BOOK REVIEWS 225

The Voter Decides. BY ANGUS CAMPBELL, GERALD GURIN AND WARREN E.

MILLER. (Evanston, 111.: Row, Peterson and Co. 1954. Pp. xiii, 242. $4.75)
This volume is the result of an intensive analysis of voting behavior in the
1952 presidential election, conducted by the Survey Research Center of the
University of Michigan. The study was sponsored by the Committee on Politi-
cal Behavior of the Social Science Research Council, which served in an ad-
visory capacity during the course of the study. Funds were provided by the
Carnegie Foundation. Over 2,000 persons, representing all parts of the country,
were interviewed prior to the election, each for an hour, and 1,614 of these were
reinterviewed after the election, each for a half hour. Although this is a relative-
ly small sample, it may be assumed that unusual care and skill was exercised
by the staff in selecting the persons to be interviewed so that they would be
representative of the entire voting population.
A similar but much smaller survey was conducted following the 1948 elec-
tion, and at many places in this study the voting behavior in the two elections
was fruitfully compared. The differences in the two elections were frequently
striking, giving rise to the observation of Professor V. 0. Key in the foreword
that failure to recognize these variations has frequently caused prognosticators
to come to grief. It is difficult, however, to hazard which of these two elections
was atypical. The authors of this volume evidently regard the 1948 election as
the more typical, but this assumption is certainly open to question.
The objectives of the study, carefully outlined in the opening chapter, were
to identify voters and nonvoters, Republicans and Democrats, with regard to
(a) socio-economic characteristics, and (b) attitudes and opinions on political
issues, and perceptions of the parties and the candidates. Comparisons of these
groups are made with the corresponding groups in the 1948 election, and spe-
cial attention is given to undecided and changing voters. Part I of the study
summarizes the facts brought to light concerning the behaviors and attitudes of
voters in the 1952 election, including the time of decision, the extent of public
interest and participation in the campaign, the public perception of parties and
the candidates, and the demography of the vote. This part follows rather closely
the pathbreaking study by Paul Lazarsfeld and associates in The People's Choice
(New York, 1948), which surveyed the voting behavior in Erie County, Ohio,
through a sample of 600 persons.
Part II is concerned with the motivation of voting behavior, and it is in this
part that the authors push their inquiries into new areas and undertake new
methods of analysis. Here the authors analyze and classify their data in an at-
tempt to explore the major psychological variables which influenced the voter
in his choice. The emphasis is not upon group influences and pressures, which
were stressed in the Lazarsfeld study, but rather upon the motivation of the
individual. The data collected from the sampling interviews are collated with
respect to three basic factors: (1) personal identification with one of the parties;
(2) concern or orientation with respect to national issues, and (3) orientation to
the two presidential candidates. Some attention is given to the influence of as-
sociates and groups, the belief in the efficacy of voting, and the sense of c^vic
obligation to vote. These latter topics are relegated to the appendix.
226 THE AMERICAN POLITICAL SCIENCE REVIEW

What were the significant findings of the study? In what respects do they
correspond with or differ from those of preceding studies? Do they confirm or
discredit the opinions of politicians and political writers who have undertaken
to explain the reasons for the outcome of the 1952 election? In what regards
do they throw new light on voting behavior and open up new hypotheses for
further study and exploration? Do they support or negate the often expressed
opinion that the vote of women was an important factor in the election of
Eisenhower? That his promise to go to Korea influenced the outcome? Or that
his popularity was the decisive factor in the result?
The study throws much new light on the voting behavior in the 1952 elec-
tion, confirming certain commonly held impressions, but also introducing many
new and unsuspected factors. The authors are meticulous in limiting their con-
clusions to those strictly justified by the data; others may find in the numerous
tables summarizing the results of the interviews many other clues to voting be-
havior which they think worthy of further study.
The 1952 election brought to the polls 61.5 million voters, as compared with
only 48.5 four years earlier. Why the higher turnout? The authors reason that
the voters were more highly motivated, which is another way of saying that
public interest ran higher. But the detailed data give important clues to the
higher interest: the greater attractiveness of the candidates, particularly Mr.
Eisenhower; the greater concern over issues, especially foreign. But there were
many other explanations, which the authors, sticking to their data, do not ex-
plore. These include, for example, the fact that the outcome in 1948 was re-
garded generally as a foregone conclusion, always a depressant of voting, where-
as in 1952 the outcome was generally regarded as in doubt. Far greater funds
were spent by both sides in the campaign, and television for the first time be-
came an important factor. But one astonishing fact brought out in the study
is that voting declined from 1948 to 1952 in the metropolitan centers, although
markedly increasing elsewhere.
When does the voter make up his mind? The authors state that whereas in
1948 as late as October some 19 per cent of the public had not decided on the
choice between the two candidates, with the late deciders voting two to one for
Truman, in October 1952 the corresponding figure was only six per cent, with
the division of the late deciders between the two major candidates being the
same as for voters who had decided earlier. While the data are somewhat con-
tradictory, they indicate that immediately after the national conventions two-
thirds of the voters made up their minds once and for all; 60 per cent of these
favored Eisenhower. The results were not altered by the campaign, as they were
in 1948.
The authors state categorically that the Eisenhower victory was not due to
the women's vote, though the statistics given by sex indicate that 41 percent of
the women interviewed voted Republican, while only 28 per cent voted Demo-
cratic. In 1948 the corresponding percentages were 26 and 29. The greatest
shift was in the farm vote and in rural areas, which in 1948 had gone two to one
for the Democrats, and which now switched to two to one Republican, and
 BOOK REVIEWS 227

turned out in much larger numbers. As is well known, Eisenhower drew many
votes from Democrats. Stevenson drew 74 per cent of his votes from regular
Democrats, while the regular Republicans constituted only 56 per cent of the
Eisenhower vote. One-fourth of Eisenhower's vote came from persons who had
voted for Truman in 1948. These constituted the great bulk of the undecided
and the switchers. Evidently many of these voters toyed with the idea of voting
for Dewey in 1948, but returned to the fold on election day; in 1952 they were
attracted to the Republican side, largely because of the candidate, and stayed
with it on election day.
Students of politics will be interested in the statistics concerning active par-
ticipation in the campaign. Four per cent of the respondents contributed finan-
cially, though 12 per cent were solicited; three per cent did party work, and
seven per cent attended political meetings. More than a fourth of the voters
urged others to support their candidate, including some who failed to vote
themselves! The data indicate in considerable detail the numbers of voters who
were greatly interested and concerned about the outcome of the election, those
only mildly concerned, and those who were apathetic. It is of interest to note
that the percentage of those who voted was much lower in the high interest
group of the Democrats than in the corresponding group of Republicans.
The data on party identification will also be of great interest to the student
of political behavior. One of the major findings of the study, which belies com-
mon impressions, is that party identification for the great majority of voters is
very real and continuing. Three-fourths of the voters regarded themselves as
members of one of the major parties. Although one-fourth of the respondents
stated that they were independents, on further questioning only seven per cent
indicated that they did not favor one party or the other. Even those who
switched and voted for the opposing candidate still regarded themselves as
identified with their party. An important corollary to this finding is that a large
number of voters who switched or who voted Republican for the first time do
not regard themselves as firmly attached to the party, and whether they will
remain with it will depend upon future events. There were very few switchers
to the Democrats, but the new Democratic voters, in contrast, regarded them-
selves as strongly attached to the party. What this forbodes for the future, only
time will tell.
The analysis of issue orientation is, at least for this reviewer, somewhat in-
conclusive. Whether the Korean war was the decisive factor, causing many
loyal Democrats in the low income groups to leave the party, cannot be ascer-
tained from the study. Generally, only about a third of the voters were informed
on the stand of the parties on issues; another third thought there was no differ-
ence in their stands; and the other third did not know. The issue orientation of
the Democrats in the South who switched to Eisenhower was practically the
same as that of the regular Republicans of that area.
This is a notable study of political behavior. The attempt to identify and
evaluate several significant factors of voter motivation—partisanship, issue
orientation, and the appeals of the candidates—utilizes a new technique of
228 THE AMEHICAN POLITICAL SCIENCE REVIEW

political analysis. However, as the authors point out, these factors are closely
related to one another and are not neatly separated in the mind of the voter.
For this reason, it is questionable whether this technique can ever provide fully
satisfactory explanations of voter motivation. The authors are to be com-
mended for the great caution which they exercised in drawing conclusions from
their data. Teachers and research students of the subject owe a heavy debt to
them for a volume which will serve as a sourcebook and an inspiration for con-
tinuing studies.
JOSEPH P. HARRIS.
University of California (Berkeley).
Government and Science. BY DON K. PRICE. (New York: New York University
Press. 1954. Pp. 203. $3.75.)
The endless frontier that is science has advanced so far, so rapidly, so dy-
namically that every human institution has felt its revolutionary impact.
Government is no exception; and Don Price would be the first to acknowledge
that his book, pioneer effort that it is, has fallen short of fully comprehending
this impact. Yet in his Stokes Lectures he has carried this analysis considerably
farther than any contemporary writer. Here is a creative mind operating on
data both old and new, bringing new insights and a new dimension to the study
of government.
The work opens with the thesis that scientific research and professional dis-
cussion, rather than political conflict and pressure groups, have historically
been the principal determinative factors in national policy and administration.
A considerable number of examples of such influence have been woven together,
not enough to prove the point but sufficient to intrigue the reader. The union of
popular pragmatism and the prestige of our scientists has created an atmos-
phere such that the present day is witnessing the full flowering of scientific in-
fluence on governmental policy.
It is in such a setting that the dilemma has risen between the traditional free-
dom of the scientist and the imperious demands of a government whose very
existence may someday be threatened. The problems thereby created are an-
alyzed. An important contribution toward responsible freedom lies in contracts
with universities and other institutions in lieu of the alternative of direct gov-
ernment operation. On the other hand, the current undiscriminating passion
for "security" is a threat greater than any which might lie in a strong and re-
sponsible executive.
How far and through what means science and the scientists should advise
government, to what extent they should determine its actual policy—these
are the two major questions to which the remainder of the book is devoted.
The questions raised are important and clear. The author's answers are more
implied than explicit.
Science in government, including government subsidy of basic research, can
and should be administered within the governmental hierarchy, at least to the
extent that there should exist policy-determining officials in line of authority