Professional Documents
Culture Documents
congressman, and state and local offices. These data provide the basis
for the analysis presented in this article.2
I. THE DISTRIBUTION OF STRAIGHT AND SPLIT TICKETS
A voter who does not wish to vote a straight ticket, in a presidential
election may split his vote in a variety of ways, ways which have quite
different political implications. We, shall divide the ballot-splitters into
the following five categories, based on certain assumptions about the
motivations of voters using these different patterns of ballot-splitting:
1. Those who vote a straight ticket except for President.
2. Those who vote a straight ticket except for congressman or senator.
3. Those who vote a straight ticket at the national level (president,
senator and congressman) but a straight ticket for the opposite
party at the state and local level.
4. Those who vote a straight ticket at the national level but split
their ticket at the state and local level.
5. Those who split their ticket both at the national and local levels.
When we divide the voters according to these different ballot pat-
terns, considering Eisenhower and Stevenson voters in the North and
South separately, we find the distribution presented in Table 1. Data
from our study of the 1952 presidential election are also included for
purposes of comparison.3 Without attempting a full discussion of the
contents of this table we may comment on the following conclusions
which can be drawn from it:
1. A very sizeable segment of the American electorate crosses party
lines when it votes in presidential elections, something on the order
of one-third of the voters in 1952 and two-fifths in 1956. The most
common splitting pattern is a straight vote for national candidates
and a split among the state and local candidates. However, a
significant number of voters split at both levels. Splitting an other-
wise straight ticket in favor of a national candidate of the opposite
party is less common.
2. Voters in the North and South differ greatly in the way they have
marked their ballots in the last two presidential elections. Eisen-
8
At the time this analysis was undertaken we had in hand an unpublished manuscript
entitled "The Split-Ticket Voter in 1952," written by Professor Daniel M. Ogden, Jr., of
the Department of Political Science of Washington State College, and based on data col-
lected by the Survey Research Center in its 1952 study. We wish to acknowledge Profes-
sor Ogden's generosity in making this manuscript available to us.
* See Campbell, A., Gurin, G., and Miller, W. E., The Voter Decides (Evanston: Row,
Peterson & Co., 1954).
STRAIGHT AND SPLIT TICKET VOTING 295
North South
19S2 1956 1952 1956 1952 1956 1952 1956 1952 1956
Voted straight ticket 66% 56% 70% 68% 17% 34% 95% 96% 66% 61%
Voted straight except for Pres- 1 6 •J * 39 31 —i — 4 6
ident
Voted straight except for sen-
ator or congressman 2 3 4 3 4 2 3 2 3 3
Voted straight at national
level, straight for opposite
party at state and local level * * * * — 6 1 — * 1
Voted straight at national lev-
el, split at stateandlocallevel 23 20 18 21 4 6 1 2 18 17
Voted split ticket at both levels 8 15 8 8 36 21 — — 9 12
100% 100% 100% 100% 100% 100% 100% 100% 100% 100%
Number of cases 522 575 335 356 72 118 77 121 1,006 1 ,189
1
For our purposes the South consists of all the states east of New Mexico and south of the Ohio River, including
West Virginia and Maryland. The North consists of all the remaining states.
1
This table includes only those voters in our samples who reported a vote for president, congressman or senator,
and state and local candidates. Excluded are 15 per cent of the 1952 voters and 8 per cent of the 1956 voters who
either did not vote for one or more offices on the ballot or did not give us full information about their vote.
s
The asterisk is used in these tables to denote frequencies of less than one per cent.
' The dash is used in these tables to denote a zero frequency.
TABLE II. PISTBIBUTION OP STRAIGHT AND SPLIT TICKET VOTEBS WITHIN DEMO-
GRAPHIC GROUPS: EISENHOWER VOTERS IN THE NORTH
Straight
Straight natl., Straight Voted
Not Number
Straight , except straight nation- , ,, incom-
Total of
ticket J o r sen for other
for sen. al, split a * b ° t h
al plete
leTeU cases
*T: oroong. party locally ticket
dent . .,
locally
Straight
Straight
Straight nat'l., Straight Split Voted
except Not Number
Straight except straight nation- at incom- rp.i 1
ticket lor for sen. for other al, split both plete ascer- lotai OI
irreai- or cong. party locally levels ticket tained cftses
dent locally
ballot. Of the Eisenhower voters in the single choice states, 59 per cent
voted a straight ticket; in the other states it was 48 per cent. Of the
Stevenson voters in the single choice states, 69 per cent voted a straight
ticket; 60 per cent voted a straight ticket in the multiple choice states.
Since the distributions of party identifiers in the two types of states do
not differ we must conclude that the sheer ease of voting a straight ticket
facilitates this type of voting.6 This fact becomes an important consider-
ation in our subsequent discussion.
We turn now to the political motives of the voters and the ways they
relate to the use of the straight or split ticket. We present below three
explanations of the straight or split ticket vote, drawn partly from
political folklore and partly from earlier research findings, and subject
each to such tests as the data of our study make possible.
1. The straight ticket is the product of indifference and lack of con-
cern with political affairs. Since marking the party circle, pulling
the party lever, or otherwise voting a party ticket requires the
least effort on the part of the voter, it is most attractive to the
voter who is least willing to make a discriminating effort in the
selection of candidates. The split ticket, conversely, is the mark of
the motivated voter, the voter with sufficient interest to distin-
guish among the candidates and to take the trouble to choose
among them.
2. The straight ticket is the expression of high political motivation.
The motivated voter is more likely to organize his ballot in a highly
structured pattern. His motivation may derive from an interest in
parties, candidates, or issues or any combination of these. Since
the party label on the ballot makes it possible for him to shape his
vote in support of his motivation, he is likely to be a straight-
ticket voter. In contrast, the indifferent, unmotivated voter, feel-
ing no need to organize his ballot toward any general political
objective, makes his selections among candidates on a capricious,
quasi-random basis. In doing so he is very likely to split his ballot.
3. The straight ticket is the mark of the party-oriented voter.
Whether his general interest in politics is high or low, his party at-
tachment governs his voting behavior; he goes down the line for his
6
We may also note that the single choice type of ticket appeared to hold Democratic
identifiers more securely to their presidential nominee than did the multiple choice ballot.
In the single choice states, 91 per cent of the "strong" Democrats voted for Stevenson,
nine per cent for Eisenhower; in the multiple choice Btates, the percentages were 86 and
14. "Weak" Democrats gave Stevenson 70 per cent of their votes in the single choice
states, 30 per cent to Eisenhower, and in the multiple choice states, 64 and 36 per cent.
The votes of Republican identifiers in the two types of states did not differ.
STRAIGHT AND SPLIT TICKET VOTING 301
party. Ballot splitting is found either where party identification
is weak or absent or where there is conflict between party identi-
fication and other political motives.
The Straight Ticket as an Indifferent Vote. The supposition that the
straight ballot bespeaks a casual, disinterested vote has a considerable
currency in this country. The popularity of the concept of "voting for
the man," which we have observed in our earlier studies, carries with it
the implication that a straight ticket voter is likely to be an uninformed,
indifferent person who takes the easy way out in the voting booth.
The data we have already seen in Tables II and III suggest that this
explanation of straight ticket voting must be quite inadequate. We do
not find there, for example, any significant tendency for the higher edu-
cational and occupational groups to split their tickets as this theory
would lead us to expect. Such differences as we find among educational
and occupational groups are small and inconsistent.
Looking more directly at the factor of indifference, we may divide
our respondents into those who said they "cared a great deal" whether
TABLE IV. RELATION OF BTBAIOHT AND SPLIT TICKET VOTING TO THE
IMPORTANCE OF VOTING: NOBTH
"We would like to know how strongly1 you felt about the importance of voting in this
election. Would you say you oared a great deal whether or not you voted, cared some-
what, or didn't care too much thia time?' i
Cared Cared
Cared somewhat Cared somewhat
great deal or not great deal or not
too much too much
or not they voted in the 1956 presidential election and those who were
less concerned about voting. This distribution is heavily loaded at the
"cared a great deal" end, but we find that those voters who seemed to
attach relatively little importance to whether or not they voted were not
therefore more likely to be straight ticket voters (Table IV). They were,
in fact, more likely to split their ballot than were the voters who felt
more strongly about the importance of voting.7
The Straight Ticket as a Motivated Vote. Paradoxically, it appears from
Table IV that, while some people may vote a straight ticket because
they are not highly motivated politically, others vote a straight ticket
for the very reason that they are so motivated. In order to test this latter
proposition more adequately than our single question on the felt im-
portance of voting makes possible, we need to draw from our data a
measure of the major political motivations which influence the vote. In
our study of the 1952 elections we postulated three types of political
motives which we were able to show accounted for a substantial part
of the total variance in voting for President. These three factors related
respectively to parties, candidates, and issues; they were defined in the
following terms:
Party identification. We refer by this term to the sense of personal
attachment which the individual citizen feels toward the political
party of his choice. We equate strong identification with high
significance of the party as an influential standard. In the course of
the pre-election interview we asked each respondent, "Do you
usually think of yourself as a Republican, a Democrat, an Inde-
pendent, or what?" If he identified himself as a Republican or a
Democrat, he was asked, "Would you call yourself a strong (Repub-
lican or Democrat) or a not very strong (Republican or Democrat)?"
These questions permit a division of the sample into those strongly
identified with one party or the other and those not so identified.
Candidate partisanship. This variable is defined as the structuring of
political events in terms of a personal attraction to the major per-
sonalities involved, in this case the presidential candidates of the
two parties. The extent to which our respondents reacted to the
personal qualifications of the two presidential candidates and the
relative attractiveness of one over the other was assessed from a
series of open-ended questions asked in the pre-election interview.
' Thisfindingis not changed when the voters of single choice and multiple choice states
are compared. There is more straight ticket voting in the single choice states but those
voters who "cared a great deal" whether or not they voted were more likely to vote a
straight ticket than those who did not in both groups of states.
STRAIGHT AND SPLIT TICKET VOTING 303
may be, the simplicity of the single choice ballot facilitates straight
ticket voting. Significantly, this is particularly marked among those
voters who by our measurements have no active political motivation.
These relatively unmotivated people are notably more likely to vote a
straight ticket if they can do so by making a single mark on the ballot
than if the voting procedure is more complicated.
One additional comment regarding Table V is in order. We see that
in the single choice states, 45 per cent of those voters who reported no
supporting motivation for their vote voted a straight ticket; in the
multiple choice states the comparable figure was 27 per cent. We are not
surprised to find a small number of apparently unmotivated people in
our sample of voters; our three-factor theory of political motivation
does not include all of the influences that might bring a person to the
polls. It is among these people, however, that we would expect the factor
of least effort to be most effective, and the fact that less than half of
them actually vote a straight ticket requires some explanation.
If we assume that a straight ticket vote by an indifferent, politically
unmotivated voter is an expression of his wish to remove himself from
the voting situation with the least possible effort, we must conclude that
Three • 64 77
Two * 39 66
One * 54
None *
Summary 45 49 65 77
Number of cases 38 108 179 122
Note: Asterisk denotes number of cases too small for computation of percentages. Dash denotes zero frequency.
STRAIGHT AND SPLIT TICKET VOTING 307
Party Identification:
Supports vote for President 741 73 76 70 89 S3 69 * 47
Neutral (Independent) 40 36 38 43 33 27 • *
Conflicts with vote for President 18 * * 20 * * * * *
> Entry is the percentage of the group who voted a straight ticket.
1
Number of cases too small for computation of percentages.
Although the data are not altogether consistent, the candidate and
issue factors both appear to have an independent influence on the ten-
dency to vote a straight ticket. Neither of them is as strong as the party
factor, but they are both visible.
The reader will also observe from both Table V and Table VII that
when the political motives we have been discussing come into conflict,
split ticket voting tends to increase. Table V, which summarizes the
interaction of the three motivational variables, demonstrates clearly
that the presence of a conflicting motive is associated with a lower pro-
portion of straight ticket voting than is found when an equal number of
supporting motives are present without conflict. Table VII shows the
same data with the individual motives specifically identified. These
tables confirm and extend similar data we reported from our 1952 study.
STRAIGHT AND SPLIT TICKET VOTING 309
ballots we can at least identify the most obvious points at which coattail
influence was or was not present in 1956. Looking back to Table VI we
see that a considerable number of people who normally think of them-
selves as Democrats voted for Mr. Eisenhower. It is evident that a large
part of this defection resulted from a personal attraction to the candidate
rather than from party or issue considerations. Most of these Democrats
supported some or most of the Democratic candidates on the ballot, but
almost a third of them voted for a Republican congressman. They either
voted a straight Republican ticket or a straight Republican ticket at the
national level only. In either case it seems that their vote for President
influenced their vote for congressman. Since these people made up six
per cent of the total vote for Republican congressmen in the North it
must be concluded that a visible coattail effect was present. The number
of Republicans who voted for Stevenson was very small in 1956 but if
we give him all the credit for such defections as did occur among Re-
publican identifiers it is clear that his ability to carry Republican votes
for Democratic congressmen could not have been more than two per
cent of their total vote in the Northern states.
There are undoubtedly other coattail effects which we cannot easily
assess. The ability of a presidential candidate to hold the members of his
own party in line is an important aspect of his influence on the vote. His
success in winning the votes of the Independent voters to his ticket may
be more significant than the defectors he may attract from the other
party. Without attempting to evaluate these effects exactly, we would
conclude from the data we have seen that even though a Presidential
candidate runs far ahead of his fellow partisans for congressional office,
they may owe him a greater debt than the election returns imply on the
surface.
III. DISCUSSION
tisanship. All these variables are clearly related to the two-party nature
of American.politics and we have assumed, in accordance with the gen-
eral principle of motivational strength, that when they combine in a
mutually supportive manner they increase the pressure toward full
support of one of the parties. Our data have shown that straight ticket
voting is lowest among voters for whom none of these motives is present
and highest among those whose party, candidate, and issue motives all
impel them to support the same party. Deviations across party lines
are least common among those voters who identify with their party,
prefer its presidential candidate, and approve its issue positions.
However, this statement does not fully account for the variations we
find in straight and split ballot voting. The fact that the American ballot
system permits the voter to choose an entire party slate with a single
decision, and in a large proportion of the cases with a single movement of
a pencil or voting machine lever, introduces an additional factor into the
explanation of the vote. If the voter were compelled to go through the
individual positions on the ballot, making independent decisions on
each one separately, we would expect the unmotivated voter to be very
unlikely to vote a straight ticket. Not being guided by any general
political motive according to which he could organize the entire ticket
he would presumably be influenced by a variety of unrelated consider-
ations that would take him haphazardly from one side of the ballot to
the other. The fact is, however, that we find a substantial number of
straight tickets among voters who seem to have no partisan commitment
to either parties, candidates, or issues. To explain their presence in our
data we must invoke a second motivational principle, the tendency
toward least effort. We conclude that in the absence of relevant political
motivation, these people are voting a straight ticket because the pecu-
liarities of the voting system make a straight ticket vote the easiest way
for them to complete the task of voting. The more the ballot form facili-
tates the marking of a straight ticket, the more likely they are to choose
this alternative.
We must also take account of the fact of conflict between political
motives. We find that the factor we have called party identification
exercises a strong controlling influence on the way a voter marks his
ballot, but when his candidate preference and issue partisanship conflict
with his sense of party attachment he tends to try to satisfy both com-
ponents of the conflict by supporting candidates from both parties. His
conflicted motivation reduces the pressure toward a party-oriented vote
and opens the way to a divided ballot.
We may summarize this discussion by proposing a four-fold classi-
312 THE AMERICAN POLITICAL SCIENCE REVIEW