On the Reputation of Political Parties: Emotional Appeal and Competence in a Cross-National Study
By Chris Levy
Second of two parts
Recently, TNS collected crossnational data on the reputation of political parties in the eyes of the general public, using a method originally developed as a management tool for measuring the reputations of corporations in the private sector. By analyzing these data, we can examine possible influences on the reputation of parties, which may explain the pattern of results uncovered. Does the reputation of parties depend on the performance of the governments that are the product of their competition? Or is it more deeply rooted, depending, for example, on citizens’ wider belief in the values of the democratic system to which parties belong?
In part 1 of “On the Reputation of Political Parties,” we established three hypotheses to explain the nature and dynamics of political party reputation:
- Hypothesis 1: Political parties will have a lower reputation in countries where populations hold values that are more post-materialist in orientation.
- Hypothesis 2: There will be a higher reputation for parties in countries where governments are seen to be performing well.
- Hypothesis 3: The reputation of parties will be highest in countries where there is a strong preference among the citizens for democracy as a system of government.
Combinations of these three hypotheses allow for a range of scenarios explaining party reputation. As Figure 1 illustrates, reputation may be influenced from within the political system, either from below by governmental performance (hypothesis 2, at the left of the figure), or from above by preference for democracy (hypothesis 3, in the middle). Hypotheses 2 and 3 are not mutually exclusive, which means a third possibility—on the right—is that party reputation is influenced from both directions within the system.
We have also hypothesised that party reputation is subject to outside factors—specifically, the cultural changes that have occurred in the different countries as a result of postwar economic growth (hypothesis 1). The three scenarios shown in Figure 1 therefore also have variants, with additional influence coming from this post-materialist culture shift. Yet another possibility is that party reputation may derive from this source, but none of the factors influencing it come from within the political system.
To test these various scenarios, we first identified data sources which best described the variables we wished to include in the analysis—that is, we looked for indicators of respondents’ post-materialist values, satisfaction with government performance, and preference for democracy. We then used a statistical procedure called a linear regression to examine the relationships between these variables on the one hand and the TRI*M index—our measure of party reputation in the various countries—on the other.
Data to test hypotheses 1 and 3 were taken from the World Values Survey (WVS), which covers twenty-eight countries in the TRI*M sample. For data on hypothesis 2 (pertaining to government performance) we turned to the European Social Survey (ESS), which includes fewer countries.
Because two different data sources were used, there were two different sets of country samples. In order to allow for direct comparison between hypotheses, we conducted the statistical analysis first for all available countries with TRI*M scores in either sample, and then again for the core of fifteen European countries represented in both samples (see Table 1).
Hypothesis 1: Political parties have a poorer reputation in countries where post-materialist values are relatively more prevalent.
The indicator for post-materialist values was taken from the World Values Survey. In each country in the WVS respondents were asked the following questions:
If you had to choose, which one of the things on this card would you say is most important?... And which would be the next most important?
(a) Maintaining order in the nation
(b) Giving people more say in government decisions
(c) Fighting rising prices
(d) Protecting freedom of speech
Responses (a) and (c) were designed to represent materialist values relating to physical well-being. Responses (b) and (d) were post-materialist, representing the values pertaining to more than physical survival which, some authors claim, are more prevalent in wealthy societies. We coded respondents to these questions according to whether their answers indicated they were “materialist” (a and c), “post-materialist” (b and d), or “mixed” (any other combination of responses).
Table 2 shows the percentage of post-materialists in each country, according to this classification.
It is clear from Table 2 that post-materialist theorists are correct, in that a higher proportion of citizens holds post-materialist values in the wealthier countries. Eight of the top ten countries are advanced Western industrial democracies, while none of the bottom ten are. These attitudes, however, did not have any relationship with party reputation, as measured by TRI*M. Carrying out the same analysis using the core fifteen countries resulted in a slightly stronger, but still insignificant, relationship (R²=0.09). While it is true that post-materialist values are more prevalent in advanced industrial economies, they do not seem to affect the reputation of political parties in the way the post-materialist theorists claim.
Hypothesis 2: The reputation of political parties is higher in countries where the population is satisfied with the performance of the incumbent government.
The indicator for the independent variable in this hypothesis was taken from the European Social Survey (ESS):
Now thinking about the [country] government, how satisfied are you with the way it is doing its job?
Respondents were asked to answer on an eleven-point scale, where zero equalled extremely dissatisfied, and ten equalled extremely satisfied.
This question provides a very good indicator of satisfaction with government performance, especially because it does not specify performance in a particular area—say, the economy—but leaves respondents to interpret “performance” themselves. This gives a built-in control for differing expectations of the policy areas in which government should be involved, which vary among countries due to cultural and historical factors. Table 3 shows the mean scores for government satisfaction for all the countries in the EES that also have TRI*M scores.
As Table 3 indicates, satisfaction with government performance is low across Europe. Even in Spain, at the top of the list, citizens in 2004 were, on average, more dissatisfied than satisfied with their national government. Citizens in Poland and Portugal were especially unhappy with government performance. While the range of variations in satisfaction (which, in fact, describes only different degrees of dissatisfaction) was fairly small, there was a strong link here with party reputation (R²=0.63). Reducing the sample to the core fifteen countries produced a nearly identical result (R²=0.64).
In considering these results, we must emphasize that this relationship does not allow for the direction of causation—that is, we can’t tell if party reputation is low because respondents are unhappy with the performance of their governments, or if they are dissatisfied with government performance because they don’t think much of their parties—and that there is a lack of control here for external variables. Such a strong relationship, however, is certainly indicative of the conclusion that citizens evaluate their parties as a collective institution largely according to the performance of the governments they produce.
By separating the TRI*M index into the five question batteries on which it is based and treating the data from these questions as a series of conventional poll questions, we can analyze this further, looking, in particular, at the differences in response to the emotional components of the index and those that measure assessments of party competence.
The first point here, the results of our linear regression show, is that there is a strong relationship between government satisfaction and the “master”’ reputation question. (“How would you rate the overall reputation of political parties in [your country]?”)
The strong relationship between government performance and responses to the question, “How would you rate the success achieved by political parties in [your country]?” demonstrates that government performance ratings are a good predictor of party competence evaluations. This is to be expected, as it is logical that if there is one component of the TRI*M index where we would expect to see a particularly good relationship with government performance, it is the competence question; both of these measures relate to practical considerations.
Most interestingly, though, government performance is linked with the emotional aspects of the index, too—favorability of opinion (“How favorable is your opinion on political parties in [your country]?”) and trust (“In the long run, how much do you believe you can trust political parties in [your country]?”), suggesting that the emotional components of the reputation of party institutions—within Europe, at least—are sensitive to short-term evaluations of the level of government performance. This runs contrary to notions of political support that work on the premise that citizen dissatisfaction is mainly directed at particular politicians at particular times, serving to insulate parties and the principles of the regime from negative evaluations.
For example, one argument holds that publics are highly capable of distinguishing among different levels within the political system—such as politicians, governments, institutions, and principles—and they evaluate these separately. The relationship between citizen satisfaction and favorability and trust, as demonstrated here, offers a different interpretation: that these may, to some extent, be a function of governmental performance.
Further investigation is required into the effects of performance over time, however; party reputation is formed in the context of the accumulated judgments and evaluations that result from the continued experience of individuals with a particular political system. For this reason, it is dangerous to come to firm conclusions on the relationship between trust and short-term performance without the longitudinal time-series data required to understand the accumulated effects of performance levels over a number of years.
Hypothesis 3: The reputation of parties is highest in countries where there is a strong preference amongst the citizens for democracy as a system of government
The indicator of preference for democracy was taken from the WVS survey and consisted of the following battery of questions:
I’m going to describe various types of political systems and ask what you think about each as a way of governing this country. For each one, would you say it is a very good, fairly good, fairly bad, or very bad way of governing this country?
…Having a strong leader who does not have to bother with parliament and elections
…Having experts, not government, make decisions according to what they think is best for the country
…Having the army rule
…Having a democratic system
For our analysis, we converted these ratings of democracy to a four-point scale where zero equalled very bad and three equalled very good. Table 4 shows the average rating for each country in the WVS sample.
As can be seen, preference for democracy was high, with average citizen ratings in all countries in the sample at the “fairly good” level at least. In some cases, notably Greece, Denmark, and Sweden, there was almost overwhelming support for democracy as a very good way of governing the country. Differences in support for democracy were thus small in nature, and described only subtle variations in enthusiasm for the system.
When analyzed against the TRI*M index data for party reputation, preference for democracy showed a moderate relationship with levels of party support (R²=0.15). Analysis using the core fifteen countries produced a similar result (R²=0.18). This was not very strong compared to the relationship we have already seen coming from the opposite direction within the political system, but preference for democracy did seem to be a minor explanatory factor.
Citizens in democracies, then (or at least in the democracies analyzed here) have faith in the principles of the regime, and subtle variations in the degree of their enthusiasm do not greatly affect the way in which parties are evaluated. The values citizens hold regarding democracy are relevant to how they evaluate parties, but these judgments are ultimately more performance-based in nature.
The results of the Global Corporate Reputation study show two patterns when examined geographically. The first of these, as shown in last month’s issue of Public Opinion Pros, is that parties in a number of Asian countries are evaluated highly. It is difficult, however, to come to any conclusions on this when three of the seven Asian political systems in the study are semiauthoritarian in nature—we need to ask, do the high reputations of parties in Singapore, Hong Kong, and Malaysia reflect the fact that surveys were conducted in environments where the expression of political criticism is discouraged? Or are they, in fact, a result of the strong economic performance which, the leaders of these countries claim, justifies the curtailment of political freedoms? We must also consider Japan, where party reputation at the time of the study was considerably lower than elsewhere.
We can be clearer about the second pattern concerning variation among European countries: Scandinavian publics are more positive towards their parties than those in Western Europe, while parties are held in particularly low regard in Southern and, especially, Eastern Europe.
One matter that is not ambiguous is the lack of a relationship between post-materialist values and party reputation. It is hard to dispute that these values become more widespread the more wealth a country generates, but the evidence here shows they have not had the effect on political parties that post-materialists suggest—we did not find that parties had a lower reputation in countries with higher numbers of post-materialists.
The clearest findings relate to the influence on party reputation from specific objects (politicians and governments) and diffuse objects (principles of government) within the political system. In Figure 1 three different models were developed: performance-driven, values-driven, and dualistic. The evidence reviewed here strongly supports the first of these models—we saw a very strong relationship between party reputation and evaluations of current government performance. Furthermore, analysis of the individual TRI*M component questions shows that short-term performance evaluations have a relationship with the emotional, as well as the competence, elements of the party reputation equation. Trust and favorability toward parties appear in this light as changeable rather than deeply rooted. There is also a relationship, albeit a very moderate one, between preference for democracy and party reputation. While the results primarily support the performance-driven model, there are still dualistic elements to explaining reputation from within the system.
With this in mind, we can say that government performance and evaluations of parties collectively are related, but in a way that cannot be ascertained here. Although there may be some moderate relationship between preference for democracy as a way of governing and party reputation, they exist on planes that are largely unrelated to one another. We saw earlier that WVS data show democracy is positively viewed in all countries, including in those where its introduction has been more recent. Slight variations in how enthusiastic populations are about democracy are not enough to predict the differing reputations of political parties.
It appears, than, that democracy is in rude health where it exists, and that those living in these conditions are, by and large, satisfied that it offers the best system of governing. While democracy is well-accepted, though, citizens do not take a universally uncritical view of its parties—views which are largely determined by the performance of the governments produced by party competition.
Chris Levy is research executive, TNS Social. The author would like to thank Bill Blyth, Dave Addis, and Cynthia Pinto at TNS for their help, along with Dr. Steffen Hermann and Anne Uekermann at the Global TRI*M Centre in Munich. Data from the GCR study are available for download at http://www.tns-global.com. Click on the "Business Solutions" link, followed by the TRI*M link, and finally the Global Corporate Reputation button.
|
TRI*M and the Global Corporate Reputation Study
The data used here to measure the reputation of political parties are taken from the 2004 TNS Global Corporate Reputation (GCR) study, which uses a commercially developed approach to stakeholder assessments of reputation known as TRI*M.
The TRI*M methodology was developed by TNS in the 1990s as a stakeholder management tool for the private sector. Originally, the main application was in a customer satisfaction environment for both services and consumer goods. This initial focus on quality of product and service was later broadened to cover the general nature of firms’ corporate reputation, only part of which relates to what they produce.
Corporate Reputation TRI*M works on the premise that reputation, as evaluated by individuals, derives from two main sources: competence and emotional appeal. Competence is judged by product and service quality and the financial success of the firm. Emotional appeal consists of the general favorability (or unfavorability) of individuals’ opinions toward the firm, and the extent to which they believe the firm can be trusted in the long run. TRI*M studies measure reputation by asking respondents to rate the company on these four constituent parts, in addition to a general rating of reputation. An algorithm is then used to convert data from these five answers into an overall index score which expresses overall reputation. This index score can then used as a benchmark for evaluation, both against other firms’ reputation and to monitor changes in reputation over time.
TRI*M Global Corporate Reputation Survey
[REPUTATION COMPONENT]
Questions: How would you rate the overall reputation of … excellent, very good, good, fair, or poor?
Accept one response for each item
1.1 … [your country] as a business location
1.2 … the health service in [your country]
1.3 … the supermarket you mostly use
1.4 … your primary bank / provider of financial
services
1.5 … your main telephone provider
1.6 … political parties in [your country]
1.7 … the mail service in [your country]
[FAVORABILITY OF OPINION COMPONENT]
Questions: How favorable is your opinion on … extremely favorable, very favorable, fairly favorable, not very favorable, or not favorable at all?
Accept one response for each item
2.1 … [your country] as a business location
2.2 … the health service
in [your country]
2.3 … the supermarket you mostly use
2.4 … your primary bank / provider of
financial services
2.5 … your main telephone provider
2.6 … political parties in [your
country]
2.7 … the mail service in [your country]
[TRUST COMPONENT]
Questions: In the long run, how much do you believe you can trust …definitely, probably, fairly likely, probably not, or definitely not?
Accept one response for each item
3.1 … [your country] as a business location
3.2 … the health service in [your country]
3.3 … the supermarket you mostly use
3.4 … your primary bank / provider
of financial services
3.5 … your main telephone provider
3.6 … political parties in [your country]
3.7 … the mail service in [your country]
[SUCCESS COMPONENT]
Questions: How would you rate the success achieved by … excellent, very good, good, fair, or poor?
Accept one response for each item
4.1 … [your country] as a business location
4.2 … the health service in [your country]
4.3 … the supermarket you mostly use
4.4 … your primary bank / provider of financial services
4.5 … your main telephone provider
4.6 … political parties in [your country]
4.7 … the mail service in [your country]
[QUALITY COMPONENT]
Questions: How would you rate the quality of product and service of … excellent, very good, good, fair, or poor?
Accept one response for each item
5.1 … [your country] as a business location
5.2 … the health service in [your country]
5.3 … the supermarket you mostly use
5.4 … your primary bank / provider of financial services
5.5 … your main telephone provider
5.6 … political parties in [your country]
5.7 … the mail service in [your country]
|
Method of Analysis
For each hypothesis, the method of analysis was the same. After identifying the data source which best described the variable in question, a linear regression (R²) was calculated, using the indicator in question as the explanatory variable and the overall TRI*M score as the response variable. A high R-squared, naturally, would support the particular hypothesis. In the one case where such a strong relationship was evident, further analysis was carried out by examining the relationship of the independent variable to data for each of the five constituent questions of the TRI*M Index.
This served two purposes: First, it allowed for a more detailed exploration of the confirmed hypothesis. The key finding was that there is a very strong relationship between the TRI*M index and evaluations of government performance. Disaggregating the index allowed us to see whether this held equally as true for the emotional-attachment elements of party reputation (as measured by the favorability and trust questions) as for the performance elements.
Secondly, because commercial reasons prohibit transparency about the TRI*M algorithm, presenting the raw data for index components served as a safety-check for the viability of the relationship between independent variable and dependent index score.
Data Sources for the Analysis
Relevant external data were not available for all countries in the TRI*M sample. The most extensive was the World Values Survey, which covers twenty-eight of the countries in the sample. Unfortunately, the comprehensive nature of the geographical coverage of the WVS is not matched by its regularity. Since 1981, four waves of the study have been carried out at intervals of four to five years. Data for the most recent wave in 2005 are not yet available, so we were forced to fall back on data from 1999/2000, which introduced problems of time-scale with the TRI*M data collected a full four years later.
The major alternative to the WVS was the European Social Survey. This had the advantage of being conducted biannually, which meant we could use 2004 data that accurately described attitudes at the time of the GCR survey. Against this had to be set the limitations of geographical coverage that rendered the extra-European TRI*M data of little use.
The choice between these two data sources was therefore a question of balancing sample size and geographical coverage against the need for controlling for time differences by using data from 2004. The resolution of these conflicting imperatives varies depended on the explanatory variable in question. Specifically, as argued by Mishler and Rose, deep-rooted beliefs and values are much slower to change than evaluations of government performance.
For this reason, data on hypotheses 1 (post-materialist values) and 3 (preference for democracy) were taken from the WVS on the assumption that figures would not have changed to a great extent since 2000. Nonetheless, it would be a worthwhile exercise to replicate the analysis when the 2005 WVS data become available. For data on hypothesis 2 (government performance), ESS data were selected on the basis of the fact that government performance—and incumbent governments themselves—can change a great deal in a four-year period.
Limits to the Method of Analysis
Linear regressions are excellent at explaining the strength of a relationship between two variables, but they cannot explain causation. For example, if hypothesis 3 (preference for democracy) holds true, then is it the case that higher party reputation results from a greater preference for democracy, or do we conclude that citizens will prefer democracy if they hold a high opinion of their parties?
We must also remember that analyses of the type employed here hold problems relating to the influence of external variables. A typical problem encountered in the social sciences is the inability of the investigator to create the ceteris paribus conditions necessary to establish empirical relationships between variables with complete certainty. This lack of physical control over the experimental group is sometimes compensated for by attempting to impose statistical controls accounting for all variables that are suspected of influencing the dependent variable. That this has not been attempted here is due to the large number of potential intruding influences—in testing each individual hypothesis, we would need to control for the other two hypotheses, for demographic factors, and for other dynamics not covered by the hypotheses which could conceivably exert some influence.
For example, the length of time a country has been democratic may also be an issue: specialists on transitional regimes argue that political authorities may experience problems related to the change from a planned to a market-based economy, and the cognitive availability of a recent alternative regime for comparison. Huntington has noted that this may lead to problems of "authoritarian nostalgia" regardless of the performance of the new regime, especially if the previous dictatorship was not especially harsh, corrupt, or incompetent.
Additional Reading
Arter, D. 1999. Scandinavian politics today. Manchester: Manchester University Press.
Bell, D. 1960. The end of ideology: On the exhaustion of political ideas in the fifties. Glencoe: Free Press.
Biorco, R., and R. Mannheimer. 1995. Citizens and political parties. In Citizens and the state, ed. H. Klingemann and D. Fuchs. Oxford: Oxford University Press.
Dalton, R. 1995. Political support in advanced industrial democracies. In Critical citizens: Global support for democratic government, ed. P. Norris. Oxford: Oxford University Press.
———. 1996. Citizen politics: Public opinion and political parties in advanced Western democracies. New Jersey: Chatham House.
Easton, D. 1979. A framework for political analysis. Chicago: University of Chicago Press.
Huntington, S. 1991. The third wave: Democratization in the late twentieth century. Norman: University of Oaklahoma.
Inglehart, R. 1981. Post-materialism in an environment of insecurity. American Political Science Review 75:4.
Inglehart, R. 1997. Post-materialist values and the erosion of institutional authority. In Why people don’t trust government, ed. J. Nye, P. Zelikow, and D. King. Cambridge, Mass.: Harvard University Press.
Klingemann, H. 1999. Mapping political support in the 1990s. In Critical citizens: Global support for democratic government, ed. P. Norris. Oxford: Oxford University Press.
Klingemann, H., and D. Fuchs. 1995. Citizens and the state: A relationship transformed. In Citizens and the state, ed. H. Klingemann and D. Fuchs. Oxford: Oxford University Press.
Krupavicius, A. 1998. The post-communist transition and institutionalisation of Lithuania’s parties. In Parties and democracy, ed. R. Hofferbert. Oxford: Blackwell..
Lijphart, A. 1971. Comparative politics and the comparative method. American Political Science Review 65:3.
Lipset, S. 1963. Political man. London: Mercury.
Lockerbie, B. 1993. Economic dissatisfaction and political alienation in Western Europe. European Journal of Political Research 23.
Mair, P. 2002. Mass electoral behaviour in Western Europe. In Comparative democratic politics: A guide to contemporary theory and research, ed. H. Keman. London: Sage.
Mill, J. S. 1861. Considerations on representative government. London: Routledge.
Mishler, W., and R. Rose. 1999. Five years after the fall: Trajectories of support for democracy in post-communist Europe. In Critical citizens: Global support for democratic government, ed. P. Norris. Oxford: Oxford University Press.
———. 2001. What are the origins of political trust? Testing institutional and cultural theories in post-communist societies. Comparative Political Studies 34:1.
Norris, P. 1999. The growth of critical citizens. In Critical citizens: Global support for democratic government, ed. P. Norris. Oxford: Oxford University Press.
Sartori, G. 2005. Parties and party systems: A framework for analysis. Essex: ECPR Press.
|
|
|
|
|
|
|
|
