference between carrying

D* =

the party can recruit

>

now

is

dlp\ otherwise

case of limited

in the

mobilization and

D* = L — X

in the

case of

full

mobilization.

(4A.41)

Another reasonable criticism of the model would run as follows.

We

argued that parties lose opportunities among workers because

they decrease the salience of class as they adopt supraclass egies.

But

we used

strat-

N{t) to measure the cumulative effect of these

strategies in equation (4A.1), while in fact N{t) contains protest

vote which

is

not an effect of strategies at

excluded the part of N{t) due to protest impact on workers. Our defense sonably that

if

there are people

because the party is

is

wishy-washy

is

that

who

Thus we should have voting from having an

all.

one might also argue

vote for the socialists,

in its class orientation,

a party of generalized discontent rather than of the

endowed with a project seem quite plausible, let

us

that the protest vote has

no effect on workers.

for the future.

reait

because

is it

working class

But since both arguments

examine the consequences of assuming

Are Socialist Leaders Vote Maximizers?

136 '

In this case,

we would

N* =

\

have, as

in the full

model:

—^^ L + -A— kW*, + q q c

c

among workers would become

but the carrying capacity

W* = X -

— ^ kW*. q +

(d/p)

c

or

=

1

1

1

+

c)]

+

q + C/ k(d/p)[c/(q

+

c)]

X

(4A.43)

more simply,

and, to rewrite

If a

k(dlp)[c/iq

=

yV*

(4A.42)

+

^ p(q +

c;

piq +

c)

+

ckp

(4A.44)

ckd

party followed the pure supraclass strategy,

its

carrying

capacity would be

Y\ = L +

X -

(d/p)

+

q

if

,

X/L

c

>

c

(dip)

q + c (4A.45)

and they would be the same as under other assumptions otherwise. If the

party followed the pure class-only strategy,

capacity would include

its

carrying

the workers and the protest vote of

all

allies

Y*^

Q defined by (4A.30).

= X +

QL,

(4A.46)

The difference between carrying

capacities

would be

D* =

[1

-

(d/p)

^

-

Q]L,

if

X/L

> (d/p)^,

otherwise

D* =

[1

-

{dlp)\{pld)X

— QL

in the

limited mobilization.

case of

Appendix

137

D* =

(1

— Q)L —

X,

in the

case of

full

mobilization.

(4A.47)

Tables 4.3 and 4.4 show the information contained in the

main body of the

text

in table

under the alternative assumptions

4.2

now

shows carrying capacities in the case protest voting is ignored altogether. Table 4.4 shows carrying capacities when protest vote is assumed to have no effect on workers. A comparison of these three tables will show that none of our qualitative conclusions depends on the assumptions which were just reviewed. Even numerical differences are few. The most indiscussed. Table 4.3

teresting finding

is

that

once workers are not affected by protest

voting, the pure class-only strategy turns out to be actually the

Table 4.3

Comparison of Carrying Capacities Associated with Supraclass, and Class-only Strategies

Without Protest Voting (^

Actual^

Strategy

=

— 1970 Class

the Actual,

Structure

0)

Limited

Full

Supraclass

Supraclass^

Class-only

Carrying Capacity

Country

Total Left

max

Belgium

18.9

17.3

Denmark

50.7

60.4

Finland

23.2

37.7

France

24.3

23.7

Germany Norway Sweden

28.5

23.6

65.3

81.0

30.3

41.3

56.3

37.5

Socialists

19.1

28.4

max max max

20.3

24.8 33.3

Only

Denmark

38.0

57.6

Finland

18.8

14.1

France

9.9

2.7

Germany‘S

33.1

2.0

33.1

Norway Sweden

62.8

98.7

30.3

39.5

47.1

37.5

^

Actual strategy

^

A max

in this

is

the strategy estimated to best

column

28.4

max max

fit

20.3 24.8

the model.

indicates that the middle classes together are

more numer-

ous than the proportion of the electorate a party can win with either of the two alternative strategies.

As of

1933.

Are Socialist Leaders Vote Maximizers?

138

Table 4.4

Comparison of Carrying Capacities Associated with Supraclass, and Class-only Strategies Protest Voting by Allies

Actual^

Strategy

— 1970 Class Structure

Has No Effect on Workers

Limited

Full

Supraclass

Supraclass^

Country

the Actual,

Class-only

Carrying Capacity

Total Left

max

Belgium

34.6

21.4

Denmark

60.7

61.8

Finland

48.0

56.7

France

30.0

26.8

Germany Norway Sweden

28.5

23.6

65.3

81.0

30.3

41.7

57.1

37.8

Socialists

23.7

54.7

max max max

45.6 25.3

33.3

Only

Denmark

38.0

57.6

Finland

28.1

16.2

France

9.9

2.7

Germany‘-’

9.9

2.1

33.1

62.8

98.7

30.3

45.5

51.7

42.3

Norway Sweden “

Actual strategy

^

A max

in this

is

28.4

max max

the strategy estimated to best

column

fit

26.4 24.8

the model.

indicates that the middle classes together are

more numer-

ous than the proportion of the electorate a party can win with either of the two alternative strategies.

As of 1933.

vote maximizing one for the pre-1933 allies is

SPD.

If protest

voting by

included but assumed to have no effect on workers, then

will turn out that the

it

Belgian and the French Left as well as the

Finnish Social Democrats have actually gone

some way toward

the full supraclass strategy. But otherwise neither the ordering of the strategies with regard to their vote getting effects nor the

com-

parisons of the optimal and the actual strategies are affected by the variations of these assumptions.

Hence, our conclusions are quite

robust empirically.

Vote Maximizing Strategies Altogether

we can

subject the

model

the text to the following variations:

utilized in the

main body of

Appendix

1.

139

When

is

=

L,

assumption

that the mobilization

of

negative

at

we can

use either the

allies is limited

by the

=

support the party draws from workers and then

{plcf)X or that the party continues to fully mobilize the allies

= L

and then

We

2.

can assume that the party has no control over protest

vote and exclude

We

3.

completely from consideration

it

can assume that workers are not affected by the protest

vote

which are only laborious, we can the following about vote maximizing. The condition that

Without going into sert

details,

equilibrium trade-off, as measured by

dtp

>

1

,

is

ici/p), is

as-

the

steeper than unity,

sufficient to guarantee that the pure class-only strategy

maximizes the carrying capacity of the party in the electorate as a whole in all cases, with the exception of protest voting with no on workers and

effect tion

>

is

1.

slightly

more

mobilization of allies,

full

strict,

namely, {dip)

Formally,

if

{dip)

>

1

>

{q

or,

+

when

the condi-

c)l{q

+

when

c

-

qc)

protest voting

has no effect on workers and allies are fully mobilized, {dip)

>

{q

+

c)l{q

+

c

—

(4A.48)

qc).

Otherwise the pure supraclass strategy maximizes the carrying capacity.

This result shows that there are conditions under which the pure class-only strategy

the one that

is

maximizes the steady-state

share of the electorate and that these conditions hold with a minor

modification

independently

sometimes may be called

What remains sufficient in the

is

to

of particular

which

assumptions

into question.

prove that the condition {dip)

main model

>

1

is

also

to guarantee that the pure class-only

maximize the time path of the socialist share of the The results presented above concerned only the carry-

strategy will electorate.

ing capacity, but

it

may

be possible,

in principle, that in

get to the higher carrying capacity a party

would have

order to

to experi-

would have to pursue a strategy inferior for this period. We will now show that this will not be the case the strategy which maximizes the carrying capacity also maximizes the sum of the votes the party would obtain throughout its

ence a period

in

which

history. This proof

is

it

more

difficult than the rest

of

this

appendix.

Are Socialist Leaders Vote Maximizers?

140

The theorem to be demonstrated is the following: If party leaders seek to maximize the sum of the future shares of the electorate by pursing the same strategy throughout the entire period, and if the voting process is

obeys equations (4A.

through (4A.3), then

1 )

>

maximizing strategy as long as {dip)

the

1

and

maximizing strategy otherwise. The leaders program /

= oc

r

2- iWU)

max k

[W(t}

2! /=

is to:

(4A.49)

N(t)),

1

subject to (4A.1) and (4A.2).

=

+

the

is

The Lagrangian

is:

+ ^(01

(4A.50)

I

+

|jl[X(/

+

1)

+

(1

- p)W(t) -

+

Ll{L{t

+

\)

+

{\

-

q

-

dN{t)

c)N{t)

- W{t +

D]

+ ckW{t) - N{t +

1)]

Since both the objective function and the constraints are linear,

Langrangian with regard

the derivative of the

to the strategic

variable will be either positive or negative:

^ = czn(r)iy(r) ^

(4A.51)

0.

This implies that the maximizing strategy will be one of the ex-

treme strategies, that k^

>

k^,

is,

what we need

SIGN The remaining

dW(0 dNit)

first

either k^ or k^. Since to

prove

it

is

always true

that

that

is

^ = SIGN

[I

-

(4A.52)

(d/p)].

ok

order conditions are

T

p-(/)(l

—

^{t)d

— +

p)

~

n(r)(l

|jl(/

—

—

q

1)

—

+

c)

Ctit)ck

—

fl(t

—

=

1)

0,

=

(4A.53)

0.

(4A.54) Since

we

are interested only in the sign of Sn(r)VT(/), note

only that the equations (4A.53) and (4A.54) can be solved to

Appendix

141

obtain n(/) as an explicit function of time. (1(0 will be of the

form:

= ^

(1(0

"*

(4A.55)

C2ffJ2

complex roots of (4A.53 = 4A.54), Cj’s are constants which depend on initial conditions 11(0) and |x(0), and importantly it can be shown that where the m^'s are

(1*

Note

=

real or

[

1

—

that the sign of (1*

(d/p)]

is

the

P p (, c J= G 3 3 u .3 .ti cc 3 GC £ TOGG G c 3 G o c/5


o,

O' >o,

ro

O £ o o d — '

d d d d d d d

O' X — — o 04 — 04 O o d d d d d d

rr 04 00 O' o rn r- 00 O' O' 00 d d d d d d d

— o O' q X d d d d d d

'O ri Ov

00

oo rj cr oc Cl.

Parameters

o-

c

-o 04 ro O'

00 O' ro ro — o — o q q —

sC ri

Oi

r> k-

"O

c

>'>

c o E -2 E c c c 3 OJ o

OQ

Q

Lju

U.

a Z

c o>

-a

73

C o _2 c c C3 >

c u -o

k.

0)

O Z

c/5

(U

o

The Calculations

191

scriptively impossible or because proportions of lies

associated with voting

left

outside the unit interval. Fit

model

a

A

is

workers and

according to these models range far a very

weak

criterion in evaluating

for a theory.

question which

Table A. 2

Fit

is

important

of the Model and

The naive models

Some

interpreting the numerical

in

of

Its

Naive Competitors

are;

1;

Y{t)

II;

K(/)

III;

Y(t)

IV;

Y(t)

V;

Y(t)

VI;

Y(t)

= niQ + m|X(/) = ihq + m^Yit — 1) = niQ + miX(t) + m2Y(t — 1) — W() + m|X(r) + nuLd) — mo + + m2L(t) + m^Y{l — = niQ + miX(/) + moLtO + m^Kt/ —

Where;

Y(t)

is

the socialist vote share of the electorate,

At/)

is

the proportion of

Lit)

is

the proportion of allies in the electorate, and

r

The

fit

is

al-

=

0,

1,2,...

workers

1)

+ m4t

1)

in the electorate,

represents time.

as follows;

Naive Models

Country

Our Model

II

I

III

IV

V

VI

Total Left

Belgium

0.30

0.05

0.11

0.13

n.a.

n.a.

0.08

Denmark

0.94

0.56

0.93

0.93

0.84

0.93

0.95

Finland

0.73

0.04

0.71

0.70

n.a.

n.a.

0.77

France

0.84

0.42

0.77

0.76

0.38

0.74

0.74

Germany Norway Sweden

0.92

0.67

0.90

0.93

0.92

0.93

0.93

0.98

0.00

0.98

0.98

n.a.

n.a.

0.97

0.89

0.65

0.87

0.87

0.70

0.86

0.93

Socialists

Only

Denmark

0.86

0.37

0.84

0.83

0.82

0.86

0.88

Finland

0.16

0.00

0.22

0.20

n.a.

n.a.

0.21

France

0.40

0.00

0.16

0.20

0.00

0.14

0.14

Germany Norway Sweden

0.74

0.81

0.73

0.83

0.81

0.82

0.82

0.92

0.02

0.91

0.90

n.a.

n.a.

0.92

0.85

0.59

0.83

0.82

0.70

0.82

0.89

Note:

Fit is

models.

measured as

corrected for degrees of freedom for

all

the naive

Appendix

192

results

analyzed

in the text

lated values of parameters.

ask

how much would

the

concerns the precision of the calcu-

One way

fit

to

deteriorate

examine if

this

question

is

to

the value of a particular

parameter was slightly different, say by plus or minus 0.01 or 0.05 away from

its

The

best value.

Table

results are presented in

A. 3 for the three theoretically important parameters: p, d, and k. The explained variance is quite sensitive to the values of p and d; it

is

less sensitive to the values

nonlinearly with the parameter

d

c.

Note

in the

of

which occurs

k,

in particular the sensitivity

model

in the

with regard to

case of the Danish Left and the two Nor-

wegian cases. In Denmark the best value of d was extremely low, namely, 0.002. With

model accounts

for

0.9155

vote. If the parameter

d were

this value, the

percent of the variance of the

taken to be lower by 0.01

,

left

that is, if

we

took the value of -0.008,

model would explain 0.8904 percent of the variance, 2.5 percent less. Hence although the calculated value of d is extremely the

low,

we

are reassured that

it

is

The same

positive.

is

true of

Norway. These

however, with regard

results are less reassuring,

parameter d

in the

to the

case of major socialist parties, where the

tends to be quite insensitive to perturbations.

We

have performed,

We took the parameters

therefore, the following experiment.

fit

from

each country and used them to generate the smooth time series with fit

initial

conditions of another country.

We

then measured the

of this series based on one country’s parameters to the observed

were no differences across

series of the other country. If there

countries in their parameter structure, then the

been the same regardless which

The

set

was

the null hypothesis.

The

crucial distinction to be tested

fit

should have

of parameters was used:

results are presented in table

this

A. 4.

was between Sweden and Denmark, on the one hand, and France and Germany, on the other hand, since several interpretations in the text assumed that the estimated differences between these pairs of countries were nonaccidental. The results strongly support this distinction.

The Data

Two

sets of data

were required

to

conduct the numerical analyses

and estimating computations: the distribution of people into classes

and

their distribution according to votes for parties.

These

re-

The Data

Table A. 3

193

Sensitivity Analysis: Proportion of Variance Explained

Selected Parameters Are

Around

the Best Fitting

-0.05

When

Changed by -0.05, -0.01, +0.01, +0.05 Values Smooth Series

—

-0.01

Best

+ 0.01

+ 0.05

Total Left

Parameter p

-2.5927 -0.5743

0.0882

0.1625

0.1067

-0.8675

0.8904

0.9155

0.8952

0.6510

Finland

0.7224

0.7254

0.7250

0.7243

0.7197

France

0.7988

0.8409

0.8425

0.8414

0.8133

Germany Norway Sweden

0.1639

0.8598

0.8694

0.8440

0.5473

-11.1150

0.7018

0.9765

0.7603

-2.4264

0.5148

0.8653

0.8708

0.8575

0.691

-0.6459

0.1003

0.1625

0.1772

-0.2160

Denmark

0.2399

0.8865

0.9155

0.8943

0.3341

Finland

0.7139

0.7239

0.7250

0.7254

0.7144

France

0.8325

0.8422

0.8425

0.8419

0.8290

Germany Norway Sweden

0.8307

0.8674

0.8694

0.8687

0.8428

-2.0225

0.9011

0.9765

0.9240

0.1338

0.8514

0.8701

0.8708

0.8701

0.8529

Belgium

0.1525

0.161

0.1625

0.1635

0.1647

Denmark

0.9155

0.9155

0.9155

0.9155

0.9153

Finland

0.7250

0.7250

0.7250

0.7250

0.7250

France

0.8425

0.8425

0.8425

0.8425

0.8425

Germany Norway Sweden

0.8690

0.8693

0.8694

0.8694

0.8692

0.9760

0.9765

0.9765

0.9764

0.9759

0.8664

0.8707

0.8708

0.8706

0.8659

0.7493

Belgium

Denmark

Parameter d

Belgium

Parameter k

Socialist

Only

Parameter p

Denmark Finland

France

Germany Norway Sweden

0.5214

0.8309

0.8408

0.8372

-5.9082 -0.5892

0.0557

0.1650

0.0472

-

0.6052

0.6475

0.6134

-0.1018

0.1339

0.6838

0.7044

0.6982

0.5121

-2.1776 -1.2460

0.8648

0.9334

0.8767

0.0625

0.7944

0.8485

0.8059

0.1058

1

.4080

{continued)

Appendix

194

Table A. 3

{Continued)

-0.05

+ 0.05

-0.01

Best

+0.01

0.6794

0.8338

0.8408

0.8374

0.7397

-3.1025

0.0329

0.1650

0.0746

-2.1861

France

0.6381

0.6471

0.6475

0.6472

0.6394

Germany Norway Sweden

0.7037

0.7044

0.7044

0.7044

0.7039

0.7157

0.9263

0.9334

0.9261

0.7846

0.7519

0.8464

0.8485

0.8428

0.7476

Denmark

0.8302

0.8402

0.8408

0.8405

0.8321

Finland

0.1614

0.1648

0.1650

0.1650

0.1623

France

0.4899

0.6421

0.6475

0.6432

0.5550

-7.6785

0.5926

0.7044

0.6393

-0.0776

0.9316

0.9333

0.9334

0.9333

0.9318

0.8447

0.8483

0.8485

0.8485

0.8461

Parameter d

Denmark Finland

Parameter k

Germany Norway Sweden Note:

Fit is

calculated by subtracting the ratio of error variance to vote variance

from unity.

quirements determined our choice of countries, which was totally opportunistic.

We

cases but only in

would have very much the seven countries was

liked to analyze other the time series of the

vote sufficiently long and the information concerning class structure

manageable

to warrant statistical analyses.

Austria, Italy, and Spain,

all

We

could not use

of which had major socialist and

other left-wing parties, because these countries did not have a sufficient

dom

number of

elections.

We

did not use the United King-

because the census information concerning class reported for

England and Wales was separate from cause

it

was reported

gregated fashion.

We

in

that for Scotland

and be-

an extremely detailed or excessively ag-

found

that to reconstruct the British class

would have taken resources we did not have. Finally, we did not study the Netherlands, which would have been thestructure

oretically interesting,

and Switzerland because we ran out of time

and money. To reconstruct the class structure for one country

from the beginning of the century took a team of four persons about six months. Analyses of each case took,

at the

minimum.

The Data

195

Table A. 4

Fit

of the Smooth Time Series Generated by Imposing Parameters

from One Country on Observations from Another Country; Major Socialist Parties

Observations from

Denmark

Sweden

France

Germany

-1.66 -2.64

Parameters from

0.85

0.72

0.33

0.84

-18.07 -24.10

France

-2.93

-2.67

0.65

0.33

Germany

-1.48

-5.78

-0.66

0.70

Sweden Denmark

Note:

Fit is calculated

by substracting the

ratio

of error variance to vote variance

from unity.

about another four months of work. Thus

we were

forced to close

the analysis mostly for practical reasons.

With regard

to class structure the

procedure utilized was differ-

Denmark, France, Germany, and Sweden, where detailed information was available, than for Belgium, Finland, and Norway, where only numbers of workers could be measured. Detailed information concerning the evolution of class structure was gathered for Denmark, 1901-60; France, 1901-68; Germany, 1882-1933 and 1950-61; and Sweden, 1900-1960. Full docu-

ent for

mentation of the sources and an explanation of procedures utilized to reconstruct the information is available

on request. The

ography of research reports and other writings

is

bibli-

attached to this

Appendix.

The data

are unique in a

number of ways. They portray

the

evolution of the entire adult population, not only those recognized

by census takers as economically active or gainfully employed.

The

adult population

is

partitioned into a variety of categories

narrower than social classes. Each category

is

further divided by

The categories, and to same in all countries and

sex and by relation to economic activity. the extent possible their content, are the

over time

in

each country. Finally,

in

most cases we reconstructed

the age categories necessary to determine the class composition of the population eligible to vote in national elections.

Basic data sources consisted of the enumerations of the entire population conducted by the national census bureaus. Although

we have

not attempted to assess the reliability of the original in-

Appendix

196

formation,

we have supplemented

these data with the results of

independently conducted censuses of agricultural and industrial establishments whenever these were available. the censuses of occupations to tinctions as consistent as

Our

make

We

have adjusted

the sectoral and status dis-

was possible within each country.

it

data are based on industrial or collective classifications,

rather than

on occupational or individual distinctions. For exam-

ple,

our category of manual workers

ual

workers

in

industrial

in

industry includes

whether

establishments,

man-

all

they

assembly-line workers, drivers, or Janitors. Altogether

are

we used

twenty-two categories, of which fourteen constitute core categories

01

common

to the four countries.

Manual Workers

in

Industry

Industrial sectors include mining, manufacturing

and handicrafts,

construction, and electric, gas and waterworks.

Manual workers

are those listed in the censuses as Arbeiter, ouvrier, and arbe-

jedere; unless they were obviously misclassified, as were, for ex-

ample, the Swedish barbers,

listed together

with other people

who

cut hair under “leather industry workers.”

02 Manual Workers

in

Transport

Transport includes those sectors involved

goods. Communication sectors

—

moving people and

in

post office, telegraph, etc.

—

are

not included here.

03 Manual Workers

in

Agriculture comprises

Agriculture all

agricultural activities such as forestry,

gardening, fishing, animal breeding, and dairy production. Only those manual workers

only

if

they receive

who work

for a

some payment

separate from their employer, and

wage

in

are included here,

and

cash, live in a household

do not own the land which they

farm. 11

Manual Workers

in

Nonproductive Sectors

Nonproductive sectors comprise above, that

is,

administration.

all

sectors other than those listed

commerce, communication,

Two

services,

and public

nonproductive sectors, however, are counted

separately as categories 15 and 16.

12

Nonmanual Employees

in

Productive Sectors

Salaried employees, typically office and technical personnel, in industry, transport, and agriculture, as defined for categories 01,

02, and 03.

The Data 13

197

Nonmanual Employees

in

Nonproduetive Seetors

Salaried employees, typically office, technical, and sales personnel, in the

nonproductive sectors as defined for category

1

1

14 Supervisory Personnel in Productive Sectors

Used only

in

Sweden. Such persons

are included in category 12 in

other countries.

15 Domestic Servants

16 Public Order Employees Included here are the professional military, police and

customs

nel, forest rangers,

highest ranks,

all

officials,

and the

like.

fire

Except for the

levels of the military are included here,

“manual"

they are identified in the censues as

person-

whether

“nonmanual."

or

21 Nonagricultural Petite Bourgeoisie Includes

all

those in nonagricultural sectors

who own workshops,

do not employ anyone outside

stores, shops, etc. but

their family.

Independent professionals, such as doctors and lawyers, are cluded here even

if

in-

they are employers.

22 Agricultural Petite Bourgeoisie Includes nently

all

those in agriculture

employ anyone outside

gory simply includes

all

who own

land but do not perma-

their family. In

who owned

those

Sweden

this cate-

or rented between 2

and 100 hectares of arable land.

23 Agricultural Bourgeoisie Includes

all

those

who own and/or

control agricultural establish-

ments which employ laborers outside of the family. This category is

not used in

Sweden.

24 Dependent Agricultural Labor Includes agricultural

who

laborers

received payment in kind

and/or room and board from their employer and/or

who were

le-

gally tied to land or the employer. During recent years this category includes all those

cultivating

who own some

land but cannot live only from

it.

31 Nonagricultural Bourgeoisie Includes

all

those

who own

or

manage

nonagricultural establish-

ments which employ wage labor as well as the highest echelons of state officials, if the

den

this

censuses allow their identification. In Swe-

category further includes those

who own

or rent over 100

hectares of arable land.

41 Students Living This

is

in

a Separate Household

just a fraction of students: those

who do

not live in their

Appendix

198

were classified as dependents

parents’ household. Other students

and distributed according to the category of the head of household. In France this category listed as

not used since

is

all

students are

dependents.

42 Dependent on Public Assistance: Permanently Institutionalized

This category shows greater inconsistency countries list

— than any

those

who

other.

receive

It is

— within

and across

usually a residual category used to

public

assistance

tionalized in prisons and asylums and

or

who

are

institu-

do not have any known

occupation.

43 Unemployed This category

used only

is

in

France. Elsewhere the unemployed

are listed in the category of their regular or tion.

This category

is

most recent occupa-

not a measure of unemployment.

The remaining categories change from country Denmark and Germany we have: 49 Rentiers and Retired This category includes rentiers and their

all

to country.

For

those retired, regardless of

occupation before retirement.

For Sweden:

49 Otherwise Unclassifiable Includes a small number of otherwise unclassifiable individuals, typically widows about whom no information is available. 09 Manual Workers Retired from Productive Sectors Includes all those whose preretirement occupations would have placed them in categories 01, 02, or 03. 19 Other Retired

Wage Earners

Those whose occupations before retirement would have placed

them

in categories

29 Retired

beginning with

as the first digit.

1

Petite Bourgeoisie

39 Retired Bourgeoisie and Rentiers For France between 1901 and 1936

we

have:

49 All Dependents For these years

it

was not possible

to distribute

dependents ac-

cording to the occupation of the head of household.

39 Rentiers and All Retired

The Data

199

For France after 1954 we used:

49 Inactive Heads of Households 091 19 Retired Petite Bourgeoisie

39 Retired Bourgeoisie All of these categories, unless otherwise indicated, classified

by sex and by status

—

were cross-

either gainfully

employed or

we added

the categories

dependent.

To 01

,

constitute the category of workers

02, 03, and 09 or

workers were not

some

part of retired people

listed separately.

whenever

retired

Adult dependents were classi-

by the position of the head of the household. Allies were

fied

defined as those listed in the categories 11, 12, 13, and 22 in

Denmark and Germany; in

those listed as

1

1,

12, 13, 14, 19,

and 22

Sweden; and persons listed as 11, 12, 13, 19, and 22 in France. In Belgium, Finland, and Norway we were unable to distin-

guish

among

different groups of nonworkers.

Workers were de-

fined in these countries as those listed in categories 01, 02, and

03, and

all

the remaining adults

were treated as

allies for the pur-

poses of calculating the parameters. The data for these three countries are full

not as reliable as those for the four countries for which

information was available.

In order to calculate proportions

electorate,

we used

in the

the information concerning elgibility criteria

based on age and sex. ria,

of workers and allies

We

could not take into account other

crite-

typically concerning residence.

To

interpolate the census information for the dates of elections,

we used

information derived from the censuses of industrial estab-

lishments whenever possible. These censues typically

who

workers

employed each year in establishments of some They are, therefore, much more sensitive to short-

are currently

minimal

size.

term variations.

yond

list

To

extrapolate the trends in class structure be-

the date of the last census

good reasons

we used

a straight line. There are

to believe that, as a result,

we

overestimated the

proportion of workers in the electorate during the 1970s. to believe

such survey studies, the decrease

in the

If

we

are

proportion of

workers was much more rapid than one would expect from extrapolating earlier trends.

The data concerning election results were derived mainly from Mackie and Rose (1974) and, whenever necessary, supplemented

Appendix

200

or updated from other sources^ are quite Huid,

some

Only

in

France, where party lines

difficulties appeared.

from the Radical Socialists

in

To

separate the

SFIO

1924 we used the distribution of the

them in 1919. This procedure was validated by the number given by Duverger (1965:82): our result was 16.05 and

vote between

1967 and 1968

his 16.21 percent. In

we

left

the

MRG

votes to-

gether with the Socialists. All of the vote shares used in the analysis constitute proportions of the entire electorate, not of the votes actually cast.

The data concerning class teams which were composed of

structure

were reconstructed by

the following persons:

Denmark: Richard Jankowski, Ernest Underhill France: Phillip Jackson,

Adam

Przeworski, Barnett Rubin, Er-

nest Underhill

Germany:

Jankowski,

Richard

Adam

Przeworski,

Ernest

Underhill

Sweden:

Amy

Beth Bridges, Robert Melville,

Adam

Przeworski,

Ernest Underhill

Belgium: Barnett Rubin Finland: Ernest Underhill

Norway: Robert Melville, Ernest Underhill Most of hill. hill

the

work was organized and supervised by Ernest Under-

Final data tables

were cleaned and prepared by Ernest Under-

and Michael Wallerstein.

Bibliography: Documents Available for Distribution Jankowski, Richard. 1978. The evolution of class structure of Germany,

1882-1933 and 1950-61. Data Report. Przeworski,

Adam, and

structure of

Przeworski,

Ernest Underhill. 1978.

The evolution of

class

Sweden, 1900-1960.

Adam,

Ernest Underhill, and Michael Wallerstein.

1978.

The evolution of class structure in Denmark, 1901-60; France, 190168; Germany, 1882-1933 and 1950-61; and Sweden, 1900-1960: Basic data tables. University of Chicago.

Przeworski,

Adam,

Barnett R. Rubin, and Ernest Underhill. 1980.

evolution of class structure of France, 1901-60.

The

Economic Develop-

ment and Cultural Change 28, no. 4 (July):725-52.

The Data

201

Rubin, Barnett, R.

1978.

The evolution of

class structure of France,

1901-68. Data Report. Underhill, Ernest.

1978.

The evolution of

class structure of

Denmark,

1901-60. Data Report. 1978.

Data Report.

The evolution of

class structure of

Sweden, 1900-1960.

I

Bibliography

Abraham, David. 1981 Corporatist compromise and the re-emergence of the labor/capital conflict in Weimar Germany. Political Power and Social Theory 2:59-109. 1982a. Review essay: The S.P.D. from socialist ghetto to postGodesberg cul-de-sac. Journal of Modern History 51:417-50. .

‘Economic Democracy’ as a labor

1982b.

‘Growth Strategy’ in

the

in the

Weimar Republic

.

.

and why

.

post-Godesberg, post-Keynes period.

alternative to the it

will return

Princeton University.

Typescript.

Adam, Gerard. Adam, Gerard

1983. Le pouvoir syndical. Paris: Dunod. et al.

1970. L’ouvrier frangais en 1970. Paris:

Armand

Colin.

Alford, Robert. 1963. Party

and

Rand McNally.

society. Chicago:

1967. Class voting in the Anglo-American political systems. In

Party systems and voter alignments, edited by S.

Rokkan.

New

Allardt, Erik.

M.

Lipset and Stein

York: The Free Press.

1964. Patterns of class conflict and working class con-

sciousness. In Cleavages, ideologies, Erik Allardt. Helsinki:

and party systems,

edited by

Westermarck Society.

1970. Types of protest and alienation. In

by Erik Allardt and Stein Rokkan. Studies

Mass

in Political

politics, edited

Sociology.

New

York: The Free Press. Allardt, Erik,

and

Pertti

Pesonen. 1967. Cleavages

in

Finnish politics. In

204

Bibliography

Party systems and voter alignments, edited by S.

Rokkan.

New

Allardt, Erik,

M.

Lipset and Stein

York: The FreeJ^ress.

and Wtodzimierz Wesotowski, eds. 1978. Social structure

and change: Finland and Poland

in

comparative perspective. Warsaw:

Polish Scientific Publishers.

Althusser, Luis.

1971. Ideology and ideological apparatuses. In Lenin

and philosophy. New York: Monthly Review Press. Andrae, Carl-Goran. 1969. The popular movements and the process of mobilization in Sweden. Social Science Information 8:65-79. 1975. The Swedish labor movement and the 1917-1918 revoluIn

tion.

Sweden’s development from poverty

to affluence, edited

by

Steven Koblik. Minneapolis: University of Minnesota Press. Bain, G. S., and Robert Price. 1980. Profdes of union growth:

A com-

parative statistical portrait of eight countries. Oxford: Basil Blackwell. Balibar, Etienne. 1970. In

Fundamental concepts of

historical materialism.

Reading Capital, edited by Luis Althusser and Etienne Balibar.

New

York: Pantheon Books.

La

Bartolini, Stefano. 1979.

sinistra nei sistemi partitici europei

(1917-

1978): una analisi comparata della sua dimensione e compozisione e dei problem! di sviluppo elettorale. Rivista Italiana di Scienza Politica 9.

Beer, Samuel. 1969. British politics in the collectivist age. 2d ed.

New

York: Vintage Press. 1968. Socialism. International encyclopedia of the social

Bell, Daniel.

sciences. Vol. 15.

New

York: Macmillan.

Berelson, Bernard R., Paul F. Lazarsfeld, and William N. MePhee. 1954. Voting:

A

study of opinion formation in a presidential cam-

paign. University of Chicago Press. Bergier,

J.

The industrial bourgeoisie and the rise of the workThe Fontana economic history of Europe, edited by C. M.

F. 1973.

ing class. In

Cippola. London: Penguin.

Berglund, Sten, and Ulf Lindstrom. 1978. The Scandinavian party system(s). Lund: Studentlitteratur.

Bergounioux, Alain, and Bernard Manin. 1979. La social-democratic ou le

compromis.

Paris: Presses Universitaires

Berlingucr, Enrico. 1973.

The lessons of

de France.

Chile.

The

Italian

Communist

5-6. Bernstein, Eduard. 1961. Evolutionary socialism.

New

York: Schocken

Books.

Birnbaum,

Pierre.

1978. La question des elections dans

cialiste. In Criticpie

and

J.

M. Vincent.

de pratique

politicjue, edited

la

pensee so-

by Pierre Birnbaum

Paris: Gallileo.

1979. Les peuples et les gros: L’histoire d’un myth.

Grasset

et

Fasquelle.

Paris:

Bibliography

205

Bodin, Luis, and Jean Touchard. 1961. Front populaire, 1936. Paris;

Arniand Colin. Boltanski, Luc. 1982. Les cadres:

La formation

d’ un

group

social. Paris:

Lcs Editions dc Minuit. Booth, D. E. 1978. Collective action, Marx’s class theory, and the union

movement. Journal of Economic Issues 12:263-85. Bprrc, Ole. 1974. Denmark’s protest election of December 1973. Scandinavian Political Studies

9:

1977a. Recent trends

197-204. in

Danish voting behavior.

In

Scandinavia

Denmark, Norway, and Sweden, edited by Karl H. Cerny. Washington, D.C.: American Enterprise Inat the polls:

stitute for

Recent political trends

in

Public Policy Research.

1977b. Personal correspondence.

Hans Jorgen Nielsen, Steen Sauerberg, and Torben Worre. 1976. Vaelgere I 70’ Erne. Copenhagen: Akademisk Forlag. Bprre, Ole, and Jan Stehouwer. 1968. Partistyrke og Social Struktur, I960. Aarhus: Akademisk Boghandel.

Bprrc, Ole,

1970. Eire Eolketinksvalg,

1960-1968. Aarhus; Akademisk

Boghandel. Braga, Giorgio. 1956.

II

Communismo fra

gli Italiani.

Milan Comunita.

1973. Le comportement electoral en Erance. Paris:

Brand, Philippe.

Presses Universitaires de France.

Edward. 1922. Die Entwicklung der Arbeiterbewegung

Bull,

den

in

drei

skandinavischen Liindern. Archives fur Geschichte des Sozialismus

10:329-61.

Edward,

Bull,

way around

Jr.

1955. Industrial workers and their employers

in

Nor-

Economic History' Review 3:64-84. Cazals, Remi. 1978. Avec les ouvriers de Mazamet. Paris: F. Maspero. Chodak, Szymon, cd. 1962. Systemy Partyjne W.spolczesnego Kapitalizmu. Warsaw: Ksiazka Wiedza. Christiansen, Niels Finn. 1978. Reformism within Danish social democ1900. Scandinavian

i

racy until the 1930s. Scandinavian Journal of History 3:297-322.

Conradt, David, and Dwight Lambert. 1974. Party system, social structure, sis

and competitive

politics in

West Germany: An ecological analy-

of the 1972 federal election. Comparative Politics 7:61-87.

Converse, Philip E. 1958. The shifting role of class

and

attitudes.

coby, T.

&

In

Readings

Newcomb, and

in social

psychology

E. Matley. 3d ed.

New

in political

behavior

edited by E.

Mac-

York: Holt, Rinehart

Winston. 1969.

Of

time and partisan stability. Comparative Political

Studies 2(July):139-71

Converse, Philip E., and Henry Valen. 1971. Dimensions of cleavage

and perceived party distances litical

Studies 6: 107-52.

in

Norwegian voting. Scandinavian Po-

Bibliography

206

W.

Craig, F.

S., ed. 1969. British

parliamentary election results, 1918-

1949. Glasgow: Political Refe^rence Publications.

Damgaard,

Erik. 1974. Stability

and change

in the

Danish party systems

over a half century. Scandinavian Political Studies 9:103-25. Daniel, Jean. 1978. Editorial. Delruelle, Nicole,

Le Nouvel Observateur.

Paris.

Rene Evalenko, and William Fraeys. 1970. Le com-

portement politique des electeurs beiges. Brussels: Presses univerde Bruxelles.

sitaires

Derfler, Leslie. 1973. Socialism since left.

New

York:

St.

Marx: A century of the European

Martin’s Press.

Desmet, R. E., and Rene Evalenko. 1956. Les elections beiges: Explication

de

la repartition

geographique de suffrage. Brussels:

Institut

de

Sociologie Solvay.

Dogan, Mattel. 1960. Le vote ouvrier en Europe occidentale. Revue Frangaise de Sociologie 1:25-44. 1965. Le vote ouvrier en France: Analyse ecologique des elections

de 1962. Revue Franqaise de Sociologie 6:435-71

Downs, Anthony. 1957. An economic theory of democracy. New York: Harper and Row. Droz, Jacques. 1966. Le socialisme democratique. Paris: Armand Colin. Dupeux, George, with the collaboration of Francois Goguel, Jean Stoetzel, and Jean Touchard. 1958 French national election survey.

Ann

Arbor: Interuniversity Consortium for Political and Social Re-

search (ICPSR)

#7278.

Duverger, Maurice. 1965. Political parties.

New

York: John Wiley

&

Sons. Edel, Matthew.

1979.

A

note on collective action, marxism, and the

prisoner’s dilemma. Journal of Economic Issues 13:751-61

Elvander, Nils. 1979. Scandinavian social democracy:

Its

strength

and

weakness. Stockholm: Almqvist och Wiksell. Engels, Frederick, n.d. cratic draft

A

contribution to the critique of the social demo-

program of 1891 Moscow: Foreign Languages Publishing

House. 1942. The origin of the family, private property,

New

and

the state.

York: International Publishers. 1958. The condition of the working class in England in 1844.

New

York: Macmillan. 1960. Introduction (1895) to Karl Marx, The class struggles in

France, 1848 to 1850. Moscow: Progress Publishers. Engels, Frederick, and Karl Marx. 1935. Correspondence 1846-1895.

New

York: International Publishers. 1975. Social democratic. Edited by Roger Dangeville. Paris:

10/18.

Ensor, R. C. K. 1908.

Modern

socialism as set forth by the socialists in

Bibliography

207

speeches,

their

and programmes.

writings,

New

York:

Charles

Scribner’s Sons.

Esping- Anderson, Gosta. 1979. Comparative social policy and political

advanced welfare

conflict in

Denmark and Sweden.

state:

Interna-

tional Journal oj Health Services 9:269-93.

1985. Politics against markets: The .social democratic road to

power. Princeton: Princeton University Press. Fiechtier, Jean-Jacques. 1965.

Le socialisme fran^ais: De

affaire

1’

Drey-

fus a la grande guerre. Geneva: Librairie Droz. Furet, Francois. 1963.

Pour une definition des classes inferieures a Fepo-

que modeme. Annales: Economies, Societes, Civilisations

Raymond. 1954. Le

Fusilier,

parti socialiste suedois:

18.

Son organisation.

A

Les Editions Ouvrieres.

Paris:

Gay,

Peter. 1970.

lier

The dilemma of democratic socialism.

New

York: Col-

Books.

Geertz, Clifford. 1973. The interpretation of cultures.

New

York: Basic

Books. Geiger, Theodor. 1932. Die Soziale Schichtung des Deutschen Volkes. Stuttgart.

Gerstle, Jacques. 1979.

Goldberg, Samuel.

Le langage des

socialistes. Paris: Stanke.

New

1958. Introduction to difference equations.

York: John Wiley and Sons. Goldthorpe,

J.

H.

1969. The affluent worker in class structure.

et al.

Cambridge: Cambridge University Press.

New

Gomick, Vivian. 1977. The romance of American communism. York: Basic Books.

Gramsci, Antonio.

New

1971. Prison notebooks.

York: International

Publishers.

Green, Nathanael. 1971. Introduction. European socialism since World

War

I.

Chicago: Quadrangle Books.

Guerin, Daniel. 1970. Anarchism:

From

theory to practice.

New

York:

Monthly Review Press. Guinchard, Axel Johan Josef. 1913. Historical and 2 vols. Stockholm: P. A. Norstedt

&

statistical

handbook.

Soder.

Gunsche, Karl-Ludwig, and Klaus Lantermann. 1979. Historia de

Nueva Sociedad. 1981. The German Social Democratic

la in-

ternacional socialista. Mexico:

W. From

Guttsman, 1933:

Hamon, Leo,

L.

Party,

ghetto to government. Boston: George Allen ed.

&

1875Unwin.

1962. Les nouveaux comportements politiques de la

classe ouvriere. Paris: Presses Universitaires de France.

Haupt,

George.

1980.

L’Historien

et

le

mouvement

social.

Paris:

Maspero.

Heady, Bruce. 1970. Trade unions and national wages Journal of Politics 31:407-39.

policies.

The

Bibliography

208

Heidar, Knut. 1977.

The Norwegian Labour

Party; Social

democracy

in a

periphery of Europe. In SociaJ democratic parties in Western Europe, edited by William E. Paterson and Alastair H.

Thomas. Beckenham:

Croom Helm.

A

Helander, Voitto. 1982.

incomes policy system

liberal-corporatist sub-system in action:

in

The

Finland. In Patterns of corporatist policy-

making, edited by Gerhard Lehmbruch and Philippe C. Schmitter.

SAGE.

London:

New

Henderson, Arthur. 1918. The aims of labour. 2d ed.

W.

York: B.

Huebsch. Hentilii,

Seppo. 1978. The origins of the folkhem ideology

in

Swedish

democracy. Scandinavian Journal of History 3:323-45. Keith. 1977. Belgium: Political change in a segmented society.

social Hill,

Electoral behavior:

New

A comparative handbook,

In

edited by Richard Rose.

York; The Free Press.

Comment on Sam Peltzman, Toward a more of regulation. Journal of Law and Economics. 19:2381976.

Hirschleifer, Jack.

general theory 40.

Hobsbawm,

Eric.

1962. The age of revolution.

New

York:

New

Ameri-

can Library. 1964. Labouring men; studies in the history of labour. London:

Weidenfeld and Nicolson. 1973. Revolutionaries. 1978.

New

York:

New American

Library.

The forward march of labour halted? Marxism Today

(September) :279-86.

Horkheimer, Max. 1973. The authoritarian

state.

Telos 15:3-24.

Howard, Dick. 1974. Re-reading Luxemburg. Telos 18:89-107. Hunt, Richard N. 1970.

German

social

democracy 1918-1933. Chicago;

Quadrangle Books. Jaffre,

Jerome.

1980.

The French

electorate in

March 1978.

In

The

French National Assembly Election of 1978, edited by H. R. Penniman. Washington, D.C.: American Enterprise Institute for Public Policy Research. Jaures, Jean. 1971. L’e.sprit

du socialisme.

Johanson, Curt. 1972. Lantarbetarna torica Upsalensia XXXll.

i

Paris:

Denoel.

Uppland, 1918-1930. Studia His-

Uppsala: Scandinavian University Books.

Johansson, Ole. 1967. Socialdemokratins valjare 1911 och 1914. Historisk Tidskrift. Joll,

James. 1966. The second international, 1889-1914.

New

York:

Harper and Row. Katznelson,

England

Ira.

in

1982. Class formation and the state: Nineteenth-century

American perspective. Paper presented

ference on States and Social Structures, Mt. Kisko,

at the

New

SSRC ConYork.

Bibliography

209

New

Kautsky, Karl. 1971. The class

York;

W. W.

Norton.

Kircheimer, Otto. 1966. The transformation of the Western European party systems. In Political parties

and political development,

edited by

Joseph LaPalombara and Myron Weiner. Princeton: Princeton University Press.

Klingemann, Hans-Dieter. 1983. Sozialstrukturelle Bestimmungsgriinde

Veranderungen des Wahlervcrhaltens.

Langfristiger

Basistabellen.

Manuscript.

Konopnicki, Guy. 1979. Vive

centenaire du P.C.F. Paris: Editions

le

Libres.

Krantz, Olle, and Carl- Axel Nilsson. 1975. Swedish national product,

1861-1970. Lund. Ladyka, Teodor. 1972. Polska Partia Socjalistyczna (Frakcja Rewolucyjna)

w Latach

1906-1914. Warsaw; Ksiazka

Lafferty, William A. 1971.

i

Wiedza.

Economic development and

the response of

labor in Scandinavia. Oslo: Universitetsforlaget.

Landauer, Carl. 1959. European socialism. 2 vols. Berkeley: University of California Press.

The Guesdists and the small farmer: Early erosion of French marxism. International Review of Social History 6:212-25. 1967. The origins of socialist reformism in France. International Review of Social History 2 ;8 - 07. 1961.

1

Laski, Harold.

1935.

Democracy

1

1

Chapel

in crisis.

Hill:

University of

North Carolina Press.

Lehmbruch, Gerhard. 1982. Neo-corporatism and

the function of repre-

sentative institutions. Paper presented at the Conference tation

on Represen-

and the State, Stanford University.

Leser, Norbert. 1976. Austro-Marxism:

temporary History Levin, Leif,

Bo

11:1

A

reappraisal. Journal of

Con-

33-48.

Jansson, and

Dag Sorbom.

1972. The Swedish electorate

1887-1968. Stockholm: Almqvist och Wiksell. Lipset, Martin S. 1963. Political

The changing

1964. politics.

man. Garden

class structure

City:

Doubleday.

and contemporary European

Daedalus 93:271-301.

Lockwood, David. 1960. The ‘new working

class’.

Archives Europeens

de Sociologie 1:248-60. Lukacs, Gyorgy.

Mass.

MIT

1971. History

and

class-con.sciousness.

Cambridge,

Press.

Luxemburg, Rosa. 1967. The Russian revolution and Leninism or Marxism? Ann Arbor: University of Michigan Press.

The mass strike, the political party, and Rosa Luxemburg Speaks, edited by M. A. Waters. 1970.

In

finder Press.

the trade unions.

New

York: Path-

Bibliography

210

Lyman, Richard. 1957. The

Chapman and

first

Labour government, 1924. London;

Hall.

Mabille, Xavier, and Val R. Lorwin. 1977. In Social

democratic parties

in

The Belgian

Western Europe, edited by William E.

Paterson and Alastair H. Thomas. London;

McDonald,

Ian,

socialist party.

and Robert Solow. 1981.

Croom Helm. Wage bargaining and em-

ployment. American Economic Review 71;896-908.

McKibbin, Ross. 1974. Evolution of the Labour Party. London; Oxford University Press.

Mackie, Thomas and Richard Rose. 1974. The International Almanac of Electoral History.

New

York; Free Press.

McLellan, David. 1971. The thought of Karl Marx.

New

York; Harper

and Row.

Mandel, Ernest. 1971. The formation of the economic thought of Karl Marx. New York; Monthly Review Press. ^Alarcus, Steven. 1975. Engels, Manchester and the working class.

New

York; Vintage Books.

Markham, James M. 1983. Blue collar Ruhr Valley moves to the right. International Herald Tribune 26-27 March. Martin, Andrew. 1975. Is democratic control of capitalist economies possible? In Stress and contradiction in modern capitalism, edited by Leon N. Lindberg et al. Lexington; Lexington Books. Martin, Penny Gil. 1972. Parts' strategies and social change: The Norwegian Labour Party. Ph.D. diss., Yale University. University microfilm #72-31450. Marx, Karl. n.d. The poverty of philosophy. Moscow; Progress Publishers.

1879. Interview. The Chicago Tribune. 5 January. 1952. The class struggles in France, 1848-1850.

Moscow;

Pro-

gress Publishers.

1960. The eighteenth brumaire of Louis Bonaparte.

Moscow;

Progress Publishers. 1974. Inaugural address to the International sociation. In

The First International and

Working Men’s As-

after, edited

by David Fer-

nbach. London; Penguin.

Marx, Karl, and Frederick Engels. 1967. The communist manifesto. Edited by Harold Laski.

New

York; Pantheon Books.

1969. Selected works in three volumes

(MESW). Moscow;

Pro-

gress Publishers.

Numero Hors Serie. Bargaining: Monopoly power vi’. union power.

Le Matin. 1978. Le dossier des Menil De, George. 1971

.

legislatives 1978.

Cambridge; Harvard University Press. Michelat,

Guy, and Michel Simon.

1975.

Categories

socio-profes-

Bibliography

21

sionelles en milieu ouvrier et

comportement

1

Revue Fran-

politique.

qaise de Science Politique 25:291-316.

1977. Closse, religion et comportement politique. Paris: Editions Sociales-Messidor.

Michels, Robert.

1962. Political parties:

oligarchical tendencies of

A

sociological study of the

modern democracy

New

York: Collier

Books.

New

Miliband, Ralph. 1970. The state in capitalist society.

York: Basic

Books. 1972. Parliamentary socialism:

A

study in the politics of labour.

2d ed. London: Merlin Press. 1977.

Marxism and

politics.

Oxford: Oxford University Press.

and protest: The and the origins of

Mitchell, Harvey, and Peter N. Steams. 1971. Workers

European labor movement, the working classes, social democracy, 1890-1914. Itasca, Illinois:

Peacock

E.

F.

Publishers.

Le Monde. 1973. Les forces politiques

et les elections

de mars 1973.

Neimark. M. A. 1976. Belgiiskaya Socjalisticheskaya Partya: Ideologia i

Politika.

1945-1975. Moscow: Izdatelstvo ‘Nauka’.

Newmann, Sigmund, comparative Nicolaus,

ed. 1956.

politics.

Martin.

Modern

political parties;

approaches

to

Chicago: University of Chicago Press.

1967.

Proletariat

and the middle class

in

Marx:

Hegelian coreography and the capitalist dialectic. Studies on the Left

7:22-49. Offe, Claus, and Helmuth Wiesenthal. action: Theoretical notes

Political

power and

on

1980.

social class

Two

logics of collective

and organizational forms.

In

social theory, edited by Maurice Zetlin. Green-

wich: JAI Press.

Okasaki, Ayanori. 1958. Histoire du Japon: L’ Economie et la population. Paris: Presses Universitaires

Olson, Mancur,

Jr.

de France.

1971. The logic of collective action. Rev. ed.

New

York: Schocken Books. Organization for Economic Co-operation and Development. 1979. Collective Paris:

bargaining and government policies

in ten

OECD

countries.

OECD.

Panitch, Leo. 1981. Trade unions and the capitalist state.

New

Left Re-

view 125:21-46. Pappi, Franz U.

1973. Parteiensystem und sozialstruktur in der Bun-

desRepublik. Politische Vierteljahresschrift 14:191-213. 1977. Sozialstruktur, gessellschaftliche wertorientierungen und

wahlabsicht. Politische Vierteljahresschrift 18:195-230. Parti

Communiste

Fran^ais. 1971

Le capitalisme monopoliste

.

Traite marxiste d’ economie politique:

d’Etat. 2 vols. Paris: Editions Sociales.

Bibliography

212

Paterson, William E.

1977.

The German

Social Democratic Party. In

Social democratic parties in Western Europe, edited by William E.

Paterson and Alastair H. Thomas. London: Paterson, William E. and A. H.

Thomas,

Parties in Western Europe. London:

Pesonen,

1968.

Pertti.

reactions.

New

An

Groom Helm.

eds. 1977. Social

Democratic

Groom Helm.

election in Finland: Party activities

and

voter

Haven: Yale University Press.

1974. Finland: party support in a fragmented system. In Electoral Behavior:

A comparative handbook.

Ed. by Richard Rose.

New

York: The Free Press. Petersson, Olaf, and

Bo

Sarlvik. 1975.

The 1973

election.

Vol 3 of Gen-

eral Elections 1973. Stockholm: Gentral Bureau of Statistics.

Pizzomo,

Alessandro.

1966.

Introduzione

alio

della

studio

par-

tecipazione politica. Quaderni di Sociologia 15:235-86.

1978. Political exchange and collective identity

in

industrial

The resurgence of class conflict in Western Europe since 1968, edited by Golin Grouch and Alessandro Pizzomo. London:

conflict. In

Macmillan. Plekhanov, Georgij V. 1965. Sochynenya. Vol. Poulantzas, Nicos.

New

1973. Political

power and

11.

Moscow.

social classes. London:

Left Books.

1974. Les classes sociales dans

le

capitalisme aujourd’hui.

Paris: Seuil.

Przemiany 1963.

w

strukturze klasy robotniczej

Warsaw: Ksiazka

Adam.

Przeworski,

i

w

krajach kapitalistycznych.

Wiedza.

1975. Institutionalization of voting patterns, or

is

mobilization the source of decay? American Political Science Review.

69:49-67. Przeworski,

Adam,

Ernest Underhill, and Michael Wallerstein.

1978.

The evolution of class structure in Denmark, 1901-60; France, 190168; Germany, 1882-1933 and 1950-61; and Sweden, 1900-1960: Basic data tables. University of Ghicago. Typescript. Rabier, Jean-Glaude. 1978.

On

the political behavior of French workers.

Acta Sociologica 21:355-70. Rabier, Jean-Glaude, and Ronald Ingelhart. nities Study.

Ann

Social Research

1973. European

Commu-

Arbor: Intemniversity Gonsortium for Political and

(IGPSR) #7330.

Ranger, Jean. 1969. L’evolution du vote communiste en France depuis 1945. In Paris:

Le communisme en France, edited by Frederick Bon

Armand

et al.

Golin.

Riker, William H. 1962.

The theory of political

coalitions.

New

Haven:

Yale University Press. Rimbert, Pierre. 1955. Partis politiques Paris:

et classes sociales

Fondation des Sciences Politiques.

en France.

Bibliography

213

Roemcr, John E. 1979. Mass action

not individually rational: Reply.

is

Journal of Economic Issues 13:763-67. Rokkan, Stein. 1966. Norway: Numerical democracy and corporate pluralism.

Political opposition

In

New

Robert Dahl.

cleavages

Haven: Yale University Press.

Geography,

1967.

Norwegian

in

western democracies, edited by

in

ments, edited by S.

M.

and social

religion,

Crosscutting

systems and voter align-

In Part}'

politics.

class:

New

Lipset and Stein Rokkan.

York: The Free

Press.

Rokkan, Stein, and Henry Valen. 1962. The mobilization of the periph-

Approaches

ery. In

Stein

to the study

of political participation, edited by

Rokkan. Bergen: Michelsen

Institute.

Rose, Richard. 1980. Electoral participation: Beverly

Hills:

A comparative

analysis.

Sage Publications.

Rose, Richard, and Derek Urwin. 1969. Social cohesion, political parties

and strains

regimes. Comparative Political Studies 2:7-67.

in

Rosenberg, Arthur. 1965. Democracy and socialism. Boston: Beacon Press.

Roth, Guenther.

1963. The social democrats in imperial

study in working class isolation

and national

Germany: A Totowa:

integration.

Bedminster Press. Sainsbury, Diane. 1980. Swedish social democratic ideology toral politics, Siirlvik,

Bo.

elec-

1944-1948. Stockholm: Almqvist och Wiksell.

1960. Swedish national election survey.

teruniversity

and

Consortium for

Political

Ann

Arbor: In-

and Social Research (ICPSR)

#7366.

ICPSR

1964. The social determinants of voting in Sweden.

#7339. 1966. Political stability and change

Swedish

in the

electorate.

Scandinavian Political Studies 1:188-222. 1967. Party politics and electoral opinion formation: issues in

Swedish

politics,

A

study of

1956-1960. Scandinavian Political Studies

2:167-202. 1969. Socioeconomic determinants of voting behavior

Swedish

electorate.

in

the

Comparative Political Studies 2:99-135.

1970. Voting behavior

in shifting ‘election

winds’:

An overview

of the Swedish elections, 1964-1968. Scandinavian Political Studies

5:241-83. 1974. Sweden: tal

The

social bases of the parties in a

perspective. In Electoral behavior:

ed by Richard Rose.

New

A comparative handbook,

edit-

York: Free Press.

1977. Recent electoral trends polls, edited

developmen-

in

Sweden.

In

Scandinavia at the

by Karl Cerny. Washington, D.C.: American Enterprise

Institute for Public Policy

Research.

214

Bibliography

George

and

1958. The communists

Sartre, Jean-Paul.

the peace.

New

York:

Braziller.

1960. Critique de la raison dialectique. Paris: Gallimard. 1973. Elections, piege a cons. Les

Temps Modernes 318:1099-

1108.

Scase, Richard. 1977. Social democracy in capitalist society. London:

Groom Helm. ^-Schelling, Thomas C. 1978. Micromotives and macrobehavior. New York: W. W. Norton. The primary goals of political parties: A clarifitheory. American Political Science Review. 68:840-

Schlesinger, John. 1975. cation of positive

49.

Schmitter, Philippe C.

New Corporatism,

1974.

edited by Frederick Pike and

Dame: University of Notre Dame Schmitter, Philippe C.

Thomas

Stritch.

1984. Neo-corporatism and the state.

German

Notre

Press.

Paper no. 106. Florence: European University Schorske, Carl E. 1972.

The

the century of corporatism? In

Still

social

Working

Institute.

democracy 1905-1917 : The de-

velopment of the great schism. New York: Harper and Row. Schumpeter, Joseph. 1942. Capitalism, socialism, and democracy.

New

York: Harper and Row.

Schwerin,

Don

1980. The limits of organization as a response to

S.

wage-price problem. In Challenge to Governance, edited by Richard Rose. Beverly Hills: Sage. Sellier, Francois.

1976. Les problemes du travail en France 1920-74.

Rapport au 4^ congres mondial de I’AIRP. Geneva. Typescript.

The transformation of Austrian socialism. University of New York Press.

Shell, Kurt. 1962.

State

Skirda, Alexandre. tuels, edited

1979. Presentation, in

by Jan Waclav Makhaiski.

Soderpalm, Sven Anders. 1975. The ocratic road to power. In

ence,

crisis

New

Le socialisme des

York:

intellec-

Paris: Seuil.

agreement and the social dem-

Sweden’s development from

poverty' to afflu-

1750-1970, edited by Steven Kublik. Minneapolis: University

of Minnesota Press.

SOFRES.

1981.

Soikkanen, Hannu. 1978. Revisionism, reformism and the Finnish bour movement before the

first

la-

world war. Scandinavian Journal of

History 3:347-60.

Sombart, Werner. 1909. Socialism and the social movement. London:

M. Dent

&

Sons.

Sondages Revue Fran^dise de Editions

/’

opinion publique. 1960, 1973. Paris:

Le Chancelier.

Stehouwer, Jan. 1967. Long term ecological analysis of electoral tics in

J.

Denmark. Scandinavian

Political Studies 7.'.9A-

\

17.

statis-

Bibliography

215

Stephens, John D. 1981

The changing Swedish

.

electorate: Class voting,

contextual effects, and voter volatility. Comparative Political Studies

14:163-204. Streeck, Wolfgang.

economic

crisis in

1984. Neo-corporatist industrial relations and the

West Germany.

In

Order and Conflict

Contem-

in

porary' Capitalism, edited by John H. Goldthorpe. Oxford: Clarendon Press.

Sully, Melanie A. 1977.

The

socialist party of Austria. In Social

cratic parties in Western Europe, edited

Alastair H.

demo-

by William E. Paterson and

Thomas. London: Croom Helm.

Svensson, Paul. 1974. Support for the Danish social democratic party

1924-1939

— growth

and response. Scandinavian Political Studies

9:127-46. Sylos Labini, Paolo. 1972. Sviluppo economico e classi social!

in Italia.

Quaderni de Sociologia 21:371-443. Tarschys, Daniel. 1977. The changing basis of radical socialism

dinavia. In Scandinavia at the polls, edited by Karl H. Cerny.

ington, D.C.:

American Enterprise

Tawney, R. H. 1923. The

Institute of Public Policy

British labour

movement.

New

Scan-

in

Wash-

Research.

York: Yale

University Press. Theil, Henry. 1976. Econometrics.

Thomas,

Alastair H. 1977.

democratic parties

in

New

The Danish

E. P. 1963.

&

Sons.

social democratic party. In Social

western Europe, edited by William E. Paterson

and Alastair H. Thomas. London:

Thompson,

York: John Wiley

Croom Helm.

The making of the English working

class.

New

York: Vintage. Tingsten, Herbert. 1973. The Swedish social democrats. Totowa: Bed-

minster Press.

Tomasson, Richard. 1969. The extraordinary success of cial democrats. Journal of Politics 31:772-98. 1970. Sweden: Prototype of

dom

modern

society.

the

New

Swedish

so-

York: Ran-

House.

Touchard, Jean. 1977. La gauche en France depuis 1900. Toutain, J-C. 1963. La population de

la

Paris: Seuil.

France de 1700 a 1959. Cahiers

de Tlnstitut de Science Economique Appliquee. Serie AF. 3:3-247. Urry, John. 1973. Towards a structural theory of the middle class. Acta

Sociologica 16:175-87. Usitalo,

Hannu. 1975. Class structure and party choice: A Scandinavian

comparison. Helsinki: University of Helsinki Research Reports, no. 10

.

Valen, Henry, and Willy Martinussen. 1977. Electoral trends and foreign politics in

Norway: The 1973 Storting election and

the

EEC

issue. In

Scandinavia at the polls, edited by Karl H. Cerny. Washington, D.C.:

American Enterprise

Institute of Public Policy Research.

Bibliography

216

1983. Dimensions of union growth in postwar Western

Visser, Jelle,

Working Paper

Europe.

no.

Florence:

89.

European

University

Institute.

Wallerstein, Michael. 1983. ior.

Working

class solidarity

University of Chicago. Typescript.

Windmuller, John

P.

1975.

The

authority of national trade union con-

federations: a comparative analysis. In icy,

and rational behav-

edited by D. B. Lipsky. Ithaca,

Windmuller, John international

P.

Union Power and Public Pol-

NY:

Cornell University.

1981. Concentration trends

in

union structure: an

comparison. Industrial and Labor Relations Review.

35:43-57.

Wittman, Donald A. 1973. Parties as ical

utility

maximizers. American Polit-

Science Review. 67:490-98. 1983. Candidate motivation:

American

Political Science Review.

Wright, Erik Olin. cieties.

New

Left

A

synthesis of alternative theories.

77:142-57.

1976. Class boundaries in advanced capitalist so-

Review 98:3-42.

Name

Index

Abraham, David, 71

Bodin, Luis, 82

Adam, Gerard,

B0rre, Ole, 72, 93, 150, 154,

77, 172

Alford, Robert, 5, 145, 166

171, 177

Allardt, Erik, 87, 150, 154

Braga, Giorgio, 87

Allende, Salvador, 83

Branting, Hjalmar, 16, 32, 57

Althusser, Luis, 8

Braud, Philippe, 151, 154, 155,

Andrae, Carl-Goran, 79

171 Bull,

Edvard

Jr.,

33

Bain, G. S., 77 Balibar, Etienne,

78n

Bartolini, Stefano,

30

Bebel, August, 16, 28 Beer, Samuel, 41 Bell, Daniel,

83

Cazals, Remi, 49

Chodak, Szymon, 22 Converse, Philip, 6, 61, 153, 156 Craig, F.

W.

S.,

57

Berelson, Bernard R., 5 Bergier,

J.

F., 31

Berglund, Sten, 150, 153, 154, 156, 171

Damgaard, Erik, 150, 154, 172 Daniel, Jean, 125

Danielsson, Alex, 15, 20

Bergounioux, Alain, 46, 49

Deat, Marcel, 43, 125

Berlinguer, Enrico, 52, 109

Derfler, Leslie, 25, 28, 41

Bernstein, Eduard, 21, 23, 24,

Downs, Anthony, 125

29, 41, 42, 44, 50, 75, 179

Droz, Jacques, 14

Birnbaum, Pierre, 57, 183

Dupeux, George, 154

Bismarck, Otto von, 181

Duverger, Maurice, 83, 200

Name

218

Elvander, Nils, 116, 118

Jaffre,

Engels, Frederick, 8, 17, 20, 26, 29, 32, 40, 46,

Ensor, R. C. K.,

24-

48-50, 79

22-25

Index

Jerome, 151, 155, 171

Jaures, Jean, 21, 23, 24, 41, 51

Johansson, Ole, 159 Joll,

James, 15

Esping- Anderson, Gosta, 61, 62, 116, 150, 152-56, 161, 171, 172, 174, 177

Kautsky, Karl, 23, 32, 41-43, 50, 79

Klingeman, Hans-Dieter, 151, Fiechtier, Jean-Jacques, 17, 23,

Konopnicki, Guy, 20

49, 118

Krantz, Olle, 37

Furet, Francois, 31 Fusilier,

160, 162, 177

Raymond, 50 Lafferty, William, 17

Gay, Peter, 21, 24, 118, 179

Landauer, Carl, 20, 41, 42, 50

Geertz, Clifford, 182

Laski, Harold, 21

Gerstle, Jacques, 72

Lembruch, Gerhard, 77

Goldberg, Samuel, 98

Lenin, V.

Goldthorpe,

Leser, Norbert, 16

J.

H., 102

Gomick, Vivian, 51 Gramsci, Antonio,

I.,

24

Lindstrom, Ulf, 150, 153, 154,

1, 7, 8,

21,

156, 172 Lipset, S. Martin, 5, 7, 102, 162

47, 51, 73, 126, 143

Green, Nathanael, 71

Lockwood, David, 182

Guerin, Daniel, 15

Lorwin, Val, 28

Guesde, Jules, 23, 41, 49

Lukacs,

Guinchard, Axel, 48

Luxemburg, Rosa,

Hansson, Per Albin, 24, 51

Mabille, Xavier, 28

Hardie, Keir, 36

Helander, Voitto, 75

McDonald, Ian, 54 MacDonald, Ramsay, 41 McGurk, John, 20 McKibbin, Ross, 41 Mackie, Thomas, 199

Henderson, Arthur, 56

McLellan, David, 16

Haupt, George, 15, 42, 59

Heady, Bruce, 75 Heidar, Knut, 44, 79,

Hentilla,

1

18

Seppo, 51, 118, 121

Gy orgy,

3, 19

17, 18

Malatesta, Errico, 15

Hill, Keith, 87, 150, 161

Man, Henrik

Hirschleifer, Jack, 123

Mandell, Ernest, 32

Hobsbawm,

Eric, 21, 32,

Horkheimer, Max,

2,

34n

18

Howard, Dick, 18 Hunt, Richard, 42, 43, 71, 118

de,

43

Manin, Bernard, 46, 49 Marcus, Steven, 32

Markham, James M., 57 Martin, Andrew, 109 Martin, Penny Gil, 50,

Ingelhart, Ronald, 155

152, 155

1

15,

1

18,

Name

Index

219

Martinussen, Willy, 151, 152, 155

Rose, Richard, 6, 199

Rosenberg, Arthur, 49

Marx, Karl,

1, 8,

14-17, 26, 32,

40, 41, 43, 46-48, 54, 79, 87,

Sainsbury, Diane, 104, 105

181

Sarivik, Bo, 62, 64, 87, 152,

Menil, George de, 54

155, 156, 159, 172, 174, 177

Michels, Roberto, 8, 19, 20, 41,

109

Sartre, Jean-Paul, 9, 18,

Scase, Richard,

1

18

Miliband, Ralph, 19, 20

Schelling,

Millerand, Alexandre, 13, 22, 23,

Schlesinger, John, 125

25, 42

53

Thomas, 54

Schmitter, Philippe, 78

Mitterand, Francois, 162 Mitchell, Harvey, 77

Schorske, Carl, 16, 17, 27, 118, 119

Neimark, M. A., 41

Schumaher, Kurt, 44, 82 Schumpeter, Joseph, 4

Neumann, Sigmund, 83

Schwerin,

Nicolaus, Martin, 43

Sellier, Francois,

Nillson, Carl-Axel, 37

Soderpalm, Sven Anders, 83, 166 Solow, Robert, 54

Okasaki, Ayanori, 34

Sombart, Werner, 31

Olson, Mancur

Stehouwer, Jan, 72

Jr.,

81

Stein,

Don

S.,

75

77

Lorenz von, 31

Paepe, Cesar de, 20, 42, 50

Stephens, John, 64, 153, 161

Panitch, Leo, 77

Streeck, Wolfgang, 75

Pappi, Franz von, 62, 149, 151,

Sully, Melanie,

84

Sylos Labini, Paolo, 34

162, 171

Paterson, William, 44, 82

Pesonen,

Pertti,

149, 150, 154

Tarschys, Daniel, 61, 72

Petersson, Olaf, 152, 156, 172

Tawney, R. H., 18

Pizzomo, Alessandro,

Theil, Henri, 181

9, 123

Plekhanov, Georgij V., 101, 102

Alastair, 28, 82, 83

Thorez, Maurice, 82

Poulantzas, Nicos, 19, 78n Price, Robert,

Adam,

Tingsten, Herbert, 15-17, 20,

22-24, 26, 32, 41, 42, 49, 50,

77

Proudhon, Pierre-Joseph, 14 Przeworski,

Thomas,

61, 77

51, 53, 121 Togliatti, Palmiro, 51

Tomasson, Richard, 44 Rabier, Jean-Claude, 151, 154, 155, 171

Touchard, Jean, 51, 82, 125 Toutain, J-C., 34

Riker, William, 109

Rimbert, Pierre, 72

Urry, John, 43

Rokkan, Stein, 64

Urwin, Derek, 6

Name

220

Uusitalo,

Hannu, 150, 152, 154,

155, 171

Wallerstein, Michael, 54

Wesotowski, Wtodzimierz, 150, 154

Valen, Henry, 64, 151-53, 155, 156 Visser, Jelle, 75, 77, 78

Vollmar, Georg von, 24

Index

Windmuller, John, 75

Wittman, Donald, 124, 125 Wright, Erik Olin, 57

Subject Index

Address

to the

Communist League

Manifesto of

1959, 57

(1850), 48 Allies: definition of,

among

portion

16; Electoral

41-42; pro-

adults, 45;

toward, 50-

socialist strategy

Calculation procedures, 65-66, 85, 187-89;

51, 80-81 Anarchists: Congress at

de-Fonds 1870,

14;

toral participation,

Chaud-

and elec-

14-15

Austrian Socialist Party: electoral progress of, 27, 29-30; Linz

Programme of 1926, 16

fit

of the model,

189-90; sensitivity of parameters,

191-95

among actual among allies, 88;

Carrying capacity: voters, 94;

among workers,

69; associated

with pure strategies, 131-33; cross-national differences of,

94-96;

in the electorate,

89-94

Belgian Parti Ouvrier, 20; electoral

progress of, 27, 29-30;

founded

in

1885, 50; Labor

Plan of 1933, 43;

Programme

Class structure, evolution

cording to Marx, 26; reconstructed for various countries,

of 1894, 41

Belgian Socialist Party,

49-50

of: ac-

192-201

Communist Manifesto,

8, 26, 33,

167

Brabant Socialist Party (Parti Socialiste

Labour Party, 41; decision form government in 1924,

British to

Braban^on), 20, 42 Daily Tribune

26

(New York,

1850),

Subject Index

222

Danish Social Democratic Party;

sured, 68; equilibrium trade-off

and vote maximizing

actual

compared, 110, 114,

strategies

117; electoral progress of,

and cumulative trade-off mea-

28-

measured, 70; immediate tradeoff measured, 67; and other

30; electoral strategies of, 86;

parties,

73-74; and trade

and farmers, 115; Programme

unions,

74-78

of 1888, 42; range of choice available to, 108-9; trade-off

faced by, 72-73; and trade unions,

1

Finnish Social Democratic Party; actual and vote

15-19

strategies

Dutch Socialist Party, electoral progress of, 28-31

maximizing

compared,

1

10,

electoral strategies of, 86;

Electoral participation, 13-25;

anarchist views of, 14-15; ef-

on party

strategies,

20-22,

184; effect on workers’ organization, 18-20; socialist of,

views

15-18

differences they

make, 106-9,

128-31; effectiveness of egies measured, 86;

strat-

how

108-9; trade-off faced by, 73 First International (International

Workingmen’s Association); Hague Congress of, 22; Marx’s Inaugural Address to, 48

choose them, 80-81, of,

82-84; and intertemporal trade120-126; language

50-51; limited

French Socialist Parties (Parti Socialiste Fran^aise,

Socialiste), 43, tual

84-86, 103-6; instruments

offs,

to,

French Communist Party, 72

Electoral strategies, socialist; and

parties

14;

electoral progress of, 27, 30;

range of choice available

fect

1

SFIO,

Parti

82-83, 182;

and vote maximizing

strategies

compared, 110, 114;

electoral progress of, 30; electoral strategies of, 86;

intertemporal trade-offs ana-

of,

lyzed, 121-26; Marseille

vs. full su-

pure strategies defined, 104-5,

Programme of 1892, 42; Nantes Programme of 1894,

127-28; strategies of various

42; range of choice available

praclass strategies, 110-12;

parties, 85,

1

ac-

10-26; and trade

to,

108-9; Tour Programme of

unions, 118-19; with regard to

1902, 23; trade-off faced by,

craftsmen, artisans, and small

72

merchants, 42; with regard to farmers, 42, 165-66; with re-

gard to salaried employees,

43-

44, 178-79; with regard to

women and

youth, 44

Electoral trade-off,

66; and tion,

55-56, 60-

communist competi-

71-73; cross-national

differences of, 73-79; delayed

German Communist

Party (KPD),

71

German

Social Democratic Party

(SPD), 3, 17, 80, 82, 182; actual

and vote maximizing

compared, 110, 114; electoral progress of, 27-30, strategies

64; electoral strategies of, 86,

Subject Index

223

88; Erfurt Congress of 1891,

Programme of 1891,23,33,41,49;Gorlitz Programme of 1921, 51; Gotha Programme of 1875, 49; Programme of 1927, 42-43; range 24, 28; Erfurt

of choice available

to,

108-9;

Swedish Peoples’ Party

in

Finland, 74

Swedish Social Democratic Party (SAP), 20, 50, 53, 68, 82-83; actual

and vote maximizing

strategies

compared, 110, 114,

117; discussion of electoral par-

Resolution of the Eisenach

ticipation,

Congress 1870,

campaigns, 104-5; electoral

16; trade-off

faced by, 71

15-17; electoral

progress of,

G1o5 Wolny (London, 1867), 26

28-3

1

,

64; elec-

toral strategies of, 86,

1

15;

intertemporal trade-offs ana-

International Herald Tribune

(1983), 57

Labour and

the

1889, 16,

New

Programme of 22,41,49; Pro-

lyzed, 120-26;

Social Order

(1918), 57

gramme of 1897, 23; Programme of 1911, 42, 50; Programme of 1920, 41; range of choice available

Norwegian Labor Party, 80; actual strategies

electoral

17, 50,

and vote maximizing

compared, 110, 114; progress of, 28-30;

electoral strategies of, 86,

range of choice available

1

15;

to,

to,

108-9;

trade-off faced by, 70, 72; and trade unions,

115-19

Trade Unions, 74-77; and tradeoffs facing parties,

77-79,

115-19

108-9; trade-off faced by, 7273; and trade unions,

116-19

Voting, and class, 5-6; model

and surveys compared, 147-

The Origins of Private Property, Family, and the State, 17

Protest vote, 87,

58; theory of,

9-11, 141-47

Voting, by middle classes: of

Beamte und Angestellte in Germany, 61; of farmers in various

134-38

countries, 165, 168-70; histor-

Salaried employees, as proportion

of labor force, 43

Second International, Socialist goals,

toral

various

countries, 164-67; impact of 15, 17

22-23

Stagnation of socialist and

particular groups distinguished,

177-78; left

29-31, 58; and electrade-off, 88, 126; and

voting,

ical trends of,

1,

its

impact on workers’

voting observed, 172-79; of

nonworkers

in

Sweden, 62; of

petty bourgeoisie in various

party strategies, 88, 113-19,

countries, 166, 168-70; of

126; and workers as a minority,

small farmers in France, 5; and

34-40, 126

trade union

Suffrage,

36-37

membership, 174-

77; in various countries,

model

Name

224

Voting, middle classes (contd.)

and surveys, 150-56; ers,

.

work-

vs.

s

countries,

Index

model and surveys,

150-56

166-67, 171-72

Voting, by workers: of agri-

Workers, 31-34, 50; as a minor-

31-40; as a proportion of

cultural vs. industrial workers,

ity,

149; historical trends of, vari-

adults, 27,

ous countries, 158-64; and

tion of the electorate,

industrialization, 163; in

den, 5, 64;

in

various

Swe-

33-35; as a propor-

37-40;

and socialism, according

to

Marx, 32-33, 46-47, 49-50

I

«

I

t

BOSTON PUBL

C

L

BRARY

paperstoneshistoOOprze paperstoneshistoOOprze

INI INI

1 1

paperstoneshistoOOprze

of th/s

Boston Public Libmry

The Date Due Card

in

the pocket

in-

dicates the date on or before which this

book should be returned

to the Library.

Please do not remove cards from this pocket.

The authors analyze from 1890

the history of electoral socialism

to the present in

seven European countries:

Belgium, Denmark, Finland, France, Germany, Norway, and Sweden. They develop a theory that interprets individual voting behavior as an effect of social pro-

They use this becomes or does not

cesses of collective identity formation.

when

theory to understand

become

class

—

— an important determinant of voting behavior.

Their main finding

is

that left-wing political parties

have historically faced an electoral dilemma. Since

workers were never sufficiently numerous to provide an electoral majority, working-class parties

seek support across class

were forced

to

This pursuit of votes

lines.

necessarily resulted, however, in diluting the salience of class as a basis of individual behavior and eventually in

losing potential support

among

workers. Consequently,

even as socialists entered into the competition" for votes, their great expectations could never

be

fulfilled.

Today,

with Western Europe experiencing deindustrialization, the authors argue that the working-class base of left-

wing

parties, as well as their ability to reach across class

lines, is

being eroded.

Przeworski and Sprague employ a mathematical

model

to reconstruct voting patterns of

workers and

other groups since the turn of the century, their perspective, they use this

model

"ib

validate

to predict results

of survey studies conducted since ihc 15 Jos. Highly readable, with dices,

all

technical material relegated to appen-

Paper Stones represents signal contributions

number of fields and

will find a

to a

wide readership among

political scientists, historians, sociologists,

and

all

those

interested in socialism.

“The

authors’

emphasis on

sents a significant

new

political strategies repre-

interpretation. Their ability to

derive important time trends and explanations from a limited data set and rigorous modeling sets an important

new standard

for research in political history.”

— Sidney Verba, Harvard University Adam Przeworski

is

Martin A. Ryerson Distinguished

Service Professor of Political Science, University of

Chicago, and the author of Capitalism and Social

Democracy.

John Sprague

is

professor of political science, Wash-

He

the author of Voting

ington University,

St.

Patterns of the U.S.

Supreme Court, and

Adam tists.

Louis.

is

is

co-author with

Przeworski of Systems Analysis for Social Scien-

BOOKS OF RELATED INTEREST Nationalism and the State John Breuilly

“This

is

a hold account, rigorously

argued and lucidly presented. Breuilly

supports his case with an impressive array of historical detail and a wealth of cumulative comparisons. rational state, he provides a

By singling out the pivotal role of the modern welcome corrective to so many sociological

accounts which would ‘reduce’ nationalism to economic, cultural, or even socio-psychological levels of analysis.”

— Anthony

Geography Quarterly Paper ISBN: 0-226-07412-9

D. Smith, Political

1985 X, 422 pages First

World Nationalisms

Class and Ethnic Politics in Northern Ireland and Quebec

Katherine O’Sullivan See

major study of the origin, development, and dynamics of ethnic separatist movements, using two richly textured case studies of Northern

This

is

a

Ireland and

Quebec. See offers an illuminating

historical analysis that

goes

beyond the previous explanations of ethno-nationalism and explains why Northern Ireland remains caught in sectarian struggle while Quebec has

made

a successful transition to a secular state.

1986

200 pages Cloth ISBN: 0-226-74416-7 Paper ISBN: 0-226-74417-5

A Lost Left Three Studies

in

Socialism and Nationalism '

David Howell

A

Lost Left

is

•

an incisive, multileveled history of the development of British

socialism during the critical period 1914-28, a time in which less attentive to the salience of cultural distinctions

nationalism. Howell’s study illustrates key

leftists

became

and embraced

moments and themes

British experience of socialism and nationalism by focusing

in the

on the careers

of James Connolly, John Maclean, and John Wheatley.

1986

350 pages Cloth ISBN: 0-226-35513-6 Paper ISBN: 0-226-35514-4

THE UNIVERSITY OF CHICAGO PRESS