Abstract
Representative democracy entails governments that are both responsive and responsible. Mair argued that political parties find it increasingly difficult to balance these two tasks. With an increase in international commitments and interdependence, governments cannot always follow the wishes of their citizens but need to be responsible instead. Our study examines the responsiveness–responsibility link from the angle of citizen perceptions. We argue that when governments are seen as responsive they build a “buffer” of support, allowing them to make decisions that are not necessarily responsive but possibly responsible. By being responsive, governments build a reservoir of goodwill, which they can use to survive more difficult periods. Using data from the 2012 European Social Survey, we test whether perceived responsiveness feeds into this reservoir and whether this reservoir consequently increases perceived responsibility. We find support for this link, suggesting that responsiveness and responsibility do not need to be trade-offs but can complement each other.
Mair (2009) highlighted that representative democracy entails governments that are both responsive and responsible. He argued that political parties—and especially those who generally form the government—find it increasingly difficult to balance these two tasks. With an increase in international commitments and economic interdependence, governments often claim that they have to act responsibly and in their effort to provide responsible governance they cannot always follow the wishes of its voters. Responsive and responsible decisions may of course overlap, but the moments where they do not require governments to make a choice between what people want and what is the responsible thing to do.
We contribute to the debate about the tension between responsiveness and responsibility by empirically assessing the relationship between responsiveness and the leeway that governments have to take responsible, and sometimes unpopular, decisions. Responsiveness and responsibility are complex concepts and may be defined and operationalized in different ways. Whereas responsiveness could be taken to mean the short-term match between what people want and what they receive from political parties and leaders in terms of policies and policy output (see e.g. Kang and Powell, 2010; Stimson et al., 1995), responsibility appears to invoke a more comprehensive interpretation of long-term interests. Sometimes, it is necessary for governing political parties and leaders to take into account long-term needs of the people and the country that have not explicitly been articulated as specific demands from the electorate. In addition, the room for action of contemporary governments is often circumscribed by international and supranational agreements as well as by decisions made by former governments (Bardi et al., 2014: 237).
The theory we propose argues that when governments are perceived by citizens to act responsive to their demands and wishes, they grow a “buffer” of support. This buffer, or “reservoir of goodwill,” may eventually increase the scope for governments to make decisions that citizens perceive as not being in accordance with their short-term interests, that is, decisions that are non-responsive but rather responsible. The rationale behind our theory is a development of Easton’s (1965) notion of a reservoir of goodwill—consisting of public support that is generated by good performance and responsive policy-making—which governments can utilize in order to act responsible. While Easton theorized that the reservoir could be used in times of crisis and/or poor performance, we argue that such a buffer also constitutes an important asset to governments when they need to act responsible rather than responsive, which is frequently the case in this era of globalization and international interdependence.
Using survey data from the sixth round of the European Social Survey (wave 6 in 2012), we first test whether perceptions of responsiveness are associated with higher levels of political support, that is, contributing to a reservoir of goodwill. In a second step, we assess to what extent this reservoir of goodwill creates leeway for governments to act responsible in terms of citizens recognizing responsible government actions. We find that when people perceive the government as more responsive, they also tend to express higher levels of political support. In turn, those who contribute to this reservoir of goodwill also perceive the government as more responsible. Our analyses thus suggest that responsiveness and responsibility are not necessarily conflicting democratic values and that political support acts as an important mechanism in this relationship. This implies that governments should not underestimate the importance of acting responsive since responsive actions are rewarded with support, which in turn increases citizens’ acceptance of political decisions that are not in accordance with short-term demands of the electorate.
Theoretical framework: Perceptions of responsiveness and responsible government
Our main argument is that responsiveness is essential for responsible government and that political support is an important mechanism in this relationship. As recently argued by Esaiasson and Wlezien (2017), the actions of parties and political representatives may be more or less responsive to the demands from citizens. These actions are then judged by citizens who form their own perceptions about the responsiveness of the political system. Citizens react according to these beliefs in terms of voting in elections and expressing support/discontent and trust/distrust toward the functioning of the political system. The reactions of citizens are then supposed to feed back to the political representatives and should preferably affect their future actions. Thus, in this way, governments that act responsive are rewarded with support and trust on behalf of the public. In the following, we outline the process in which responsive government generates a reservoir of goodwill built up by support, and how this reservoir facilitates difficult and sometimes unresponsive decision-making and policy implementation.
Responsiveness and the reservoir of goodwill
As one key principle of democracy, by its very nature, governments should at least be minimally responsive to what a majority of people want (Dahl, 1998; Powell, 2004; Stimson et al., 1995). Democratic government is supposed to mean government by and for the people, and when government policies are seen to reflect people’s ideas and preferences, those governments may indeed be evaluated positively. Thus, people are more likely to express support for the political system when responsiveness is high (Armingeon and Guthmann, 2014; Torcal, 2014). In fact, different aspects of political support, such as citizens’ satisfaction with democracy, have actually been used as explicit indicators of responsiveness (e.g. Morlino and Quaranta, 2014), assuming that they are part of the same concept. However, we argue that responsiveness and support are conceptually different, where the former constitutes an important determinant of the latter. Citizens are simply more likely to evaluate their political system positively when it produces policies that match their preferences. 1
A government that manages to act in accordance with the preferences of a majority of the citizens should thus be perceived as worthy of support and could therefore be perceived as enjoying a high degree of legitimacy. This notion of legitimacy draws on Easton’s (1965, 1975) idea of political support. Easton argues that support is multidimensional and directed toward different objects—from support toward specific political actors and institutions (specific support) to a more diffuse type of support for the political community and the political regime as such (diffuse support) (cf. Norris, 1999, 2011). Easton defines diffuse support as an “attachment to a political object for its own sake” that “taps deep political sentiments and is not easily depleted through disappointment with outputs” (1965: 274). Diffuse support refers to public evaluations of what an object is or represents, not what it does. Specific support, on the other hand, is mainly generated by citizens’ evaluations of the performance and outputs of the political system. It is thus based on public perceptions of policy performance and responsiveness and may thus be sensitive to short-time variations in system outputs (Linde and Ekman, 2003; Norris, 1999; Weatherford, 1987).
In line with Easton, we argue that the two types of support are closely connected and that specific support is a necessary condition for diffuse support. Performance is a very important factor in all types of political systems, regardless of the type of regime. Good performance is recognized by citizens in terms of support for the actions and policies of the government. Specific support is then translated into diffuse support for the underlying principles of the political system. The latter type of support could be regarded as “a reservoir of good will that helps members to accept or tolerate outputs to which they are opposed” (Easton, 1965: 273).
Logically, this process may also proceed in the other direction, that is, that long periods of poor performance—in particular a low level of perceived responsiveness on behalf of the citizens—result in political discontent. In the end, this may negatively affect diffuse support for the political system as a whole, and therefore in the end jeopardize the legitimacy of the political regime. 2 If such a reservoir of goodwill facilitates the process of making decisions that are perceived as unfavorable by citizens, as argued by Easton, we argue that it should also be of importance with regard to the tension between responsive and responsible decision-making highlighted by Mair (2009).
An important question that arises from our line of thinking concerns the formation of the reservoir of goodwill in more general terms. Could it be that the kind of support we are interested in is generated by socialization rather than personal experiences with the system, such as the perceived level of responsiveness (or other experiential factors)? Easton (1975) argues that specific support is generated through two main processes connected to individual experiences. The first is positive evaluations of the extent to which the political system meets the demands of citizens. This notion comes very close to our argument of responsiveness being an essential determinant of support. Second, as we have discussed, “positive or negative evaluations may also be stimulated not by explicit actions on the part of the authorities but by their perceived general performance” (1975: 438). When it comes to diffuse support, Easton is often credited as to regard it as mainly determined by childhood and continuing adult socialization. It is, however, important to note that Easton explicitly views direct experience also as an important source of diffuse support, since long-term negative perceptions of the performance of the regime may negatively affect people’s views of the legitimacy of the regime and its institutions (1975: 445–446).
Taking stock of the results from the last decades of empirical research, the evidence points toward a primacy of experiential processes in generating political support and trust, both specific and diffuse (Dahlberg and Linde, 2016; Dinesen and Sønderskov, 2018; Linde, 2012; Magalhães, 2014; Mishler and Rose, 2001). Thus, in line with this recent body of research, we expect perceptions of responsiveness—which is admittedly an experiential factor—to be significantly related to our measures of the reservoir of goodwill.
Government responsiveness and responsible government
All governments have to maintain a balance between citizens’ demands for policy responsiveness and the need for responsibility, or, in other words, between democracy and effectiveness (Mair, 2009: 10). Our theory stipulates that—in the first step—decision-making perceived as responsive by citizens adds to the reservoir of diffuse support. In the second step, governments may use this reservoir when they need to make responsible decisions. Such a buffer of support facilitates responsible decision-making since the decisions are more likely to be accepted by the citizens, even though these decisions may not be in accordance with the general short-term policy preferences of the electorate or may even be outright unpopular. This argument aligns to some extent with that of Immergut and Abou-Chadi (2014), who argue that electorally vulnerable politicians are less likely to support policies that seem unpopular but perhaps necessary.
Responsibility is somewhat tricky to define in a straightforward manner. Our understanding of the concept draws on the definition provided by Bardi et al. (2014) who define responsibility as: the necessity for /…/ parties and leaders to take into account (a) the long-term needs of their people and countries, which have not necessarily been articulated as specific demands and which underlie and go beyond the short-term demands of those same people /…/; (b) the claims of audiences other than the national electoral audience, including the international markets that ensure their financial alimentation, the international commitments and organisations that are the root of their international credibility, and, in the European context in particular, the heavy transnational conditions of constraint that are the result of a common currency and common market. (Bardi et al., 2014: 237)
Empirical expectations
To sum up, our theoretical argument is the following. Governments acting in ways that are perceived as responsive by citizens are rewarded with support. This performance-based support fills up a reservoir of goodwill, which may be used in times when governments need to act responsibly rather than responsive. The reservoir of goodwill then creates leeway for responsible decision-making because citizens who are content with the performance of the political system are more likely to accept that the government sometimes has to make decisions based on long-term responsibility rather than short-term responsiveness. In the empirical section, we test this argument in a two-step analysis. First, we expect to find a positive relationship between perceived responsiveness and political support (the reservoir of goodwill). Second, we expect a positive relationship between political support and perceptions of responsible government.
These hypothesized relationships should not be interpreted in a strictly causal manner. First, our analyses rely mostly on cross-sectional data. Second, the central concepts—responsiveness, political support, and responsibility—are arguably conceptually related and to some extent also overlapping. Rather, we believe that responsiveness and responsibility are mutually reinforcing phenomena. A responsive government, which enjoys high levels of support, is more likely to be able to make responsible decisions that in the end are beneficial to the long-term interests of the country and its citizens. When citizens recognize this, the government is rewarded with support, and the government is perceived by citizens as actually being responsive since it produces policies that in the end actually benefit the citizens. Thus, responsiveness and responsibility are mutually reinforcing.
Data and methods
In order to find out whether government responsiveness feeds into a reservoir of goodwill, and whether this reservoir consequently helps governments in being more responsible, we mainly use data from the ESS 2012. This wave of the ESS is specifically relevant for our purposes as it includes a special battery of democracy related questions, asking respondents about their ideas and experiences with their national democracy. The wave includes questions about what respondents find important aspects of democracy (e.g. elections, a free media, courts, etc.), whether they feel that aspects of democracy function well in their own country (e.g. whether elections are free and fair or whether minorities are treated equally), and what type of institutional set-up would be best for democracy more generally (e.g. whether governments should be formed by one or more political parties).
The ESS includes individual-level data for several European countries. We further supplement the database with a number of country-level variables that—except for two (ideological congruence and aggregate political support)—are used in additional models that are presented in the Online Appendix. Ideological congruence and aggregate political support are further discussed below. We include all democracies for which data are available, resulting in a total of 25 countries: Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Great Britain, Hungary, Ireland, Iceland, Italy, Israel, Lithuania, the Netherlands, Norway, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, and Switzerland.
Our theoretical argument is sequential, including two steps. This results in two dependent variables: one that measures the reservoir of goodwill in the first step and subsequently a measure of perceptions of responsibility in the second step. To test our argument, we would ideally use indicators of the three concepts (perceived responsiveness, political support, and responsibility) measured at three different points in time (responsiveness at t−2, the reservoir of goodwill t−1, and responsibility at t). However, data on responsiveness and responsibility are rare, which leaves us with no other choice than taking advantage of the available survey items from the ESS 2012. While we are not able to solve this problem, we still aim to address this issue. For one, we include an aggregate measure of the reservoir constructed from earlier waves of the ESS. Moreover, we also try to confront the sequential problem using ideological congruence at the country level. Although this variable does not solve the problem, it may be used as an important robustness check, since it is (a) disconnected from citizens’ perceived responsiveness and (b) is measured before the 2012 wave of the ESS. In the following, the particular measures we use are described and discussed in more detail.
Dependent variables
The reservoir of goodwill
Since our argument stipulates that specific support and political trust contributes to a buffer of goodwill that facilitates decision-making that is at odds with the general demands of citizens, we need empirical indicators of political support. In accordance with a large body of survey research on system support, political support is here measured by the question asking whether or not the respondent is satisfied with the way democracy works. In addition, we use the corresponding question about satisfaction with the national government. In the ESS, the questions are formulated in the following way: “On the whole, how satisfied are you with the way democracy works in [in country]?” and “Now thinking about the [country] government, how satisfied are you with the way it is doing its job?” Respondents then choose an answer between 0 (extremely dissatisfied) and 10 (extremely satisfied).
Although its meaning and measure have been debated in the literature (see Canache et al., 2001; Linde and Ekman, 2003), we follow Linde and Ekman (2003) and Peffley and Rohrschneider (2014) who argue that the satisfaction with democracy question should be used as an indicator of public evaluations of the actual performance of the democratic political system. In their conceptualization of democratic evaluations, Peffley and Rohrschneider (2014; 185) regard both satisfaction with democracy and the evaluation of the government as two types of support that are closely related to what they call the “democratic reality,” as opposed to “constitutional ideal,” which has more to do with the public’s more abstract evaluations of democracy as a political regime, that is, diffuse support in Easton’s sense.
The two indicators—satisfaction with the way democracy works and satisfaction with the way the national government works—together form our first dependent variable. They are combined into an index ranging from 0 to 10, with each indicator weighing equally (Cronbach’s α = 0.79). Figure 1 illustrates the variation between countries for the two indicators separately, as well as that for the combined index of the reservoir of goodwill, which ranges from 3.0 in Bulgaria to 7.0 in Switzerland. When looking at the distribution, a clear pattern emerges, where we find high levels of support in the Nordic and North European countries and substantively lower levels in Southern Europe and in the post-communist East European democracies.
Mean levels of the indicators and index of the reservoir of goodwill.
Countries that were hit hard by the financial crisis, such as Portugal, Spain, Cyprus, and Slovenia, demonstrate fundamentally lower levels of support that countries that were not that severely affected. This corresponds neatly with earlier research on the impact of the crisis on public support for democracy in the wake of the crisis (Armingeon and Guthmann, 2014; Torcal, 2014). Overall, Figure 1 demonstrates substantial variation among European democracies when it comes to the reservoir of goodwill and its individual indicators.
In addition to this indicator of support, which is based on the same ESS 2012 wave as the other variables in the model, we include an aggregate measure of the reservoir of goodwill. This measure uses the same indicators—satisfaction with democracy and satisfaction with the government—but from earlier ESS waves. We include the percentage of support for each country, averaged out over time, to generate an indication of the size of the reservoir before the 2012 measurement of the ESS. Including this measure also aligns with our theoretical expectation that a reservoir builds up over time and then feeds into a perception that the government is responsible.
Perceptions of responsibility
Our second dependent variable is the perception of responsibility. As discussed earlier, the concept of responsibility is somewhat complex. Using the definition suggested by Bardi et al. (2014), responsibility refers to (a) long-term interest of the people and county that have not been articulated as specific demands and (b) claims of non-electorate audiences, such as international agreements and conventions, supranational institutions, and commitments made by previous governments. These are of course concepts that are intricate to operationalize using survey data and ideally we would like to have an “objective” measure on the country-level. So far, however, we have not been able to find any such indicators that could be used as a proxy for responsible decision-making. We are thus left with the 2012 wave of the ESS, which in fact includes a few indicators that come close to Bardi et al.’s two elements of responsibility, in terms of citizens’ perceptions. Two questions, each ranging from 0 (does not apply at all) to 10 (applies completely), are particularly relevant here: (…) [P]lease tell me to what extent you think each of these statements applies in [country].
The rights of minority groups in [country] are protected. Politicians in [country] take into account the views of other European governments before making decisions.
These two indicators are used to create an index of perceived responsibility, ranging from 0 to 10, where both items weigh equally (Cronbach’s α = 0.41). Figure 2 shows the mean levels of the perception of responsibility and its indicators, by country. The index of responsibility ranges between 4.7 (Italy) and 7.0 (Sweden, Denmark, and Finland), suggesting a relatively high mean score considering that the mid-point of the index is 5. Especially for the indicator tapping the assessment of the protection of the rights of minorities is generally high: Only Portugal, Italy, and Spain have average scores of below 5, while many countries have scores of over 6.
Mean levels of the indicators and index of perceptions of responsibility.
In general, the pattern found here is similar to the one demonstrated in Figure 1, with higher levels in Northern Europe and lower levels in Southern and Eastern Europe, although the variation is less pronounced.
The indicators that we include aim to tap perceptions of responsibility as we outline above. However, people’s perceptions might also be closely related to what they think is desirable. If this were the case, it would be difficult to disentangle the relationships between support and responsibility as they could both be seen as evaluations of good performance of the government. To test whether the perceptions of responsibility are connected to people’s ideas about how things should be ideally, we correlate our perception-of-responsibility variables with the importance-of-responsibility variables that are also included in the ESS 2012. These latter indicators refer to the same phenomena (minority rights, other governments) but ask how important these things are in a democracy. The correlations are surprisingly low, ranging from 0.01 (minority rights) to 0.13 (combined responsibility indices). Thus, the issues included in our measurement of responsibility are not necessarily seen as normatively good by those observing them and vice versa.
Independent variable
Our main independent variable, which we use in the first part of the analysis, is the perception of responsiveness. The ESS includes a question that asks whether people think it is best that the government changes its policies according to what most people think or that the government should stick to its planned policies. Depending on their answer, people get one of the following questions, each ranging from 0 (never) to 10 (always): [P]lease tell me how often you think the government in [country] today changes its planned policies in response to what most people think? [P]lease tell me how often you think the government in [country] today sticks to its planned policies regardless of what most people think?
The first of these questions indicates a person’s perception of responsiveness, while the second question indicates the opposite. We thus combined these two questions by reversing the scale of the second question, making an index that ranges from 0 (no responsiveness) to 10 (mostly responsive). This simple measure actually comes quite close to one of the suggested measurements of responsiveness discussed by Powell in his influential article on the chain of responsiveness (Powell, 2004: 102). Figure 3 shows the variation between countries of the mean scores on the perception of responsiveness. The overall average perception of responsiveness is relatively low, with most countries having an average score of below the midpoint of 5. The main exception is Switzerland, which has a mean score of 6.20. Countries on the other end of the scale include Spain, Estonia, and the Czech Republic, which all have score of just over 3.
Mean levels of perceptions of responsiveness.
It is important to notice that we are here gauging the way citizens evaluate the responsiveness of the government and thus not how important they view responsive government. As shown by Bowler (2016), also using data from the ESS, a strong majority (about 65%) of Europeans supports the idea that the government should change its planned policies in response to what most people think. There is considerably less support for the “responsible,” or “trustee,” model of government, which is favored by roughly 18% of the respondents. Our theoretical argument is, however, based on an evaluative notion of responsiveness, that is, how responsive citizens view government to be rather than how responsive government ought to be.
As an additional test, we include a measure of ideological congruence instead of the individual perception of responsiveness. To be sure, congruence and responsiveness, or perceptions of responsiveness, are different concepts as congruence mostly captures the overlap in ideology between representatives and citizens. However, ideological congruence also captures one aspect of how representation works and may be seen as a measure that is closely related to how parliaments or governments respond to people. Since such a measure is an aggregate one, and measured at previous points in time, it will help us to assess whether support and perceptions of responsiveness simply co-vary because they both indicate an evaluation of performance or whether it might be the actual performance of representation that causes—in part—support.
We focus here on what Golder and Stramski (2010) call “many-to-many congruence,” that is, the degree of overlap between citizens’ and representatives ideological preferences. This is a particularly relevant indicator when interested in substantive representation of interests (Golder and Stramski, 2010). It is further relevant as it does not compare an “averaged-out” position of the government (which are often coalitions in Europe) but looks at how preferences are structured in society and consequently how this is related to the structure of preferences in the parliament. We use the calculations and data from Golder and Stramski (2010), which is based on the comparative study of electoral systems. They provide data for 17 of our countries, measured between 1996 and 2005.
Control variables
In addition to the standard socio-demographic variables gender, age, and education, we include a number of variables that have been shown to be important in explaining political support. Many studies have demonstrated that status as electoral winner or loser is one of the most powerful determinants of political support. People who voted for a party that ended up in government tend to display higher levels of satisfaction with the way democracy works and, not very surprisingly, also more satisfied with the performance of the government (Anderson et al., 2005; Esaiasson, 2011). Similarly, people’s evaluation of the economic situation of their household has been shown to be positively correlated with political support (Clarke et al., 1993). Moreover, we control for the respondent’s self-reported level of political interest. Taking into account the results from earlier research, it is not entirely clear what to expect. It has been argued that people with an understanding of political processes are more likely to have more optimistic views about democracy in general (Anderson and Tverdova, 2003). However, others argue that politically interested and well-informed citizens are likely to be more critical to the functioning of the political system (Norris, 1999, 2011). Earlier research has also demonstrated conflicting results when it comes to the relationship between generalized trust and political support but in line with the majority of empirical studies of the determinants of political support, we nevertheless include interpersonal trust as a control. The variable perceived closeness to a political party is related to political representation, which in recent research has been demonstrated to be an important determinant of democratic discontent (Dahlberg et al., 2015). At last, following many other studies, we also include a variable measuring the respondent’s self-placement on the ideological left–right scale. Since we have not found earlier studies, we are very uncertain about what to expect with regard to potential determinants of perceived responsibility. Since our main interest concerns the relationship between this variable and the reservoir of goodwill (political support), we chose to include the same control variables in these models. A more detailed overview of the indicators used in the analyses can be found in the Appendix.
Empirical strategy
Considering the structure of our data, we use a multilevel approach. As standard errors are likely to be too small in a regular ordinary least squares analysis, it is also more likely to make type I errors (Steenbergen and Jones, 2002). A multilevel model reduces this risk. Moreover, we use a random-intercepts model to allow for variation in the overall levels of political support or perceived responsibility. As a further robustness check, we ran the models also for each country separately in order to make sure that no country influences the results too much. These results are largely in line with the ones that are reported here. All full models including information about model fit and variance can be found in the Online Appendix.
Results
In the first step of our analysis, we examine whether people’s perceived responsiveness, as well as many-to-many congruence, is related to their support for their political system and thus whether it feeds into the overall reservoir of goodwill. Figure 4 shows the results of the first analyses and includes variables that have all been rescaled to a 0 –1 scale to allow for clearer comparisons of the size of the effects. The full models, the null model, and a model including only perceived responsiveness are reported in the Online Appendix. The figure shows that responsiveness is positively related to support for the political system and that this result is statistically significant. When people see their government as more responsive, they also tend to express more satisfaction with the working of democracy and the government. This provides strong support for our expectation that responsiveness adds to the reservoir of goodwill. Moreover, as the full models in the Online Appendix show, perceived responsiveness alone accounts for more than 13% of the between-country variation in support for the political system and almost 4% of the variation within countries. All independent variables together explain almost 54 and 20%, respectively. The right-hand panel of Figure 4 further illustrates that many-to-many congruence, where higher scores imply greater congruence, has a significant and positive effect on levels of support meaning that contexts with greater congruence tend to have higher levels of support. This further supports our theoretical proposition—even though congruence is not identical to responsiveness. Compared to the variables indicating alternative explanations, both perceived responsiveness and congruence appear to have a fairly big effect.
Effects of perceived responsiveness and congruence on the reservoir of goodwill. Note: Results are based on a multilevel analysis with random intercepts. The coefficients are graphically illustrated including 95% confidence intervals.
Furthermore, Figure 4 shows that the results of many of the variables indicating alternative explanations align with what we would expect from previous research. People having voted for a party ending up in government are significantly more supportive than political “losers.” Moreover, people who evaluate their family income as negative are less likely to support their political system. Men, people with higher education, higher interpersonal trust, more political interest, more ideologically to the right, and those who feel close to a political party, tend to be overall more satisfied.
In the second step of the analysis, we move on to test whether the reservoir of goodwill consequently helps governments to be more responsible, at least in terms of how they are perceived by their citizens. Figure 5 illustrates the results of the two full models again including only rescaled variables. It shows that having a reservoir of goodwill indeed appears to boost the perception that governments are responsible. The effect of the reservoir is statistically significant and positive, suggesting that the more positively someone evaluates the performance of the government and political system in general, the more likely s/he is to see the government as acting responsible. The effect size is quite substantial, also in comparison with the other indicators. A one-point increase in political support results in a 0.23-point increase in perceived responsibility. Moreover, the complete results (reported in the Online Appendix) highlight that the reservoir alone explains about 32% of the between country variance and almost 8% of within country variation. In the full model including also the alternative explanations, this explained variance increases merely to about 34 and more than 9%, respectively. Moreover, the right-hand panel of Figure 5 shows that the alternative measure of the reservoir, measured as the percentage of support before 2012, also has a positive and significant effect on perceived responsibility.
Effects of the reservoir of goodwill on perceived responsibility. Note: Results are based on a multilevel analysis with random intercepts. The coefficients are graphically illustrated including 95% confidence intervals.
Furthermore, the results suggest that men, those who are more trusting, more politically interested, more right-wing in political ideology, and identify with a party also tend to perceive the government as more responsible. On the other hand, income, education, and age do not seem to have an effect on the perception of responsibility. Interestingly, status as political “winner” or “loser” does not matter for perception of responsible government, at least in the left-hand panel. This might seem quite surprising. However, a government that acts responsibly rather than responsive actually chooses to divert from the short-term demands of their voters in favor of more long-term national interests. Responsible actions might therefore breach the “responsiveness ties” to their voters, making neither winners nor losers more supportive of those actions. The results presented in the right-hand panel, however, show that being a political “winner” does increase the likelihood of perceiving the government as more responsible. It seems that the individual-level measurement of support absorbs this effect, suggesting a more complex relationship between perceived responsiveness, being a “winner,” and political support.
Robustness of the results
In order to check the robustness of the results, we have conducted several robustness checks. In addition to the main models presented here, we test whether several country-level controls affected the main results. The results also remain similar when including only the individual indicators for political support (satisfaction with democracy and satisfaction with the government) and perceptions of responsibility. In addition, we use a multilevel random coefficient approach to test whether the slopes for the different countries are very different from each other. We find that 95% the coefficients for the relation between perceived responsiveness and support are in the range between 0.024 and 0.240 and those for the relation between support and perceived responsibility between 0.068 and 0.394. Overall, slopes tend to vary between countries but are generally in the same direction. The results of all these models can be found in the Online Appendix.
Moreover, in order to test that the results are not affected by country-outliers, we also replicate the model for each country separately. The main results are highly similar (in size, direction, and significance) for most countries. 3 The inclusion of the electoral winner/loser variable results in a smaller sample, as it excludes all non-voters from the analysis. We therefore also replicate our models without this variable, substituting it with a measure indicating whether the respondent voted or not. The main results remain robust. Moreover, as there are many missing values in the data, we have used a multiple imputation by chained equations method (with 10 imputations, 100 iterations, and a burn-in of 10) to impute the relevant missing values. Based on a now total of 47,515 observations, the results remain highly similar to the ones presented here—the coefficients of the main variables of interest are slightly larger when using the imputed data.
The results in the light of the financial crisis
Our analyses show that citizens in those countries that were hit particularly hard by the crisis also express less satisfaction generally, but there seems to be no effect on perceptions of responsibility. While this simple country-level control cannot account for the effects of the crisis in full, it does show that the representational link has been in part affected by this event. It needs to be noted that the survey data that we use here stem from 2012. As the crisis started in 2007, it is possible that our data from 2012 are still influenced by this event and therefore do not reflect a “typical” year. We should, however, also consider that we might have expected a different result if the crisis had a specific effect. Considering that governments sometimes had to make difficult—and often very unpopular—decisions, we might not have expected the positive link between responsiveness and responsibility in light of an effect of the crisis. People perceiving the government as unresponsive might still express a lack of support, but they would consider their governments responsible as it is measured here and vice versa. Although 2012 may still prove to be “exceptional,” also in terms of the responsiveness–responsibility link, our findings seem to hold despite the occurrence of the crisis.
Concluding remarks
It has often been argued that responsiveness and responsibility are conflicting phenomena, since, as highlighted by Mair (2009), governments find it increasingly difficult to combine responsive policy-making with responsible—and sometimes unpopular—decisions. Contemporary democratic governments do indeed have to be responsible not only to the long-term interests of their own citizens, but they also face constraints from binding commitments to governments in other countries and/or by decisions and agreements made by former governments. However, in this study, we have argued that responsiveness and responsibility actually are mutually reinforcing and, consequently, mutually diminishing.
Our argument draws on Easton, who hypothesized that, over time, governments that act responsive and perform well in the eyes of the citizens generate “a reservoir of good will that helps members to accept or tolerate outputs to which they are opposed” (Easton, 1965: 273). In terms of responsiveness and responsibility, our theory implies that governments that are being perceived as responsive to short-term demands from citizens generate something like a “responsiveness capital” that eventually allows governments to make important but non-responsive decisions and that these decisions are more likely to be accepted by the citizens. In this way, responsive government actions create leeway for governments to act in a responsible way.
We test our theoretical argument empirically using cross-national survey data from the ESS. Using multiple indicators of responsiveness, political support, and perceptions of responsibility, the multilevel regression analyses provide strong support for our theory. We show that high levels of perceived responsiveness are accompanied by higher levels of support, thus adding to a reservoir of goodwill. Subsequently, we show that the reservoir of goodwill then helps to create room for government to act responsible or at least being perceived as acting responsible. Thus, responsive government seems to be a precondition for citizens’ acceptance of responsible government.
This should have implications for how governments—and political scientists—view the functioning of representative democracy in general and in particular the often assumed trade-off between responsiveness and responsibility, as highlighted by Mair and others. Governments may sometimes feel that their hands are tied by external commitments. By continuously taking the public’s short-term policy preferences seriously, they generate a capital, which may be used in situations where they need to act more responsible than responsive. This also means that governments that act unresponsive have much less freedom to make non-responsive but necessary decisions. A generally unresponsive government making such decisions will run the risk of being punished by the voters and in terms of eroding legitimacy. Logically, the mutually reinforcing process we have investigated here should also translate into something like a vicious circle in a context where a government is generally perceived as lacking in terms of being responsive to the wishes and demands of the electorate. Such a public perception could very likely run the risk of resulting in increased dissatisfaction with the democratic system and it could have an eroding effect on the reservoir of goodwill (if not already empty). In such a situation, citizens should consequently be even less inclined to accept important, and seemingly unresponsive, decisions.
With this study, we have taken a first step to illustrate that responsiveness and responsibility do not need to be traded off per se. They can rather be related in a more positive way in which responsiveness facilitates responsibility. Future studies should aim to investigate this connection more closely. More specifically, while we focused here on the perceptions of responsiveness and responsibility, future studies should also examine in what way policy responsiveness and responsible acts of the government are related. Research should make efforts to operationalize these more “objective” concepts so that the connection between actual policy responsiveness and responsible government can be tested.
Supplemental Material
Supplemental Material, sj-pdf-1-ppq-10.1177_1354068818763986 - Responsiveness, support, and responsibility: How democratic responsiveness facilitates responsible government
Supplemental Material, sj-pdf-1-ppq-10.1177_1354068818763986 for Responsiveness, support, and responsibility: How democratic responsiveness facilitates responsible government by Jonas Linde and Yvette Peters in Party Politics
Footnotes
Acknowledgements
The authors would like to thank Mark Bovens, Stefan Dahlberg, Zoe Lefkofridi, Jan Karremans, Rubén Ruiz-Rufino, and the anonymous reviewers for their valuable suggestions on earlier versions of this article. The authors also acknowledge support from the Norwegian Research Council and the Bergen Research Foundation.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Norwegian Research Council (grant no 262986) and the Bergen Research Foundation (Grant No. 811309).
Supplemental material
Supplemental material for this article is available online.
Notes
Appendix
| Variable | Source | Measurement | Categories/scale |
|---|---|---|---|
| Individual-level indicators | |||
| Reservoir of goodwill | ESS 2012 | Index of the questions ‘(…) how satisfied are you with the way democracy works in [country]?’ and ‘(…) thinking about the [country] government, how satisfied are you with the way it is doing its job?’. Indicators are added up and divided by two. | 0 (extremely dissatisfied) to 10 (extremely satisfied) |
| Responsibility | ESS 2012 | Index of ‘(…) [P]lease tell me to what extent you think each of these statements applies in [country].The rights of minority groups in [country] are protected./Politicians in [country] take into account the views of other European governments before making decisions./The government in [country] explains its decisions to voters.’ Indicators are added up and divided by three | 0 (does not apply at all) to 10 (applies completely) |
| Perceived responsiveness | ESS 2012 | Combination of ‘[P]lease tell me how often you think the government in [country] today changes its planned policies in response to what most people think?’ and ‘[P]lease tell me how often you think the government in [country] today sticks to its planned policies regardless of what most people think?’ | 0 (never) to 10 (always) |
| Winner | ESS 2012 and Parlgov | Based on ‘Which party did you vote for in that election?’ people were categorized as a winner (when the party they voted for ended up in government) or loser (when the party they voted for did not end up in government) | 0 (loser) and 1 (winner) |
| Evaluation income (negative) | ESS 2012 | ‘Which of the descriptions on this card comes closest to how you feel about your household’s income nowadays?’ | 1 (Living comfortably), 2 (Coping), 3 (Finding it difficult), 4 (Finding it very difficult) |
| Education | ESS 2012 | ‘What is the highest level of education you have successfully completed?’ | 1 (less than lower secondary) to 7 (higher tertiary) |
| Female | ESS 2012 | Sex of the respondent | 1 (male) and 2 (female) |
| Interpersonal trust | ESS 2012 | ‘(…)would you say that most people can be trusted, or that you can’t be too careful in dealing with people?’ | 0 (you can’t be too careful) to 10 (most people can be trusted) |
| Political interest | ESS 2012 | ‘How interested would you say you are in politics?’ (the original index was recoded in reverse) | 1 (not at all), 2 (hardly), 3 (quite), 4 (very) |
| Left-right scale placing | ESS 2012 | ‘In politics people sometimes talk of “left” and “right”. |
0 (left) to 10 (right) |
| Age | ESS 2012 | Age of respondent | Actual age in years |
| Feel close to a political party | ESS 2012 | ‘Is there a particular political party you feel closer to than all the other parties?’ | 0 (no) and 1 (yes) |
| Reservoir of goodwill (<2012) | ESS 2002-2010 | Index of the questions ‘(…) how satisfied are you with the way democracy works in [country]?’ and ‘(…) thinking about the [country] government, how satisfied are you with the way it is doing its job?’. Indicators are added up and divided by two. The measure indicates the % of people that indicated they are satisfied, and is an average over time (where data was available for the countries that we include). | % satisfied, before 2012 |
| Many-to-many congruence | Golder and Stramski (2010) | Measure is based on data from the Comparative Studies of Electoral Systems (CSES), and is calculated an on the basis of: Σ|F1(x) – F2(x)|, where F1(x) and F2(x) are “the cumulative distribution functions for the citizen and representative preferences” (Golder and Stramski 2010; 96). The variable is reversed re-coded so that higher scores reflect higher congruence. | The higher the value, the higher the congruence |
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
