Content uploaded by Julia Dupont
Author content
All content in this area was uploaded by Julia Dupont on Nov 28, 2018
Content may be subject to copyright.
PERCEIVED ELECTION PLEDGES – (original publication of the article available at
Science Direct: Electoral Studies, 2018)
1
Which Kind of Political Campaign Messages do People Perceive as Election
Pledges?
Julia C. Dupont (corresponding author) a)
a) University of Koblenz-Landau, Landau, Germany
Faculty of Psychology, Department of Social, Environmental and Economic Psychology
Fortstr. 7
76829 Landau, Germany
Mail: dupont@uni-landau.de
Evelyn Bytzek b)
b) University of Koblenz-Landau, Landau, Germany
Faculty of Cultural and Social Sciences, Department of Political Science
Kaufhausgasse 9
76829 Landau, Germany
Mail: bytzek@uni-landau.de
Melanie C. Steffens c)
c) University of Koblenz-Landau, Landau, Germany
Faculty of Psychology, Department of Social, Environmental and Economic Psychology
Fortstr. 7
76829 Landau, Germany
Mail: steffens@uni-landau.de
Frank M. Schneider d)
d) University of Mannheim, Mannheim, Germany
Institute for Media and Communication Studies
B 6, 30-32
68159 Mannheim, Germany
Mail: frank.schneider@uni-mannheim.de
This research was supported by a research grant from the research focus "Communication,
Media and Politics" which is part of the Forschungsinitiative Rheinland-Pfalz.
We thank Anders Larrabee Sonderlund for language editing.
PERCEIVED ELECTION PLEDGES
2
Which Kind of Political Campaign Messages do People Perceive as Election
Pledges?
Abstract
Contrary to the popular notion of the pledge-breaking politician, research has revealed that
governing political parties fulfill most of their pre-election pledges. This discrepancy between
public perception and scientific findings could be the result of citizens’ looser definition of
election pledges. Thus, as citizens’ assessment of pledge fulfillment (and with that their
satisfaction with the governing parties) hinges on their understanding of election pledges, we
examined citizens’ perception of election pledges. Our experiment (N = 705) showed that
citizens were more likely to perceive messages as pledges when they were articulated in binding
versus less-binding terms, but when they were vague as opposed to specific in terms of goals.
Political attitudes also biased the understanding of election pledges, demonstrating that it is in
the eye of the beholder which messages are considered to be election pledges.
Keywords: election pledges, election promises, campaign commitments, public opinion, political
communication, electoral accountability, issue competence
1. Election Pledges in Theory and Practice
Past research on pledge fulfillment indicates that the majority of governing parties’ election
pledges
1
is congruent with subsequent policy actions (for a recent meta-analysis, see Thomson et
1
Although the term “election pledge” is widely used (e.g., https://www.merriam-
PERCEIVED ELECTION PLEDGES
3
al., 2017). However, citizens often assume that politicians rarely keep what they may have
promised during campaigning. This phenomenon is sometimes called the ‘stereotype of the
pledge-breaking politician’ (e.g., Naurin, 2011; Thomson, 2011; Thomson & Brandenburg,
2018).
Examining the public’s understanding of election pledges is important as citizens’ accurate
perception of the content of such promises, as well as the extent to which they are delivered on,
is fundamental in terms of evaluation of incumbents’ performance as politicians (e.g., Naurin,
Soroka, & Markwat, 2015). According to the mandate theory of democracy, rational citizens will
vote for a party that endorses policies that are compatible with their own attitudes and opinions
(e.g., election pledges; e.g., Grossback, Peterson, & Stimson, 2005; McDonald & Budge, 2005).
By electing a government, voters authorize the party (or parties) to implement their planned
policy projects, hence linking voter preferences to governmental action (e.g., Klingemann,
Hoffebert, & Budge, 1994). Based on party’s performance while in power, citizens have the
opportunity to express their approval or disapproval at the next election – for example by
sustaining or withdrawing their support. In line with models on retrospective voting and electoral
accountability (Downs, 1957; Fiorina, 1981; Key & Cummings, 1966), voters must be able to
link campaign commitments to subsequent political action to maintain their support.
However, most of the empirical research on electoral accountability links voters’ assessment of
electoral performance to inflation, employment, and other growth indicators (e.g., Lewis-Beck &
Stegmaier, 2000). “There is only little research on the relationship between citizens’ perception
of election campaign pledge fulfillment and the extent to which they hold a government
webster.com/dictionary/pledge), “election promise” is also common. For this paper, we decided to use the
former because it is also used in political science research. However, as our study was conducted in
Germany, where only one term exists (“Wahlversprechen”), the decision to use “election pledge” has no
consequences for our experiment.
PERCEIVED ELECTION PLEDGES
4
accountable for broken or fulfilled promises (but see Born, van Eck, & Johannesson, 2018;
Naurin et al., 2015). This is surprising given the facts that (a) the number of election pledges in
party manifestos is increasing (e.g., for Sweden: Hakansson & Naurin, 2014); (b) election
pledges represent a popular media topic (Hakansson & Naurin, 2014; Kostadinova, 2017); and
(c) pledges actually do affect government actions (e.g., Thomson et al., 2017). In other words,
campaign pledges may help voters decide where to cast their vote prospectively (voters base
decision on the congruence with parties’ policy aims) as well as retrospectively (voters base
decision on their evaluation of parties’ past performance, i.e. have election promises been kept?)
(Naurin, 2011; Naurin & Hakansson, 2014). However, there is no empirical consensus on the
question of which specific criteria voters actually use to assess government performance
(Anderson, 2007). Regarding pledge fulfillment as a performance measure, this could be due to
scientists’ limited knowledge about citizens’ perception of election pledges in general.
Given the fundamental role of government performance – which at least in part depends on the
fulfillment of election pledges – in the context of citizens’ voting behavior, in this paper we ask
the question of which kind of campaign messages citizens perceive as election pledges. In a
nutshell, we propose that what is perceived as an election pledge depends on (a) characteristics
of the message (i.e., specificity, binding force) and (b) subjective factors such as party
identification or issue involvement.
2. Citizens’ Evaluation of Election Pledges
Past research indicates that both subjective and objective factors influence citizens’ evaluation of
election pledges (Naurin, 2011; Thomson, 2011; Thomson & Brandenburg, 2018). The
subjective factors include motivational and cognitive processes. For instance, in line with
PERCEIVED ELECTION PLEDGES
5
psychological dual-process theories (e.g. the heuristic systematic model (Chaiken, 1980); the
elaboration likelihood model (Petty & Cacioppo, 1986)), it is especially people who lack the
desire or ability to analyze political information who use information shortcuts to arrive at
evaluations about that information. Such heuristic processing may lead to cognitive biases and
systematic judgment “errors” (e.g., Kahneman, Slovic & Tversky, 1982; Kruglanski & Ajzen,
1983; Kunda, 1990; Ross, 1977; Simon, 1990, 1995). For example, politically uninformed
citizens seem to evaluate election pledge fulfillment more negatively than do their more
politically informed counterparts (Naurin & Oscarsson, 2017; Thomson, 2011; Thomson &
Brandenburg, 2018). To evaluate politicians in a resource-efficient manner, politically
uninformed citizens may rely on stereotypes (Kuklinski & Quirk, 2000).
Party identification may cause one to be biased in favor of the given party (Campbell,
Converse, Miller, & Stokes, 1960). This bias typically results in the individual perceiving his or
her party more favorably than other parties, including when it comes to kept or broken election
pledges (Thomson, 2011).
Other factors such as issue involvement (i.e. the personal relevance of a topic) also
appear to determine how people perceive government performance (Duch, Palmer & Anderson,
2000; Funk & Garcia-Monet, 1997; Thomson, 2011). For example, studies have found that
individuals who have had a personal experience related to a specific election pledge (e.g.,
unemployment) assessed election pledges more negatively than individuals who did not have
such an experience (Thomson, 2011). Likewise, citizens lend more weight to pledge statements
that are personally relevant than to statements that are not (Naurin 2011). Finally, citizens tend to
give negative evaluations on pledge fulfillment more readily than positive evaluations (Naurin et
al., 2015; Thomson & Brandenburg, 2018).
PERCEIVED ELECTION PLEDGES
6
In addition to the subjective perception of the veracity of election pledges, citizens’ evaluations
also depend on actual government performance. For example, citizens are unsurprisingly more
likely to evaluate electoral pledges as kept when these promises have actually been implemented
post election (Naurin & Oscarsson, 2017; Thomson, 2011; Thomson & Brandenburg, 2018).
Therefore, as governments usually fulfill most of their pledges (e.g., Thomson et al., 2017),
citizens’ critical and pessimistic perceptions of pledge fulfillment should mainly depend on
subjective factors (e.g., Naurin, 2011, Naurin et al., 2015).
Taken together, the findings imply that citizens’ perception and evaluation of electoral pledges
are not purely objective, but heavily influenced by subjective factors. Besides, we think that
citizens’ evaluation of election pledges fundamentally depends on the kind of campaign
commitments they refer to when they discuss election pledges. In that regard, their perception of
election pledges might be affected by mass media evaluations of election pledges. As media
coverage often is biased (e.g., Entman, 2007), citizens may perceive messages as more likely to
be election pledges if the media reports about a topic frequently and with a wording hinting
towards a pledge being made. Consequently, when analyzing which statements citizens refer to
as election pledges, it is important to also take characteristics of the message (i.e. topic, source,
degree of specificity, level of binding force) into account.
3. Criteria for Identifying Election Pledges
Our aim is to assess if and how the wording of a given statement influences citizens’ perception
of that statement as an election pledge. Thus, we (1) refer to the criteria used in comparative
pledge research to identify pledges in party programs in order to examine the extent to which
PERCEIVED ELECTION PLEDGES
7
such pledges are fulfilled (“the pledge approach”, Thomson et al., 2017, p. 528), and (2) to
supplement these criteria with potentially influential subjective factors as described above.
3.1 Characteristics of the message
Election pledges differ in terms of four key characteristics: (1) the political issue they refer to
(i.e., the topic), (2) the source (i.e., the political actor making the pledge), (3) the degree of
specificity, and (4) the level of binding force. In the following, we will focus on the last two
characteristics because we assume that, in general, citizens may perceive statements from any
political actor as election pledges. However, as the topic and source of a statement as well as
media reporting on these factors can vary, the topic and source of a statement are taken into
account as control variables.
The specificity of statements concerning future action is a central criterion for the identification
of election pledges in the pledge approach (Bara, 2005; Ferguson, 2010, 2012; Naurin, 2011,
2014; Royed, 1996). The notion of specificity is based on the idea that to assess whether a pledge
has been kept or broken, it must be sufficiently specific – for example in terms of a particular
goal or action
2
(Naurin, 2011). Relatively vague statements, such as “We will reduce the
unemployment rate”, may provide room for politicians to choose between different courses of
action, while retaining a concrete goal (Hakansson & Naurin, 2014, Naurin, 2014, see
“ambiguous pledges”; Royed, 1996, p. 67). Thus, specific statements differ from vague
statements in terms of the extent of information on policy implementation as well as goals
2
Although researchers differ in the way in which they measure the specificity of a statement, they agree
that extremely vague statements have to be excluded from the analysis. Hence statements so vague as to
leave unlimited scope for interpretation are not defined as election pledges (i.e., rhetorical statements,
commonplaces, phrases; Bara, 2005, p. 587; Maurer, 2007, p. 179; Naurin, 2011, p. 54, 2014).
PERCEIVED ELECTION PLEDGES
8
(Royed, 1996). For example: “By using new programs to strengthen the labor market, we will
reduce the unemployment rate to below 5 percent.”
Naurin (2011, p. 127) asserts that when evaluating election pledges, citizens take both specific
and vague pledges into account. However, it remains unclear whether all types of statements are
equally perceived as election pledges. It is unlikely that citizens’ perception of a political
statement as a pledge or not depends on whether a campaign commitment is specific enough to
be clearly evaluated. As a shortcut to avoid gathering information on how serious a party is
taking its pledges, citizens may, however, use statement specificity to help them gauge a party’s
true intentions to implement its policies (e.g., McKelvey & Ordeshook, 1986; Popkin, 1991).
Thus, citizens might perceive specific statements as real election pledges but vague statements as
cheap talk, empty platitudes, or slogans. As the number of specific pledges in election manifestos
is increasing (e.g., for Sweden: Hakansson & Naurin, 2014), these kinds of pledges should also
be more visible to citizens. As such, we hypothesize that specific statements are more likely to be
perceived as election pledges than vague statements (H1).
The importance of the binding nature of election campaign statements is somewhat
controversial in election pledge research (Naurin, 2011, p. 34f.). The binding nature of a
statement refers to the exact wording of the given statement. Studies on election pledges have
mostly focused on so-called hard pledges (Kalogeropoulou, 1989; Rallings, 1987; Rose, 1984).
The distinction between “hard pledges” and “soft pledges” was originally introduced by Royed
(1996) and concerns the degree of binding force of a political plan. According to this line of
research, hard statements contain unequivocal phrases such as “we promise to”, “we pledge”, or
“we will”. If a statement contains weak verbs (e.g., want, support, should, favor), they are
categorized as soft statements and sometimes excluded from analysis (Ferguson, 2010, 2012;
PERCEIVED ELECTION PLEDGES
9
Royed, 1996). Other pledge researchers, who consider the meaning of a statement to be more
important than its wording, reject this approach (Hakansson & Naurin, 2014; Naurin, 2011, p.
55, 2014). However, there is empirical evidence that the wording of election pledges affects how
voters react to them (Lindgren, 2018; Lindgren & Naurin, 2017; Thomson & Brandenburg,
2018). As the wording can prime specific cognitive schemata (Lindgren, 2018) and because
pledges have a normatively binding character (meaning that citizens expect election pledges to
be kept; e.g., Searle & Vanderveken, 1985, p. 193), we assume:
Statements of high binding force (hard statements) are more likely to be perceived as election
pledges than statements of low binding force (soft statements) (H2).
3.2 Subjective Criteria
In addition to these “objective” characteristics of a statement, previous research on election
pledges suggests that also subjective motivational factors play a role in how citizens assess
campaign messages. One important motivational factor that may affect whether citizens perceive
messages as election pledges or not relates to issue involvement (i.e. the extent of a given topic’s
personal relevance to voters; e.g., M. Sherif & Hovland, 1961; C. W. Sherif et al., 1973).
Transferred to the context of election pledges, we define issue involvement as the extent to
which citizens perceive a political statement or policy as relevant to them. Studies show that
messages are processed at a higher level when they are personally relevant due to greater
motivation and competence to engage with the topic (Petty & Cacioppo, 1986). For instance,
Thomson (2011) found that citizens with personal experience of an issue about which an election
pledge was made, were more likely to evaluate such a pledge as not being fulfilled than
PERCEIVED ELECTION PLEDGES
10
respondents without such personal experience.
3
Thus, we assume that citizens are more likely to
perceive statements as election pledges if they concern policy areas that are personally relevant
to them (H3).
According to Petty and Cacioppo (1986), the effect of involvement on information processing is
moderated by the quality of the arguments in a statement. If one interprets specificity and
binding force of a statement as indicators of an argument’s quality, it follows naturally to
hypothesize that if issue involvement is high (as opposed to low) and a statement is specific (as
opposed to vague), political statements are more likely to be perceived as election pledges than
not (H3a).
Similarly, if issue involvement is high (as opposed to low) and a statement is binding (as
opposed to non-binding), political statements are more likely to be perceived as election pledges
than not (H3b).
The second important motivational factor that may affect the extent to which citizens perceive
messages as election pledges or not, concerns party identification. Party identification – that is, a
stable psychological bond to a political party – is not only seen as a key influencing factor in
electoral decision making, but also as a strong predictor of political attitudes and behavior in
general (e.g., Johnston, 2006). Campbell et al. (1960) assume that followers see their own party
as part of their social identity and therefore always look at new information through the lens of
an ingroup member (self-stereotyping, see Mayer, 2017, p. 82). Similarly, research on motivated
3
We are aware that the perception of election pledges and the perception of election pledge fulfillment
are, strictly speaking, not the same. Nevertheless, as people’s perception is always subjective, we assume
that any perception of election pledges is biased by subjective criteria. Unclear is, however, which of
these criteria matter in regard to the examined dependent perception variable. We used the available
findings in the election pledge literature as indications on the relevance of specific subjective criteria for
people’s general perception of election pledges.
PERCEIVED ELECTION PLEDGES
11
reasoning describes how existing beliefs and attitudes affect the processing of new information
(Strickland, Taber & Lodge, 2011). Thus, for example, citizens who have a positive attitude
towards a party more readily accept information confirming this perception, and are more likely
to reject information that conflicts with their preconceptions about their party. In this way,
existing beliefs and attitudes can be maintained (Lodge & Taber, 2000; Taber & Lodge, 2006).
Party labels can thus be used as shortcuts for a positive or negative evaluation of a party (Rahn,
1993). With regard to election pledges, party supporters should perceive the statements from
their own party more positively compared to non-party supporters, that means party supporters
should think of their own party that it is serious about its policy goals and thus is able to make
binding statements (i.e. election pledges). We thus assume that people who identify with a given
party are more likely to perceive a statement as an election pledge than citizens who do not
identify with that party. (H4).
In addition to the above hypotheses, we also explore the role of political interest in evaluating
election pledges. Although we do not hypothesize a main effect of political interest on the
perception of a political statement as an election pledge, it can be assumed that politically
involved people not only set different standards in evaluating election pledges than politically
uninterested citizens (Naurin, 2011; Thomson, 2011), but also that their perception of political
content generally differs. However, diverging views exist in the literature: On the one hand,
research on prejudice and stereotypes implies that people with little involvement (e.g., little
interest, little knowledge, low affinity) rely more strongly on cognitive schemata such as
heuristics and stereotypes to better structure and manage decisions (Klauer, 2008). This means,
for example, that respondents with little political interest are more dependent on cues, such as
specificity and binding force, when it comes to determining election pledges. On the other hand,
PERCEIVED ELECTION PLEDGES
12
findings from motivated reasoning research suggest that politically informed and interested
persons’ political views often are strongly distorted by political preferences (Strickland et al.,
2011). Thus, it can be expected that the effect of party identification as described in H4 is
moderated by political interest (H5) – however, it is unclear in which direction.
4. Data and Methods
To empirically test our hypotheses, we conducted an online experiment in September 2014.
Participants were recruited via mailing lists of several German universities, and advertisements
online as well as in a regional German newspaper. As an incentive for participating, we offered
participants the chance to win 100 or 10 Euro. N = 705 participants completed the questionnaire
(58 % female, 42 % male; age M = 31.43, SD = 12.66, 16–79 years). Nearly all participants had a
university-level educational background and thus completed at least ISCED level 3 (Schneider,
2008, p. 95). The treatment consisted of two policy messages. Thus, two potential election
pledges – placed directly after one another and only separated by a set of items on the perception
of the first message-– were presented to every participant. We decided to restrict our analysis to
the message presented first because we could not control for potential carry-over effects of the
first message to the second message. For a realistic setting, the communicator of the message
alternated randomly between the governing parties of the German Bundestag (grand coalition,
legislative period 2013–2017): Christian Democratic Union (CDU) and the Social Democratic
Party (SPD). The messages were fictitious, but realistic.
4
The policy issue of the messages varied
between two typical issue areas of election pledges: economic and social politics (Bara, 2005, p.
4
Most participants perceived the messages as realistic. 46 % even considered it (very) likely that the
named party actually had made the pledge.
PERCEIVED ELECTION PLEDGES
13
590; Mansergh & Thomson, 2007, p. 323; Naurin, 2011, p. 117).
To test H1 and H2, characteristics of the message, specificity and binding force, were also
manipulated. For specificity, we distinguished between vague and specific messages: Vague
messages included neither a numerically measurable goal nor a specific course of action, but the
intended output was still measurable. For example: “We would like to reduce unemployment.” In
contrast, specific messages included a measurable goal and a specified means (action) to reach it.
For example: “By using new programs to strengthen the labor market, we will reduce the
unemployment rate to below 5 per cent.” Messages with high binding force contained the key
word werden (in English: will), whereas messages with a low binding force contained the key
word möchten (would like to).
5
We thus used a 2 (promising party: SPD, CDU) × 2 (policy area
of the message: economic, social) × 2 (concreteness of the message: high, low) × 2 (binding
force of the message: high, low) between-subjects design in our experiment. Participants were
randomly assigned to one of the 16 experimental groups.
6
After reading the message, participants were asked to respond to the following statements
(translated): “This message complies with my understanding of an election pledge” and “I would
describe these kinds of messages as election pledges” on a five-point Likert-type scale, ranging
from 1 (does not apply at all) to 5 (fully applies).
7
To measure the perception of a political
5
This operationalization is a result of a pretest (N = 56) where participants were asked to assess the
binding force of various verbs. We decided to use the verbs möchten (would like to) and werden (will) in
our experiment that were perceived differently in a statistically significant way, t(55) = 5.36, p < .001.
6
The original German version and the English translation of the treatments are presented in the Appendix
(Table A1).
7
The items used to measure the dependent variable might raise the concern that they are not well
understood by respondents as the term “election pledge” might not be known to people. However, in the
German context election pledge (“Wahlversprechen”) is not merely a technical term in the academic
literature but used frequently in the media to assess the feasibility of election pledges in the run-up to an
election and also to assess government performance after the election (e.g.
PERCEIVED ELECTION PLEDGES
14
message as an election pledge, we computed a mean score of responses to both statements (r =
.60). The mean values of respondents’ assessment of the dependent variable for the 16
experimental groups are presented in the Appendix (Table A2).
Respondents further answered a question about their issue involvement after the treatment
(“Does this statement concern you personally?”) on the same five-point Likert-type scale as
mentioned above. Party identification was asked at the end of the survey with a question
established in German electoral research: “In Germany, many people tend towards a particular
political party for a long time, even though they now and then vote for another party. How about
you: Do you, in general, tend towards a particular party? And if yes, which one?” The
respondents could then indicate whether and towards which party they lean. The variable was
dichotomized so that we can differentiate between people who identify with the party making the
statement (coded as 1) and those who identify with another or no party (coded as 0, = reference
category).
Political interest served as moderator (and control) variable and was measured before the
treatment using a well-established question (“Generally speaking, how interested are you in
politics?”). In addition to controlling for the party making the statement and the policy area of
the message, we also controlled for the level of issue competence attributed to the two political
parties (e.g., “When you think of the CDU, do you feel the party is competent in implementing
reforms in the field of labor market policy?”, asked before the treatment). Additionally, trust in
politicians served as control variable (“Please indicate for each of these institutions or groups
how much you personally trust each one of them”, asked before the treatment). Political interest,
issue competence, and trust in politicians were measured on a five-point Likert-type scale.
https://www.zeit.de/politik/deutschland/2017-08/bundestagswahl-parteien-wahlprogramm-vergleich).
Consequently, the term is well-known to respondents.
PERCEIVED ELECTION PLEDGES
15
Additional control variables were age, gender, and education. The correlation matrix for the
independent variables is shown in the Appendix (Table A3).
At the end of the survey experiment respondents were fully debriefed (e.g., informed about the
fictitious nature of the treatments and the aim of the study) and thanked.
5. Results
To test our hypotheses, we first computed a linear regression model to assess the effects of
specificity (H1), binding force (H2), issue involvement (H3), and party identification (H4) (see
Table 1). Additionally, in three extended regression models, we tested moderator effects
separately (H3a, H3b, H5) (see Table 2; only interaction effects are shown).
8
Results supported our hypotheses about the positive effects of message binding force (H2) and
issue involvement (H3). However, effect sizes (beta coefficients) were rather small. Contrary to
expectations, specificity of the message negatively affected participants’ perception of election
pledges. Thus, respondents perceived messages defined as vague as election pledges rather than
messages defined as specific. In other words, H1 was not supported.
Moreover, because party identification did not affect the likelihood that a statement was
perceived as an election pledge, H4 was not supported. However, we found a significantly
positive effect of issue competence. Specifically, participants were more likely to perceive a
statement as a pledge if the party making that statement was also perceived as sufficiently
competent in regard to the policy area of the message.
8
Due to the nature of our dependent variable we also performed ordered logistic regression models.
However, the results do not differ from the results of the linear regression models. See Appendix for the
results of the main effects ordered logit model (Table A4).
PERCEIVED ELECTION PLEDGES
16
As party identification and attributed issue competence could be correlated, we checked for that
and also computed a model without issue competence as control variable. However, the two
variables do not correlate highly (r = .33) and we did not find a significant effect of party
identification in a model without issue competence.
Table 1. Results of the Multiple Linear Regression Analysis on Factors Affecting Which Kind of
Messages Respondents Perceived as Election Pledges
B
SE
ß
Characteristics of the message
Degree of specificity of the message
–.27***
.08
–.13
Degree of binding force of the message
.23**
.08
.10
Subjective variables
Respondent’s issue involvement
.08*
.04
.08
Party identification (own party = 1, other = 0)
–.13
.12
–.04
Control variables
Party of the message (CDU = 0, SPD = 1)
–.03
.09
–.01
Policy of the message (economics = 0, social = 1)
.46***
.09
.21
Respondent’s political interest
–.01
.04
–.01
Party’s attributed issue-competence
.14**
.05
.13
Trust in politicians
–.06
.05
–0.5
Age
.00
.00
.05
Gender (male = 0, female = 1)
.07
.09
.03
Education (low = 0, high = 1)
–.09
.18
–.02
Constant
2.83***
.34
Adjusted R2
.07
F (12, 619)
5.02***
N
632
Note: The dependent variable was computed by respondents’ answers to the questions “This message complies with
my understanding of an election pledge” and “I would describe this kind of messages as election pledges” (five-
point Likert-type scale, ranging from 1 = “does not apply at all” to 5 = “fully applies”).
B = unstandardized regression coefficients; SE = standard errors; ß = standardized regression coefficients (Beta
coefficients), *p ≤ .05, ** p ≤ .01, *** p ≤ .001.
PERCEIVED ELECTION PLEDGES
17
Surprisingly, there was also an effect of the policy issue of the message on the dependent
variable: Participants were more likely to perceive a statement regarding social issues as a pledge
compared to statements about economic policy. This effect of policy on the dependent variable
could be due to the specific content of the messages. Alternatively, it is possible that citizens do
not associate promises on economic policies with real election pledges because governments’
impact on the economy is notoriously limited. Notwithstanding the above, messages of both
political parties, CDU and SPD, were similarly perceived as election pledges. Furthermore, there
were no main effects on the perception of political messages as election pledges of political
interest, trust in politicians, age, gender, and education.
Table 2. Moderator Analyses
B
SE
Issue involvement × Specificity
.08
.08
Issue involvement × Binding force
.11
.08
Political interest × Party identification
–.04
.11
Note: B = unstandardized regression coefficients; SE = standard errors. See Appendix for full model results.
In terms of any moderating effects, contrary to our hypothesis, we could not find significant
interaction effects. Hence, the effects of specificity (H3a) and binding force (H3b) were not
moderated by issue involvement. Moreover, even though political interest was assumed to
influence individuals’ reliance on heuristic cues, political interest did not moderate the effect of
party identification on election pledge perception (H5).
6. Discussion
PERCEIVED ELECTION PLEDGES
18
In sum, our results suggest that citizens are more likely to perceive vague but bindingly framed
political messages as election pledges (compared to more specific and less bindingly framed
messages). Further, respondents’ issue involvement positively predicted the extent to which
messages were perceived as election pledges. Thus, citizens’ understanding of election pledges
depended on characteristics as well as subjective impressions of the message. Turning to
characteristics of the message first, our results indicate that people were more likely to perceive
vague messages as election pledges, at least in contrast to more specific messages. The reason
could be that people are more familiar with vague than with specific election pledges because
election pledges are often formulated as short slogans that do not contain a specific goal or
course of action. Hence, they could identify vague statements more easily as election pledges
than specific ones. This can have repercussions for citizens’ evaluation of election pledges.
Vague election pledges require the individual to interpret the meaning and implications of the
pledge in order to evaluate pledge fulfillment. However, as many people lack the political
knowledge to do this correctly, their interpretations might rely on cognitive biases like “wishful
thinking” (related to that: people’s perception of what they think parties have promised deviates
from parties’ actual campaign commitments, see Naurin, 2011). Therefore, vague election
pledges may give people hope – albeit perhaps unrealistically. Further, there is empirical
evidence that vague messages are actually more likely to be broken than specific ones (Thomson
& Costello 2008, p. 11).
With regard to the degree of binding force, our findings support the assumption that citizens
expect election pledges to be predominantly worded in a binding format. This may be the case
because the wording of the message can reflect a project’s degree of certainty and commitment
(Royed, 1996). However, this perception probably also depends on linguistic and cultural
PERCEIVED ELECTION PLEDGES
19
practices. Further examination of this assumption in the context of other countries is needed – for
example also with regard to characteristics of the political system (i.e. one-party vs. coalition
governments). Moreover and somewhat surprisingly, in our study the policy of the message
influenced whether people identified statements as election pledges. Unemployment was more
strongly perceived as an election pledge than statements about the economy. Whereas this result
could depend on the specific stimulus we used and therefore needs to be conceptually replicated,
one possible explanation could be that as a result of the economic crisis of 2007/2008, people
have become more aware of the fact that economic policy is influenced by many external
influences, and thus are less likely to perceive statements on economic issues as actionable
election pledges. Another explanation could be that for our sample (which consisted
predominantly of university students), social policy is more important than economic policy.
However, in our study, respondents felt more affected by the economic situation statement than
by the social policy statement.
Regarding subjective factors, issue involvement, operationalized as a respondent’s personal
concern with the statement, positively predicted the extent to which individuals defined political
statements as election pledges. Moreover, our results showed that the degree of issue competence
that people attributed to a party was relevant for citizens’ definition of election pledges. It is
possible that people think (or hope) that the likelihood of these statements being implemented is
higher when they deem a party competent, whereas they deem messages relating to policy areas
where they perceive the promising party as not competent as mere advertising. In our view, this
result is important as it highlights the idea that people’s general understanding of election
pledges is biased by personal matters (Naurin, 2011; Thomson, 2011). This also has implications
for citizens’ assessment of government performance (pledge fulfillment, e.g., in terms of
PERCEIVED ELECTION PLEDGES
20
electoral accountability). A varying basis of evaluation, namely, which campaign messages one
perceived as election pledges, should lead to individual weightings when citizens asses the
government’s pledge fulfillment. Thus, subjectiveness is a key factor in understanding citizens’
perception as well as their evaluation of election pledges. However, we did not find any effect of
party identification on our dependent variable. Further, it was not associated with the perception
of a statement as an election pledge, nor was this relationship moderated by political interest.
Consequently, people are seemingly not more likely to perceive statements as election pledges if
these statements come from a party that they identify with as opposed to one that they do not
identify with. One explanation for this result could relate to the widespread stereotype of the
pledge-breaking politician making promises that he/she cannot keep. This line of reasoning is
supported by the results of our pretest. Here, the assessment of the word “promise” differed
strongly: it was deemed very binding by some and not binding by others. Since public opinion on
election pledges thus is in part negative, making a promise can have negative repercussions on a
party. If party supporters are aware of this, it should be reasonable for them to not perceive
statements of their own party as election pledges to avoid being disappointed about their party in
case of the party breaking the election pledge.
Finally, our study has several limitations. Due to small effect sizes, our results should be
replicated in further studies to confirm their veracity. As our sample consisted predominantly of
highly educated (and politically interested) individuals, it is not representative of the German
population. Furthermore, the experimental framing of the study reduced its external validity
because the treatment was presented without any further context information (e.g., the
communication channel of the message, competing messages). Nonetheless, because our analysis
PERCEIVED ELECTION PLEDGES
21
focused on the message itself, we considered this design to be suitable – especially in terms of its
internal validity.
In the present article, we have argued that by making election pledges, parties can be held
accountable for their policy achievements at the next election. However, the mechanics of this
system hinges on citizens’ perception of election pledges. Apparently, people are prone to
perceive political campaign messages as election pledges if they believe that the given political
party can, and should fulfill them. That are: Vague but binding messages concerning policy areas
political parties can exert control over (e.g., social policy pledges, compared to economic policy
pledges, or areas in which they are deemed issue-competent) as well as messages they should
fulfill as citizens’ assess them to be personally relevant. Our results indicate that the
characteristics of a message and various subjective factors play a role in pledge perception. In
light of this, social science research should pay more attention to the subjective perception and
evaluation of election pledges, as well as to the expectations that people place on politicians and
their pledges.
PERCEIVED ELECTION PLEDGES – (original publication of the article available at
Science Direct: Electoral Studies, 2018)
22
References
Anderson, C. J. (2007). The end of economic voting? Contingency dilemmas and the limits of
democratic accountability. Annual Review of Political Science, 10, 271–296.
doi:10.050806.155344.
Bara, J. (2005). A question of trust: Implementing party manifestos. Parliamentary affairs,
58(3), 585–599. doi:10.1093/pa/gsi053.
Born, A., van Eck, P., & Johannesson, M. (2018). An experimental investigation of election
promises. Political Psychology, 39(3), 685–705. doi:10.1111/pops.12429.
Campbell, A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). The American voter. New
York: Wiley.
Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source
versus message cues and persuasion. Journal of Personality and Social Psychology,
39(5), 752–766.
Downs, A. (1957). An economic theory of democracy. New York, NY: Harper.
Duch, R. M., Palmer, H. D., & Anderson, C. J. (2000). Heterogeneity in perceptions of national
economic conditions. American Journal of Political Science, 44(4), 635–652. doi:
10.2307/2669272.
Entman, R. M. (2007). Framing bias: media in the distribution of power. Journal of
communication, 57, 163–173. doi:10.1111/j.1460-2466.2006.00336.x.
Ferguson, M. J. (2010). Pledge redemption in a grand coalition government. Paper presented at
the MPSA Annual Conference Chicago, IL.
Ferguson, M. J. (2012). Pledge redemption in Germany: Re-examining of the Schröder II and
Merkel I Governments. Paper presented at the MPSA Annual Conference Chicago, IL.
Fiorina, M. P. (1981). Retrospective voting in American national elections. New Haven, CT:
Yale University Press.
PERCEIVED ELECTION PLEDGES
23
Funk, C. L., & Garcia-Monet, P. (1997). The relationship between personal and national
concerns in public perceptions about the economy. Political Research Quarterly, 50(2),
317–342. doi: 10.2307/448960.
Grossback, L. J., Peterson, D., & Stimson, J. (2005). Comparing competing theories on the
causes of mandate perceptions. American Journal of Political Science, 49, 406–419. doi:
10.1111/j.0092-5853.2005.00131.x.
Hakansson, N., & Naurin, E. (2014). Promising ever more: An empirical account of Swedish
parties' pledge making during 20 years. Party Politics, 1–12.
doi:10.1177/1354068814549338.
Johnston, R. (2006). Party identification: Unmoved mover or sum of preferences? Annual
Review of Political Science, 9, 329–351. doi:10.1146/annurev.polisci.9.062404.170523.
Kahneman, D., Slovic, P., & Tversky, A. (1982). Judgment under uncertainity. Heuristics and
biases. New York: Cambridge University Press.
Kalogeropoulou, E. (1989). Election promises and government performance in Greece: PASOK's
fulfilment of its 1981 election pledges. European Journal of Political Research, 17(3),
289–311. doi:10.1111/j.1475-6765.1989.tb00195.x.
Key, V. O., & Cummings, M. C. (1966). The responsible electorate: Rationality in presidential
voting, 1936–1960. Harvard, MA: Harvard University Press.
Klauer, K. C. (2008). Soziale Kategorisierung und Stereotypisierung. In L.-E. Petersen & B. Six
(Eds.), Stereotype, Vorurteile und soziale Diskriminierung (pp. 23–32). Weinheim, Basel:
Beltz Verlag.
Klingemann, H.-D., Hoffebert, R. I., & Budge, I. (1994). Parties, policies and democracy.
Boulder, CO: Westview Press.
Kostadinova, P. (2017). Party pledges in the news: which election promises do the media report?
Party Politics, 23(6), 636–645. doi:10.1177/1354068815611649.
PERCEIVED ELECTION PLEDGES
24
Kruglanski, A. W., & Ajzen, I. (1983). Bias and error in human judgement. European Journal of
Social Psychology, 13(1), 1–44. doi: 10.1002/ejsp.2420130102.
Kuklinski, J. H., & Quirk, P. J. (2000). Reconsidering the rational public: Cognition, heuristics,
and mass opinion. Elements of reason: Cognition, choice, and the bounds of rationality,
153–182.
Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480–498.
Lewis-Beck, M. S., & Stegmaier, M. (2000). Economic determinants of electoral outcomes.
Annual Review of Political Science, 3, 183–219.
doi:https://doi.org/10.1146/annurev.polisci.3.1.183.
Lindgren, E. (2018). Changing policy with words: how persuasive words in election pledges
influence voters' beliefs about policies. Mass Communication and Society, 21, 425–449.
doi:10.1080/15205436.2017.1406522.
Lindgren, E., & Naurin, E. (2017). Election pledge rhetoric: selling policy with words.
International Journal of Communication, 11, 2198–2219.
Lodge, M., & Taber, C. S. (2000). Three steps toward a theory of motivated political reasoning.
In A. Lupia, M. D. McCubbins, & S. L. Popkin (Eds.), Elemets of reason: understanding
and expanding the limits of political rationality (pp. 183–213). London: Cambridge
University Press.
Mansergh, L., & Thomson, R. (2007). Election pledges, party competition, and policymaking.
Comparative Politics, 39(3), 311–329.
Maurer, M. (2007). Fakten oder Floskeln? Publizistik, 52(2), 174–190. doi:10.1007/s11616-007-
0082-2.
Mayer, J. (2017). Die Parteiidentifikation. Eine Konstruktvalidierung neuer Maße auf Basis des
Ansatzes sozialer Identität. Wiesbaden: Springer Fachmedien.
McDonald, M. D., & Budge, I. (2005). Elections, parties, democracy. Conferring the median
mandate. Oxford: Oxford University Press.
PERCEIVED ELECTION PLEDGES
25
McKelvey, R. D., & Ordeshook, P. C. (1986). Information, electoral equilibria, and the
democratic ideal. The Journal of Politics, 48(4), 909–937. doi: 10.2307/2131005.
Naurin, E. (2011). Election promises, party behaviour and voter perceptions. Houndmills:
Palgrave Macmillan.
Naurin, E. (2014). Is a promise a promise? Election pledge fulfilment in comparative perspective
using Sweden as an example. West European Politics, 37(5), 1046–1064.
doi:10.1080/01402382.2013.863518.
Naurin, E., & Oscarsson, H. E. (2017). When and why are voters correct in their evaluations of
specific government performance? Political Studies, 65, 860–876.
doi:10.1177/0032321716688359.
Naurin, E., Soroka, S., & Markwat, N. (2015). Are broken pledges more important to voters'
evaluation of government performance than fulfilled pledges? On negativity biases in
specific accountability processes. Paper presented at the Annual Meeting of the American
Political Science Association, San Francisco.
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In L.
Berkowitz (Ed.), Advances in experimental social psychology (Vol. 19, pp. 123–205).
Popkin, S. L. (1991). The reasoning voter: Communication and persuasion in presidential
campaigns. Chicago: University Of Chicago Press.
Rahn, W. M. (1993). The role of partisan stereotypes in information processing about political
candidates. American Journal of Political Science, 37(2), 472–496.
Rallings, C. (1987). The influence of election programmes: Britain and Canada 1945–1979. In I.
Budge, D. Robertson, & D. Hearl (Eds.), Spatial analyses of post-war election
programmes in 19 democracies (pp. 1–14). Cambridge, MA: Cambridge University
Press.
Rose, R. (1984). Do parties make a difference? (2 ed.). Basingstoke: Palgrave Macmillan.
PERCEIVED ELECTION PLEDGES
26
Ross, L. (1977). The intuitive psychologist and his shortcomings. In L. Berkowitz (Ed.),
Advances in experimental social psychology (Vol. 10, pp. 173–220). New York, NY:
Academic Press.
Royed, T. J. (1996). Testing the mandate model in Britain and the United States: Evidence from
the Reagan and Thatcher eras. British Journal of Political Science, 45, 45–80.
Schneider, S. L. (2008). Applying the ISCED-97 to the German educational qualifications. In S.
L. Schneider (Ed.), The international standard classification of education. An evaluation
of content and criterion validity for 15 European countries (pp. 76–102). Mannheim,
Germany: Mannheimer Zentrum für Europäische Sozialforschung.
Searle, J. R., & Vanderveken, D. (1985). Foundations of illocutionary logic. Cambridge, MA:
Cambridge University Press.
Sherif, C. W., Kelly, M., Rodgers, H. L., Sarup, G., & Tittler, B. I. (1973). Personal
involvement, social judgment and action. Journal of Personality and Social Psychology,
27(3), 311–328. doi: 10.1037/h0034948.
Sherif, M., & Hovland, C. I. (1961). Social judgment: Assimilation and contrast effects in
communication and attitude change. Westport, CT: Greenwood Press.
Simon, H. A. (1990). Invariants of human behavior. Annual Review of Psychology, 41, 1–19.
Simon, H. A. (1995). Rationality in political behavior. Political Psychology, 16, 45–61.
Strickland, A. A., Taber, C. S., & Lodge, M. (2011). Motivated reasoning and public opinion.
Journal of Health Politics, Policy and Law, 36(6), 935–944. doi: 10.1215/03616878-
1460524.
Taber, C. S., & Lodge, M. (2006). Motivated skepticism in the evaluation of political beliefs.
American Journal of Political Science, 50(3), 755–769. doi: 10.1111/j.1540-
5907.2006.00214.x.
Thomson, R. (2011). People’s evaluations of the fulfillment of election pledges: Evidence from
Ireland. The Journal of Politics, 73(1), 187–201. doi:10.1017/s0022381610000952.
PERCEIVED ELECTION PLEDGES
27
Thomson, R., & Costello, R. (2008). Election pledges and their enactment in coalition
governments: A comparative analysis of Ireland. Journal of Elections, Public Opinion
and Policy, 18(3), 239–256. doi:10.1080/17457280802227652.
Thomson, R., Royed, T. J., Naurin, E., Artés, J., Costello, R., Ennser-Jedenastik, L.,…
Praprotnik, K. (2017). The fulfillment of parties' election pledges: a comparative study on
the impact of power sharing. American Journal of Political Science, 61(3), 527–542. doi:
10.1111/ajps.12313.
Thomson, R., & Brandenburg, H. (2018). Trust and citizens' evaluations of promise keeping by
governing parties. Political Studies. doi: 10.1177/0032321718764177.
PERCEIVED ELECTION PLEDGES
28
Appendix
Table A1: Treatment Material
Condition
Treatment (German)
Treatment (English translation)
Party: SPD, policy: social,
binding force: low, specificity: low
SPD: „Wir möchten die Arbeitslosenquote senken.“
SPD: “We would like to reduce the unemployment
rate.”
Party: SPD, policy: social,
binding force: low, specificity: high
SPD: „Durch neue Programme zur Belebung des
Arbeitsmarktes möchten wir die Arbeitslosenquote auf
unter 5 Prozent senken.“
SPD: “With new programs to strengthen the labor
market, we would like to reduce the unemployment rate
to below 5 per cent.”
Party: SPD, Policy: social,
binding force: high, specificity: low
SPD: „Wir werden die Arbeitslosenquote senken.“
SPD: “We will reduce the unemployment rate.”
Party: SPD, policy: social,
binding force: high, specificity: high
SPD: „Durch neue Programme zur Belebung des
Arbeitsmarktes werden wir die Arbeitslosenquote auf
unter 5 Prozent senken.“
SPD: “With new programs to strengthen the labor
market, we will reduce the unemployment rate
to below 5 per cent.”
Party: SPD, policy: economic,
binding force: low, specificity: low
SPD: „Wir möchten die Wirtschaft ankurbeln.“
SPD: “We would like to boost the economy.”
Party: SPD, policy: economic,
binding force: low, specificity: high
SPD: „Wir möchten die Wirtschaft ankurbeln, indem
wir die Binnennachfrage um 5 Prozent steigern.“
SPD: “By strengthening domestic demand, we would
like to boost the economy by 5 per cent.”
Party: SPD, policy: economic,
binding force: high, specificity: low
SPD: „Wir werden die Wirtschaft ankurbeln.“
SPD: “We will boost the economy.”
Party: SPD, policy: economic,
binding force: high, specificity: high
SPD: „Wir werden die Wirtschaft ankurbeln, indem
wir die Binnennachfrage um 5 Prozent steigern.“
SPD: “By strengthening domestic demand, we will
boost the economy by 5 per cent.”
Party: CDU, policy: social,
binding force: low, specificity: low
CDU: „Wir möchten die Arbeitslosenquote senken.“
CDU: “We would like to reduce the unemployment
rate.”
PERCEIVED ELECTION PLEDGES
29
Party: CDU, policy: social,
binding force: low, specificity: high
CDU: „Durch neue Programme zur Belebung des
Arbeitsmarktes möchten wir die Arbeitslosenquote auf
unter 5 Prozent senken.“
CDU: “With new programs to strengthen the labor
market, we would like to reduce the unemployment rate
to below 5 per cent.”
Party: CDU, policy: social,
binding force: high, specificity: low
CDU: „Wir werden die Arbeitslosenquote senken.“
CDU: “We will reduce the unemployment rate.”
Party: CDU, policy: social,
binding force: high, specificity: high
CDU: „Durch neue Programme zur Belebung des
Arbeitsmarktes werden wir die Arbeitslosenquote
auf unter 5 Prozent senken.“
CDU: “With new programs to strengthen the labor
market, we will reduce the unemployment rate
to below 5 per cent.”
Party: CDU, policy: economic,
binding force: low, specificity: low
CDU: „Wir möchten die Wirtschaft ankurbeln.“
CDU: “We would like to boost the economy.”
Party: CDU, policy: economic,
binding force: low, specificity: high
CDU: „Wir möchten die Wirtschaft ankurbeln, indem
wir die Binnennachfrage um 5 Prozent steigern.“
CDU: “By strengthening domestic demand, we would
like to boost the economy by 5 per cent.”
Party: CDU, policy: economic,
binding force: high, specificity: low
CDU: „Wir werden die Wirtschaft ankurbeln.“
CDU: “We will boost the economy.”
Party: CDU, policy: economic,
binding force: high, specificity: high
CDU: „Wir werden die Wirtschaft ankurbeln, indem
wir die Binnennachfrage um 5 Prozent steigern.“
CDU: “By strengthening domestic demand, we will
boost the economy by 5 per cent.”
PERCEIVED ELECTION PLEDGES
30
Table A2. Mean Values of Respondents’ Assessment of the Dependent Variable (Perception of a Message as Election Pledge)
Treatment:
Party and policy
of the message
Degree of binding
force of the message
Degree of
specificity of the
message
n
M
SD
SPD, social policy
high
high
40
3.60
0.91
high
low
46
3.68
1.05
low
high
42
3.86
0.89
low
low
44
3.80
1.03
SPD, economic policy
high
high
47
3.32
1.12
high
low
42
2.75
1.21
low
high
44
3.60
1.00
low
low
44
3.13
1.08
CDU, social policy
high
high
47
3.53
0.98
high
low
43
3.45
1.06
low
high
46
4.02
0.91
low
low
44
3.74
1.08
CDU, economic policy
high
high
47
3.34
1.07
high
low
42
3.12
1.07
low
high
42
3.35
1.14
low
low
45
3.40
1.17
Note: The scale of the dependent variable ranged from 1 = “does not apply at all” to 5 = “fully applies”
PERCEIVED ELECTION PLEDGES
31
Table A3. Correlation-Matrix of the Independent Variables
1
2
3
4
5
6
7
8
9
10
11
12
1. Degree of specificity of the
message a
1
2. Degree of binding force of the
message a
.02
1
3. Respondent’s issue involvement
–.11**
–.08*
1
4. Party identification
.05
–.03
.12***
1
5. Party of the message a
–.02
–.00
.02
.08*
1
6. Policy of the message a
–.01
–.00
–.24***
–.02
–.01
1
7. Respondent’s political interest
–.05
.01
.11**
–.02
.04
–.06
1
8. Party’s attributed issue
competence
–.00
.05
.09*
.33***
.06
–.01
–.07
1
9. Trust in politicians
–.02
.01
.11**
.16***
.01
–.07
.15***
.33**
1
10. Age
.02
–.01
–.07
.05
.00
–.06
.26**
–.10**
–.07
1
11. Gender
.00
–.03
.05
.01
–.08*
.04
–.36***
.08*
–.01
–.27**
1
12. Education
–.02
.02
.06
–.04
.05
–.03
–.09*
.06
.12**
–.34***
.12*
1
Note: a dichotomous variable, unmarked variables were treated as interval-scaled, *p ≤ .05, ** p ≤ .01, *** p ≤ .001
PERCEIVED ELECTION PLEDGES
32
Table A4. Results of Main Effects Ordered Logistic Regression Model
B
SE
Characteristics of the message
Degree of specificity of the message
–.50***
.14
Degree of binding force of the message
.35**
.14
Subjective variables
Respondent’s issue involvement
.14*
.07
Party identification
–.20
.20
Control variables
Party of the message (CDU = 0, SPD = 1)
–.07
.14
Policy of the message (economics = 0, social
= 1)
.78***
.15
Respondent’s political interest
.01
.07
Party’s attributed issue competence
.22**
.08
Trust in politicians
–.14
.09
Age
.01
.01
Gender
.04
.16
Education
–.09
.18
/cut1
–2.25
.60
/cut2
–1.69
.59
/cut3
–.80
.58
/cut4
–.34
.58
/cut5
.42
.58
/cut6
.94
.58
/cut7
2.31
.59
/cut8
2.73
.59
Pseudo R2
.02
Log likelihood
–1230.56
N
632
Note: The dependent variable was computed by respondents’ answers to the questions “This message complies with
my understanding of an election pledge” and “I would describe this kind of messages as election pledges” (five-
point Likert-type scale, ranging from 1 = “does not apply at all” to 5 = “fully applies”).
B = unstandardized regression coefficients; SE = standard errors; *p ≤ .05, ** p ≤ .01, *** p ≤ .001.
PERCEIVED ELECTION PLEDGES
33
Table A5. Results of Linear Regression Model with Interaction Effect on Issue Involvement x
Specificity
B
SE
Characteristics of the message
Degree of specificity of the message
–.46*
.20
Degree of binding force of the message
.22**
.08
Subjective variables
Respondent’s issue involvement
.04
.06
Party identification
–.13
.12
Control variables
Party of the message (CDU = 0, SPD = 1)
–.02
.09
Policy of the message (economics = 0, social
= 1)
.45***
.09
Respondent’s political interest
–.01
.04
Party’s attributed issue competence
.13**
.05
Trust in politicians
–.06
.05
Age
.00
.00
Gender
.06
.09
Education
–.09
.18
Interaction
Issue involvement x Specificity
.08
.08
Constant
2.95***
.36
Adjusted R2
.07
F (13, 618)
4.71***
N
632
Note: The dependent variable was computed by respondents’ answers to the questions “This message complies with
my understanding of an election pledge” and “I would describe this kind of messages as election pledges” (five-
point Likert-type scale, ranging from 1 = “does not apply at all” to 5 = “fully applies”).
B = unstandardized regression coefficients; SE = standard errors; *p ≤ .05, ** p ≤ .01, *** p ≤ .001.
PERCEIVED ELECTION PLEDGES
34
Table A6. Results of Linear Regression Model with Interaction Effect on Issue Involvement x
Binding Force
B
SE
Characteristics of the message
Degree of specificity of the message
–.28***
.08
Degree of binding force of the message
–.03**
.20
Subjective variables
Respondent’s issue involvement
.03
.05
Party identification
–.13
.12
Control variables
Party of the message (CDU = 0, SPD = 1)
–.02
.09
Policy of the message (economics = 0, social
= 1)
.46***
.09
Respondent’s political interest
–.01
.04
Party’s attributed issue competence
.14**
.05
Trust in politicians
–.07
.05
Age
.00
.00
Gender
.06
.09
Education
–.09
.18
Interaction
Issue involvement x Binding force
.11
.08
Constant
3.00***
.35
Adjusted R2
.07
F (13, 618)
4.80***
N
632
Note: The dependent variable was computed by respondents’ answers to the questions “This message complies with
my understanding of an election pledge” and “I would describe this kind of messages as election pledges” (five-
point Likert-type scale, ranging from 1 = “does not apply at all” to 5 = “fully applies”).
B = unstandardized regression coefficients; SE = standard errors; *p ≤ .05, ** p ≤ .01, *** p ≤ .001.
PERCEIVED ELECTION PLEDGES
35
Table A7. Results of Linear Regression Model with Interaction Effect on Political Interest x
Party Identification
B
SE
Characteristics of the message
Degree of specificity of the message
–.27***
.08
Degree of binding force of the message
.23**
.08
Subjective variables
Respondent’s issue involvement
.08*
.04
Party identification
.01
.42
Control variables
Party of the message (CDU = 0, SPD = 1)
–.03
.09
Policy of the message (economics = 0, social
= 1)
.45***
.09
Respondent’s political interest
.00
.05
Party’s attributed issue competence
.14**
.05
Trust in politicians
–.06
.05
Age
.01
.00
Gender
.07
.09
Education
–.09
.18
Interaction
Political interest x Party identification
–.04
.11
Constant
2.80
.35
Adjusted R2
.07
F (13, 618)
4.63***
N
632
Note: The dependent variable was computed by respondents’ answers to the questions “This message complies with
my understanding of an election pledge” and “I would describe this kind of messages as election pledges” (five-
point Likert-type scale, ranging from 1 = “does not apply at all” to 5 = “fully applies”).
B = unstandardized regression coefficients; SE = standard errors; *p ≤ .05, ** p ≤ .01, *** p ≤ .001.