Saying Sorry in Politics
The Effect of Apologizing on the Public Opinion
Name: Suzanne Hartog Student ID: 6167500
Master’s Thesis
Master’s Program Communication Science Supervisor: Rachid Azrout
Date of Completion: June 25th 2015
ABSTRACT
Apologizing in crisis communication has been a popular research topic for the last two decades. However, the effect of apologizing of government members on the public opinion has not been researched that much. This thesis provides a valuable insight in how apologizing affects the public opinion in the Netherlands. An online experiment was conducted as a 2x2 factorial design of independent variables (no apology versus apology and no political
alignment versus political alignment) on the image of government member, the evaluation of
the response of the government member, and the evaluation of the political system. Results
indicate that government members will be evaluated more positively on the characteristics
sympathetic, integrity and critical. The response strategy was evaluated more positively when government members offered an apology compared to offering no apology. There was no
moderation effect found between apologizing and political alignment.
INTRODUCTION
Government members often find themselves in crisis situations where they are held responsible for their actions. It is becoming more common that government members make official apologies during political crises. Mills (2001) speaks about the emergence of a new culture of apology since the start of the post-cold war era ‘where the public discourse has increasingly taken on the ethics and informality of private discourse and produced a different set of expectations from those we are used to (p. 112)’. There is a change in sensibility in communications and emotions have become central elements in politics (Mills, 2001, p. 112).
This phenomenon of apologizing has also spread to the Netherlands. Ed van Thijn (1999) even calls the political system in the Netherlands the ‘Sorry Democracy’. In the past year and a half there have been a lot of apologies in the political sphere. For example, in January 2014, the state secretary for Finance, Frans Weekers made an official apology for the way duped citizens did not receive supplements they were entitled to. After his statement he resigned from his duties as state secretary. Also, in February 2014, the minister of the Interior and Kingdom Relations Ronald Plasterk apologized for making false statements about the relationship between the Dutch security service, the AIVD, and the NSA during a broadcast of EenVandaag. In October 2014, the minister of Foreign Affairs, Frans Timmermans, said sorry for his performance during the broadcast of Pauw and Witteman where he emotionally said something classified about the victims of the MH17 crash. In February 2015 state secretary for Health, Welfare and Sport, Martin van Rijn, made official apologies for the disturbance caused by delayed payments of personal budgets (PGB). The latest example of a government member apologizing is Henk Kamp, minister of Economic Affairs, who apologized in March 2015 for the way the government neglected the interests of the inhabitants of Groningen where earthquakes were caused by gas extraction.
Although before mentioned examples illustrate that there are a lot government members apologizing in political crises, the effect of apologizing in political communication has not been explored that much. Apologizing is an extensive research topic within the literature of crisis and corporate communication (Bradford & Garrett, 1995; Kim et al. 2004; Tucker et al., 2006), but more research is needed in the political context say Tucker et al. (2006). Like Sheldon and Sallot (2009) state, reputations of politicians are extremely valuable and can influence voting behavior. There is a theoretical gap in political crisis communication, which will be addressed in this thesis.
Political institutions cannot be considered as normal organizations and the relationship between the government and citizens is fundamentally different than the relationship between cooperate organizations and their stakeholders.According to Benoit (1997) politicians have to think about getting reelected and therefore they need to take the public opinion into account. The main goal of this thesis is to determine if apologies of government members have an effect on public evaluation of government members and the political system.
Moreover, in crisis it is important to uphold the reputation with the supporters of the political party for getting reelected and therefore it is interesting to investigate if political alignment plays a role in the evaluation of apologizing in politics. Can we expect the same effects between people ideologically close to apologizing politicians and people further away? If apologies also prove successful in politics, government members can find a way to influence their reputation after crises and it will offer guidelines to politicians to act in political crises. The opposite results will also give guidelines to government members in the Netherlands and show that apologizing is unnecessary or can even be harmful for a politician’s reputation. Either way, this thesis will add more knowledge to crisis communication in the Dutch political sphere. The central aim of this thesis is to examine the effect of apologizing of government members on the public opinion and therefore the following research question is formulated:
‘What is the effect of government members apologizing on the image of government members, the evaluation of the response of the government member and the evaluation of the political
system?’
To answer the research question, this thesis is divided into four different sections. In the first section a theoretical overview of apologizing in politics and three research questions and three
hypotheses are given. In the second section the method of the thesis is explained. In the third section the results are shown and in the last section a conclusion of the thesis is drawn, followed by a discussion.
THEORETICAL BACKGROUND
‘No sensible person ever made an apology.’ – Ralph W. Emerson (1803-1882)
After crises, organizations consider different strategies to respond to crises. Research by Coombs & Holladay (2008, p. 252) shows that organizations need to transmit information to all different stakeholders who are involved and help them in dealing with the crisis. Huang (2006, p. 181) stated that a response strategy ensures that actors can provide information to observers to influence their perception of the situation. Not only crises could have negative effects for different stakeholders, but also reputations of the accused organizations could be threatened according to Davies et al (2003). Also crises threaten the reputations of political actors (Huang, 2006).
When in crisis, Schlenker (1980) distinguished four responses in crisis communication: excuses, justifications, denials, and apologies. Almost twenty years later William Benoit (1997) categorized five different response strategies: denial, evasion of responsibility, reducing offensiveness of event, corrective action and mortification.
Apologizing is an example of a crisis response strategy: the mortification strategy. The act of apologizing, say Benoit & Drew (1997), implies that an organization or single actor accepts responsibility for a crisis and is asking for forgiveness. Schmitt et al (2004) state that
the accused actor accepts responsibility expresses empathy and promises not to repeat its behavior in the future.
In corporate crisis communication several researches show that apologizing has an effect on the accused actor’s image. For example the article by Tucker et al (2006) provides evidence that leadership perceptions among followers are evaluated more positively when an actor apologizes than when an actor does not apologize. The article also describes the lack of research on the relationship between leader and follower in experimental research, which is of special interest for this thesis. Also research by Bradford & Garrett (1995, p. 877) states that the apology strategy had the strongest positive effect on the perceptions of an organization’s reputation in their corporate communication response model. Research by Kim et al (2004) suggests that apologies can play a very important role in rebuilding relationships with stakeholders. The results show that trust between the accused organizations and the involved parties was repaired with more success when the organizations apologized (Kim et al, 2004, p. 104).
Evaluation image of politician
Elaborating on theory from corporate communication, this thesis examines the effect of government members apologizing, accepting responsibility for the crisis and asking for forgiveness of the public. Research by Tucker et al (2006) shows that apologizing leads to victims of mistakes seeing their leaders as more transformational (p. 195). Evidence was found that apologizing leads to a different perception of followers of their leaders (p.205). Apologies make leaders appear more human and when leaders apologize they show that they care for their victims, followers and/or employees (Tucker et al, 2006, p. 203).
According to Tucker et al (2006) more context-based research is required to see the effect of apologizing in different situations, such as politics. Benoit wrote in his work that there are
differences between organizations repairing their image and individuals repairing their image (1997, p. 177). In this research it is also expected that there are differences between (normal) organizations and government institutions. Contradicting earlier discussed research, an experimental research by Sigal et al. (1988, p. 273) shows us that politicians who were accused of sexual and financial misconduct were evaluated more positively when they denied than when accepted responsibility and they apologized. The politicians were evaluated on honesty, ethics and trustworthiness (p. 277).
This thesis does not address specific financial and sexual accusations, but is investigating the effect of the phenomenon apologizing of government members in the Netherlands. There is not so much research investigating the relationship between government members and the public after apologizing in the Netherlands. According to the ‘Reputatiegroep’ (2012) reputations are rather harmed than strengthened when apologies are offered. The ‘Reputatiegroep’ created an annual list of apologies where they research the reputation in the media and a lot of politicians are mentioned on the ‘Sorry-list’. But the main effect of apologizing on the image of the politician by citizens has not been investigated in the Netherlands.
Outcomes of corporate crisis communication suggest that apologizing leads to a more positive evaluation of organizations (Bradford & Garrett, 1995; Kim et al. 2004; Tucker et al., 2006). Therefore it could be hypothesized that apologizing will also lead to a more positive evaluation of the politician. But research in political crisis communication (Sigal et al., 1988; Reputatiegroep, 2012) suggests that apologizing will only harm the reputation of a politician. This could be because of the different relationship between citizens and government members. Instead of needing a transformational leader (Tucker et al., 2006) it is possible that citizens need their government members to be strong and tough and do not need a humble
politician in charge of their government. Due to mixed outcomes of research, the following research question is formulated:
RQ1: Does an apology of a politician in crisis communication have an influence on the
politician’s image?
Evaluation response strategy of the politician
Not only it is interesting to measure the effect of the reputation of the politician, but it is also interesting to see how people perceive the appropriateness of the crisis response strategy. Apologies are common expressions, but the interesting question is when do citizens think they are they appropriate? According to Scher & Darley (1996) apologies are appropriate when an actor ‘has violated a social norm (p. 127). The research stated: ‘the greatest improvement in perceptions came from the addition of one apology strategy – i.e., the offering of an apology, compared to no apology (Scher & Darley, 1996, p. 137)’. As mentioned earlier, the expectations of leaders of normal organizations and the expectations of leaders of government institutions differ from each other. Will the appropriateness of an apology be different in a political context? How do citizens evaluate the response of a politician when comparing an apology to giving no apology? To investigate this, the following sub question is formulated:
RQ2: Does an apology of a politician in crisis communication have an influence on
Satisfaction political system
Furthermore, political organizations differ from other organizations as it comes to the relationship they have with the citizens. Government members do not only have a relationship with the citizens, but they also represent the political system in a country. Research in corporate crisis communication by Verhoeven et al (2012) shows that a crisis has more impact on the reputation of the corporate organization than on the reputation of the spokesperson of that same organization. The crisis is viewed as a ‘collective organizational failure’ instead of an individual mistake. What would be the effect for the political system after a government member apologizes? As mentioned earlier, the relationship between government members and citizens is complex. Government members in the Netherlands are held accountable for the ministries they represent and they bear ultimate responsibility, they call it ‘ministerial responsibility’ (Parlement, 2015). Ministerial responsibility means that the government members, ministers and state secretaries, are held accountable by the members House of Representatives for their own behavior and actions, are responsible for the behavior of the King and are responsible for the behavior and decisions of the civil servants of the ministry (Parlement, 2015). This could have its influence on the public opinion. Do they see the mistakes by the government members as individual mistakes or as ‘collective organizational failure’? What happens with the evaluation of the political system after government members apologize?
Elaborating on this theory, the public opinion is also measured by addressing political satisfaction of the system and the following research sub question will be tested:
RQ3: Does an apology of a politician in crisis communication have an influence on
Political Alignment
To add another dimension to this thesis, political alignment was taken into account. The current divide in the national government of left-wing party Partij van de Arbeid (PvdA) and the right-wing party Volkspartij voor de Vrijheid en Democratie (VVD), offers a unique opportunity to investigate the effect of political alignment in evaluating crisis response strategies such as apologizing. During a crisis it is important for politicians to keep the support of the voters to get reelected and therefore it is interesting to investigate if political alignment plays a role in the evaluation of apologizing in politics. Do we expect the same effects between people ideologically close to apologizing politicians and people further away? Respondents aligned with the political party of the politician are expected to evaluate the politician, the response of the politician and the political system better than respondents who are not politically aligned with the politician. According to Ramona Bobocel et al. (2010) the social identity theory means that ‘people are motivated to think highly of the groups to which they belong (p. 306)’. It means that apologies will more quickly lead to forgiveness when the sender and the receiver of the apology are aligned. In this thesis it means that a citizen who is aligned with the VVD will evaluate an apology of a VVD politician more positively than an apology of a politician who is not aligned with the citizen. Based on this theory it is hypothesized that political alignment has a moderating effect on the relation between apologizing and the public evaluation regarding the image of the politician, the response of the politician and the political satisfaction of the Dutch democracy.
H1: An apology of a politician in crisis communication has a positive influence on the
H2: An apology of a politician in crisis communication has a positive influence on the
public opinion about the response of the politician moderated by political alignment.
H3: An apology of a politician in crisis communication has a positive influence on
political satisfaction of the Dutch democracy moderated by political alignment.
METHODS
Design
To test the research questions and hypotheses that are formulated in this research, an online experiment embedded in a survey via Qualtrics was send out to the network of the researcher via social media (e.g. Facebook) between May 16th
and May 20th
. Qualtrics is an online tool for creating and distributing surveys. In total 297 people started the survey, but only 237 participants actually finished the questionnaire. The other respondents have been filtered out of the dataset (N=237).
This research used a 2 x 2 factorial design of the independent variables (the apology condition: no apology or apology, versus the political alignment condition: no political alignment versus political alignment) on the public opinion by measuring de differences between the groups. To examine the possible effect of the apology of politicians on the public opinion, an online experiment was distributed. The first manipulation was a politician apologizing or not. The second manipulation was the political party VVD or PvdA. The party condition was not the same as the experimental factor. The experimental factor, political alignment, was made by combining the manipulation of political party and the probability to vote for the VVD or the PvdA.
The participants were randomly assigned to one of the four conditions. The first condition was a PvdA state secretary in crisis without an apology. The second condition was a VVD state secretary in crisis without an apology. The third condition was a PvdA state secretary in crisis with an apology. The fourth condition was a VVD state secretary in crisis with an apology.
First, the participants received a questionnaire with questions about their age, gender, political knowledge, voting preference, probability to vote and political ideology etcetera. Secondly, the participants were randomly assigned to one of the four conditions as mentioned above. The participants read a news article about a VVD or PvdA politician in crisis with or without the apology manipulation. The news article was constructed based on real articles that appeared when the minister Kamp of Economic Affairs apologized in March 2015 and when the state secretary Mansveld for Infrastructure and Environment apologized in January 2013. To make the articles as neutral as possible, no names were given and state secretaries were discussed of the ministry of Infrastructure and Environment and the ministry for the Central Government Sector, relatively unknown ministries. To find two similar ministries who share policy, it was necessary to make the state secretary for the Central Government Sector up. I received no questions or feedback of the respondents. This was done to minimalize primes of specific government members. The only differences between the articles are the two manipulations to get the purest causal effects. In total 50.6% of the respondents received a condition containing an apology and 49.4% received a condition without an apology. Furthermore 50.6% of the respondents received a condition containing a VVD politician and 49.4% received a condition containing a PvdA politician. In Appendix I you can find the four different stimuli.
After the manipulation sections, respondents were asked to evaluate the characteristics of the politician, the response strategy of the politician and the political system in the
Netherlands.
Sample
As mentioned earlier 297 persons started the survey and by filtering out the unfinished surveys, a sample of 237 respondents remained (descriptives in Appendix II). First the participants were asked basic questions about socio-demographics: 51.1% of the respondents was female and the average age of the sample was 33 years old (SD=12.63). The education level of the respondents was quite high: 44.3% of the participants attended University, 27.8% attended HBO and 23.9% of the respondents was currently a student.
Of the 237 respondents, 46.8% would vote for Democraten ’66 (D66) at the elections for the House of Representatives 21.5% would vote for Volkspartij voor de Vrijheid en Democratie (VVD) and 4.6% would vote for the Partij van de Arbeid (PvdA). The latest
opinion polls were held June 4th
2015 by Ipsos, a market research company in the Netherlands and show 29 seats for VVD (19.3%), 14 seats for PvdA (9.3%) and 22 seats for D66 (14.6%) (Ipsos, 2015), so the sample is not representative for the population. Of the sample 43.5% scored higher than 5 on the probability to vote for VVD (M=6.21, SD=3.46) and 18.6% of the sample scored higher than 5 on the probability to vote for PvdA (M=4.60, SD=2.98). As stated, the party favor of the respondents is not completely representative for the population. Therefore probability to vote has been used to measure the political alignment as is explained later in this thesis.There were no significant differences found for characteristics between the groups and randomization succeeded, results are shown in Table 1.
Table 1: Pearson Correlation Coefficients between Variables (N=195), for Political Alignment (N=146). Image competen ce Image sympat hetic Opinion response Political satisfaction democracy Political knowled ge Political cynicism Age Educ ation Manipulatio n Apology Political Alignme nt Image competenc e .49** .63** .22** -.03 -.28** .06 .02 -.03 -.03 Image Sympathet ic .47** .17* .03 -.16* .19* * .04 .20** -.05 Opinion response politician .25** .05 -.28** .06 .10 .16* -.09 Political satisfactio n democrac y .10 -.36** .04 .15* .04 .02 Political knowledge -.13 .19* * .09 .-.08 .07 Political cynicism .10 .19** .03 .10 Age -,14* .08 -.05 Education -.07 -.14 Manipulat ion Apology -.06 Political Alignment
Operalization & Measurement
To test the formulated research questions and hypotheses, variables have been formed and explained below. The descriptive statistics of all the variables can be found in Appendix II.
The independent variable in this research is the Manipulation Apology. To measure the effect of an apology on the dependent variables, a separate independent variable was computed. By selecting all the respondents that received an apology as stimulus and computing a new variable Manipulation Apology (M=.51, SD=.50), it is now possible to see the differences between the two groups: 117 (50,6% ) of the respondents received an article with an apology and 120 (49.4%) received an article without an apology. To make sure that the respondents who thought they received an apology but actually didn’t, a manipulation check was built in the survey. When the manipulation was successful, the participants are taken into account. When the manipulation was unsuccessful, the respondents were left out of the analyses (N=195). Of all the remaining respondents, 95 did not receive an apology (48.7%) and 100 did receive an apology (51.3%).
The other manipulation was the manipulation of the party. Of the total respondents 50,6% received an article with a VVD politician and 49.4% received an article with a PvdA politician. To analyze the moderation, we construct from exposure to these conditions a new variable Political alignment. The respondents were asked to rate the probability to vote for the party VVD and the party PvdA on a 11-point scale from (0) Very unlikely to (10) Very likely. To compute the variable, the people who received the manipulation VVD and rated the VVD higher than 5 are political aligned. Also the people who received the manipulation PvdA and rated the PvdA higher than 5 are political aligned. These two conditions together form the variable Political Alignment (M=.50, SD=.50), 124 respondents (52.3%) were not politically aligned and 113 respondents (47.7%) were politically aligned. To ensure the effect of the manipulation, another manipulation check was built in the survey. To measure the moderating
variable Political Alignment only the respondents were taken into account where the manipulation succeeded (N=146): 77 of the respondents (52.7%) were not politically aligned and 69 of the respondents (47.3%) were politically aligned.
In this research there are three dependent variables. The first dependent variable is Image politician. The image of the politician was measured by asking the respondents to rate the
politician on a 7-point Likert-scale for the following characteristics: sympathetic, competence, confidence, trustworthiness, decisiveness, informed, leadership, integrity and critical. A factor analysis showed that there are two components with an eigenvalue higher than 1 when trying to form one variable, what was really interesting.
A closer look into literature showed that there are more examples of research where two dimensions of characteristics were formed. An example is a research by Stewart and Clarke (1992) where they identified two dimensions in the image of political leaders. They labeled the characteristics such as decisiveness, energy, being tough and well informed as competence. Characteristics such as being caring, a good listener, likeable and trustworthy were labeled as responsiveness. Therefore the decision was made to use two different variables to measure the image of a politician. The first component, Image Competence consists out of competence, confidence, trustworthiness, decisiveness, informed and leadership (α=.78, M=3.70, SD=.85). The second component, Image Sympathetic consists out of sympathetic, integrity and critical (α=.63, M=4.06, SD=.85).
The second dependent variable is Opinion response of politician. The opinion of the participants on how the politician responded to the crisis was measured by asking the participants to rate the response of the politician on a 7-point Likert scale, from very poor to excellent (M= 3.85, SD= 1.34).
The third dependent variable is Satisfaction national democracy. The respondents’ satisfaction of the national democracy was measured by asking the respondents to rate the
statement ‘How satisfied or dissatisfied are you with the way the democracy works in the Netherlands?’ on a 7-point Likert scale, from very dissatisfied to very satisfied (M=4.53, SD=1.36).
To make sure possible effects are pure, a couple control variables could be used in the analyses. Apart from the descriptive variables (e.g Age and Education), two extra variables were computed. The first variable is Political knowledge. To measure the political knowledge of the respondents, five questions about Dutch politics were asked (e.g. Who is the current Minister of Foreign Affairs?). All the questions were made into dummy variables where the correct answer is made into 1 and the not correct answers were made into 0. After that the variable Political knowledge was computed by making a new scale variable where 0 is no correct answers and 5 is 5 correct answers (M= 2.61, SD=1.17).
The second variable is Political cynicism, which was measured by asking the respondents to rate four statements on 5-point Likert scale from (1) is Totally disagree to (5) Totally agree. After that a factor analysis was done to see if the four statements together could form a scale variable, with an eigenvalue greater than 1. After that it became clear that the statements were eligible to form a reliable scale variable together (α=.70, M=2.96, SD=.67).
To determine what control variables should be taken with the different analysis, the correlation between de different variables are shown in Table 1. The correlation table shows that there are no significant correlations between the variables and manipulations, which means that there is no need for control variables in the models.
RESULTS
To answer the central research question of this thesis, all the formed subquestions and hypotheses will be tested in the result section. For the first three research questions, the manipulation check for apology was activated (N=195). For the last three hypotheses, the manipulation check for apology and the manipulation check for political party were activated (N=146). 1
RQ1: Does an apology of a politician in crisis communication have an influence on
the politician’s image?
The first sub question investigates the influence of an apology on the politician’s image. Image competence and Image sympathetic are the dependent variables and the Manipulation
apology is the independent variable. As can be seen in Table 2, ANOVA in SPSS shows that
there is a significant difference between the groups for Image sympathetic (F(1,193)=8.17, p=.005). Post hoc shows that the apology condition scores more positive (M=4.23, SD=.89) than the non apology condition (M=3.88, SD=.82).There is no significant difference between the groups for Image competence (F(1,193)=0.16, p=.692). This means that a partial positive effect was found of apologizing on the image of a government member.
RQ2: Does an apology of a politician in crisis communication have an influence on
the public opinion about the response of the politician?
The second sub question is tested by using the variable Manipulation apology as an independent variable and by using Opinion response of politician as a dependent variable. As can bee seen in Table 2, there is a significant difference (F(1,193)=4.882, p=0,028) between the group that did receive an apology (M=4.11, SD=1.43) and the group that did not receive
an apology (M=3.67, SD= 1.32). This means that a positive effect was found of an apology of a politician on the public opinion about the response of the politician.
Table 2: Results of an analysis of the main effect of Manipulation Apology (N=195)
Independent Variables No Apology (N=95) M SD Apology (N=100) M SD df F p Image Competence 3.71 .84 3.65 .97 1 .16 .692 Image Sympathetic 3.87 .83 4.23 .89 1 8.17 .005 Opinion response Politician 3.67 1.32 4.11 1.42 1 4.88 .028 Political Satisfaction Democracy 4.52 1.34 4.64 1.38 1 .40 .526
RQ3: An apology of a politician in crisis communication has an influence on political
satisfaction of the Dutch democracy.
The third sub question tests the effect of a politician’s apology on the political satisfaction of the Dutch democracy of the respondents. Political satisfaction democracy is the independent variable and the Manipulation apology is the independent variable. In Table 2, it shows that the group with an apology (M=4.64, SD=1.38) did not significantly differ from the group without an apology (M=4.52, SD=1.34). There was no effect of Manipulation apology found on Political satisfaction democracy (F(1,193)=1, p=.526).
H1: An apology of a politician in crisis communication has an influence on the image
of a politician moderated by political alignment.
The first hypothesis tests the effect of a politician’s apology on the image of a politician moderated by Political Alignment. Image competence and Image sympathetic are the dependent variables. The independent variable Manipulation apology and the moderating variable is Political Alignment. As can be seen in Appendix IV, the results show that there is no significant difference for Image Competence (F(1,191)=1, p=.255) and for Image Sympathetic (F(1,191)=1, p=.655). Hypothesis 1 cannot be supported. What is interesting to
notice in these results is that there is no significant effect of political alignment alone on the image of a politician (F(1,191)=1, p=.741) and the effect of an apology on Image Sympathetic is still significant (F(1,191)=1, p=.003) even for a smaller sample (Results are shown in Appendix III).
H2: An apology of a politician in crisis communication has an influence on the public
opinion about the response of the politician moderated by political alignment.
The second hypothesis examines the effect of a Manipulation apology and on the Image of a politician moderated by Political alignment. After executing an analysis (results in Appendix
IV) no significant differences between groups different groups were found (F(1,191)=1,
p=.617) which means that there is no moderating effect from Political alignment. Hypothesis 2 cannot be supported. Also here there is no significant effect of political alignment alone on the opinion about the response of the politician (F(1,191)=1, p=.655) and the effect of an apology on Opinion response politician is still significant (F(1,191)=1, p=.006) even for a smaller sample.
H3: An apology of a politician in crisis communication has an influence on political
satisfaction of the Dutch democracy moderated by political alignment.
The last hypothesis tests the effect between Manipulation Apology and Satisfaction political system moderated by the variable Political alignment. An analysis was executed and the
results are shown in Appendix IV. There was no moderating effect found by Political Alignment (F(1,191)=1, p=.584). The last hypothesis cannot be supported. Just like the other
hypotheses there is no significant effect of Political Alignment alone on the political satisfaction (F(1,191)=1, p=.767).
CONCLUSION & DISCUSSION
Apologizing in crisis communication has been a popular research topic for the last two decades. However, the effect of apologizing of government members on the public opinion has not been researched that much. The goal of this thesis is to provide valuable insight in how apologizing affects the image of politicians, how the response strategy is evaluated and if it affects the respondents’ political satisfaction. The results of this research contribute to knowledge concerning apologizing in the political arena in the Netherlands and offers guidelines to government members to find a way to influence their reputation after political crises. Therefore, the following research question was formulated: ‘What is the effect of government members apologizing on the image of government members, the evaluation of the
response of the government member and the evaluation of the political system?’
To answer this question an experimental research was conducted containing two manipulations: the apology manipulation and the political alignment manipulation. The effect of apologizing as a response strategy on the public opinion was assessed in three ways: the
evaluation of the image of the government member, the evaluation of the apology response of the politician and the evaluation of the political system after a government member’s apology. Also the possibility of an existing moderating effect of political alignment was hypothesized.
First, a significant difference was found between the respondents that did not receive an apology and the respondents that did receive an apology. The first significant effect was found for one dimension of the Image of Politician: the dimension Image Sympathy. The results of this research supported findings by Tucker et al (2006), where they conclude that apologizing has a positive effect on the image of a leader. Government members that apologized, scored significantly higher on the characteristics sympathetic, integrity and critical. This implies that apologizing has a positive effect on rebuilding an image after a political crisis and could offer guidelines for politicians. Interesting to notice the critical part in this dimension. It means that if someone apologizes, they evaluate politicians as being more critical. The positive effect of an apology on the sympathetic dimension of the image of government member suggests that government members can use apology as a crisis response strategy in certain situations. When government members want to enhance their image regarding the characteristics as sympathetic, integrity and being critical during a crisis, apologizing seems like a good crisis response strategy. In the Netherlands it is sometimes the case that government members have to step down during a political crisis and resign from their duties as a cabinet member. Apologizing for mistakes or their actions, can offer them a possibility to rebuild their image and keep the support of their voters.
No significant results have been found for Image Competence. Based on the results it seems that the respondents do not evaluate the competence dimension of the politician’s image more positively or more negatively after an offered apology. The government members were not evaluated differently on the characteristics competence, confidence, trustworthiness, decisiveness, informed and leadership. These outcomes contradict the theory of Tucker et al
(2006) that states that apologies lead to a more positive evaluation of the politician, but they also not support the theory by Sigal et al (1988) and the Reputatiegroep (2012) that apologizing harms the reputation of a politician. It shows that in this specific case apologizing leads to politicians be seen as more sympathetic and does not lead to politicians be seen as less competent. It could be that in other contexts, it is desirable or undesirable to apologize for mistakes. It could also be that a corrective action suffices regarding the competence of a politician.
Also, a significant effect of an apology on the evaluation of the response strategy has been found. This confirms earlier research by Scher & Darley (1996) that showed more positive evaluations of the response when being exposed to an apology. This thesis confirms the theory by Scher & Darley and shows that apologizing also leads to more positive evaluations of respondents regarding the response strategy in the political context. In this thesis respondents evaluated an apology as a more appropriate response than the no apology.
Thirdly, there was no significant effect found of apologizing on the satisfaction of the respondents regarding the political system. The corporate crisis research by Verhoeven et al (2012) which states that crisis have a greater impact on the image of the organization than on the spokesperson of the organization, is not supported in the political context. It could be as suggested earlier in this thesis, that the ministerial responsibility is enough to maintain the reputation of the political system. It suggests that apologizing does not have an effect on respondents’ evaluation of the political satisfaction about the democracy in the Netherlands.
Finally, this study brought new elements to previous research on political crisis communication by adding another dimension to apologizing in politics by investigating the moderation effect of political alignment on the relationship between apologizing and the public opinion. There were no significant results found on the evaluation of the image of the government member, the evaluation of the apology response of the politician and the
evaluation of the political system after a politician’s apology. The results do not support the social identity theory discussed in the research of Bobocel et al (2010) that states that respondents aligned with the politician will evaluate the politician more positively.
Overall, it can be concluded that apologizing has an effect on the public opinion. Apologizing has an effect on the sympathetic dimension of the image of government members and on the evaluation of the response strategy. Interestingly enough, what was not part of the research question, but became clear in the results is that there is no main effect of political alignment on the three independent variables. These results are somewhat surprising and imply that respondents who are politically aligned with the right-wing party VVD do not evaluate a VVD government member more positively than a PvdA government member, and the other way around. A possible explanation for these outcomes could be that the government members, ministers and state secretaries, represent the government as a whole and people do not evaluate the politicians based on the political party they are from. It is possible that respondents see ministers and state secretaries as part of the governmental system and therefore do not treat the government members from both parties differently. In the end, the respondents who were political aligned felt connected with one of the parties and both parties are part of the government now. It can be that citizens who support one of the parties, supports the government as a whole. Another explanation could be that the state secretaries in the manipulated articles were not known by the respondents and therefore could not replicate a feeling of ‘alignment’ with the politician to treat him/her differently than a politician from another political party.
Future research & limitations
As stated by Tucker et al (2006) more research is needed to examine the effect of apologizing in the political context. This thesis is a first attempt to close the theoretical gap in political
crisis communication. It should be mentioned that although the results from this study supports findings from earlier studies, the experimental structure of this thesis has some limitations. The ecological validity of this study is somewhat restricted. Although the respondents were not aware of participating in an experimental research, the political crisis described in the stimulus was artificial. In a natural situation there would be an endless amount of external pressures affecting the public opinion, for example: the characteristics of the politician, the specifics of the crisis, the reputation of the politician, the way the apology was formulated, the media attention and the reaction of the House of Representatives.
Also, due to limited time and resources respondents were recruited to participate in the research via the network of the researcher. Therefore, the sample was not completely representative for the population. The average age of the sample of 237 respondents was 33 years old and the educational level of the participants was quite high: 44.3% of the participants attended University and 27.8% attended HBO. Moreover, 46.8% said to vote for Democraten ’66 if the elections for the House of Representatives would be tomorrow. This
means that if the sample would be more representative, different outcomes could be found. Maybe lower educated respondents would be more affected by emotions than highly educated people. Further research is suggested to examine the effect of apologizing on a more representative sample of the population, due to suggestive positive effects of apologizing.
Another suggestion for future research is to examine the absence of an effect of apologizing on the competence of a politician. This could be explained by the fact that a corrective action suffices as a response strategy. Therefore, further research is needed to explore the use different apology strategies to political crises. Is an explicit apology the best way to offer an apology or is taking corrective action enough?
Another angle is that the results can offer guidelines for government members in crisis without having to resign from their duties. It would also be interesting to see if the public
thinks a government member should resign after a political crisis and if their would be differences between groups with an apology or groups without an apology. What is the effect of apologizing on the public opinion about whether a government member should resign and how would the results relate to theory in corporate crisis communication?
This thesis was a first attempt to close the theoretical gap of political crisis communication regarding apologizing of government members. Although sorry still seems to be the hardest word, in some political crises it is the right word and can help to enhance the image of a politician.
Endnote
1. When the filter for manipulation apology is not executed, the results for research questions 1, 2 and 3 are as follows:
a. Image Competence ((F(1,236)=1, p=.409) Image Sympethatic ((F(1,236)=1, p=.015)
b. Opinion response of politician ((F(1,236)=1, p=.106) c. Political Satisfaction Democracy ((F(1,236)=1, p.556)
When the filter for manipulation apology and political party are not executed, the results for hypotheses 1, 2 and 3 are as follows:
d. Image Competence ((F(1,236)=1, p=.695) Image Sympethatic ((F(1,236)=1, p=.716)
e. Opinion response of politician ((F(1,236)=1, p=.582) f. Political Satisfaction Democracy ((F(1,236)=1, p.285)
LITERATURE
Benoit, W. L. (1997). Image repair discourse and crisis communication. Public Relations Review, 23(2), 177–186.
Benoit, W. L., & Drew, S. (1997). Appropriateness and effectiveness of image repair strategies. Communication Reports, 10, 153–163.
Bobocel, D. R., Kay, A. C., Zanna, M. P., & Olson, J. M. (2010). Psychology of Justice and Legitimacy: The Ontario Symposium Volume 11. East Sussex: Psychology Press.
Bnr (2013) Mansveld bidet excuses aan voor achterhouden rapport. Retrieved from http://www.bnr.nl/nieuws/politiek/753996-1301/mansveld-biedt-excuses-aan-voor- achterhouden-rapport
Bradford, J. L., Garrett, D. E., Garrett, E., & Brennan, E. (2010). Communicative of Corporate Responses to Accusations of Unethical Behavior. Journal of Business, 14, 875–892.
Bradford, J. L., & Garrett, D. E. (1995). The effectiveness of corporate communicative responses to accu- sations of unethical behavior. Journal of Business Ethics, 14(11), 875-892.
Coombs, W. T., & Holladay, S. J. (2008). Comparing apology to equivalent crisis response strategies: Clarifying apology’s role and value in crisis communication. Public Relations Review, 34, 252–257.
Davies, G., Chun, R., da Silva, R. V., & Roper, S. (2003). Corporate reputation and competitiveness. New York: Routledge.
Huang, Y. (2006). Crisis Situations, Communication Strategies, and Media Coverage: A Multicase Study Revisiting the Communicative Response Model. Communication Research 33(3), 180-205.
Kim, P. H., Ferrin, D. L., Cooper, C. D., & Dirks, K. T. (2004). Removing the shadow of suspicion: the effects of apology versus denial for repairing competence- versus integrity-based trust violations. The Journal of Applied Psychology, 89(1), 104–118. Ipsos (2015). Barometer van de week. Retrieved from
http://www.ipsos-nederland.nl/ipsos-politieke-barometer/barometer-van-deze-week
Mills, N.: (2001). ‘The New Culture of Apology’, Dissent (Fall), 48(4), 113–116.
Nobles, M. (2008). The Politics of Official Apologies. Cambridge: Cambridge University Press.
Nos (2013) Mansveld biedt excuses aan. Retrieved from http://nos.nl/artikel/468848- mansveld-biedt-excuses-aan.html
Parlement (2015). Ministeriele verantwoordelijkheid. Retrieved from
http://www.parlement.com/id/vh8lnhrp1wzn/ministeriele_verantwoordelijkheid
Reputatiegroep (2012). Reputatiemanagement, excuses en de sorrylijst van de Reputatiegroep. Retrieved from http://hetreputatieblog.nl/reputatiemanagement-excuses-en-de-sorrylijst-van-de-reputatiegroep/
Schlenker, B. R. (1980) Impression Management: The Self- concept, Social Identity, and Interpersonal Relations. Monterey, CA: Brooks/Cole.
Schmitt, Manfred, et al. "Effects of objective and subjective account components on forgiving." The Journal of social psychology 144.5 (2004): 465-486.
Scher, S. J., & Darley, J. M. (1997). How Effective Are the Things People Say to Apologize? Effects of the Realization of the Apology Speech Act. Journal of Psycholinguistic Research, 26(1), 127–140.
Sigal, J., Hsu, L., Foodim, S. & Betman, J. (1998). Factors Affecting Perceptions of Political Candidates Accused of Sexual and Financial Misconduct. Political Psychology 9(2), 273-280.
Trouw (2015) Kamp biedt Groningers excuses aan. Retrieved from
http://www.trouw.nl/tr/nl/4492/Nederland/article/detail/3878592/2015/03/02/Kamp- biedt-Groningers-excuses-aan.dhtml
Tucker, S., Turner, N., Barling, J., Reid, E. M., & Elving, C. (2006). Apologies and transformational leadership. Journal of Business Ethics, 63, 195–207.
Verhoeven, J. W. M., Van Hoof, J. J., Ter Keurs, H., & Van Vuuren, M. (2012). Effects of apologies and crisis responsibility on corporate and spokesperson reputation. Public Relations Review, 38(3), 501–504.
Van Thijn, E. (1999) Voorbij de ‘De Sorry-Democratie’. Over de strijd tussen politiek en bureaucratie. S&D 4, 167-171.
APPENDIX I: Stimuli used in thesis (Bnr, 2013; Nos, 2013; Trouw, 2015)
APPENDIX II: Descriptive statistics of the variables of sample (N=237) Variable Chronbach’s α M(%) SD Gender (Female) 51.1% Age 33.01 12.63 Education (University) 44.3% Student (Yes) 23.9% Ideological position 6.20 2.09 Voting behaviour (D66) 46.8%
Manipulation apology (Yes) 50,6%
Manipulation party (VVD) 50,6%
Image politician 1 .78 3.70 .85
Image politician 2 .63 4.06 .85
Opinion response of politician 3.85 1.34
Political satisfaction democracy 4.53 1.36
Political knowledge 2.61 1.17
Political cynicism .70 2.96 .67
APPENDIX III: Results of the main effect of Manipulation apology and Political alignment (N=146) No Apology M SD Apology M SD No Political Alignment M SD Political Alignment M SD df F p Manipulation Apology Image Competence 3.70 .84 3.65 .99 1 .00 .963 Image Sympathetic 4.06 .83 3.97 1.03 1 9.01 .003 Opinion response Politician 4.00 1.37 3.75 1.46 1 7.649 .006 Political satisfaction 4.73 1.28 4.77 1.31 1 3.03 .084 Manipulation Political Alignment Image Competence 3.68 .86 3.68 .97 1 .11 .741 Image Sympathetic 3.80 .85 4.25 .95 1 .20 .655 Opinion response Politician 3.57 1.33 4.21 1.43 1 .827 .365 Political satisfaction 4.57 1.29 4.93 1.27 1 .09 .767
APPENDIX IV: Results of the moderation effect of Political alignment and Manipulation apology (N=146) No Apology M No political Alignment SD No Apology M Political Alignment SD Apology M No Political Alignment SD Apology M Political Alignment SD df F p Image Competence 3.79 .85 3.56 .87 3.62 .84 3.74 1.13 1 1.31 .255 Image Sympathetic 3.86 .78 3.73 .93 4.25 .84 4.25 .95 1 .20 .655 Opinion response Politician 3.73 1.35 3.41 1.30 4.25 1.35 4.16 1.55 1 .252 .617 Political satisfaction 4.59 1.19 4.54 1.41 4.85 1.37 5.03 1.15 1 .30 .584
