Measuring political legitimacy in two dimensions: internal and external measures

Research on political legitimacy encompasses two distinct traditions, one institutionalist and another drawing on political culture accounts of legitimacy. Recent contributions argue in favor of an integration of these two vantage points in the study of political legitimacy. Following these lines of reasoning, we investigate empirically whether a two-dimensional approach in political legitimacy research integrating the institutionalist and political culture perspectives can contribute to establish a more comprehensive and nuanced view on political regimes and their legitimacy. We conceptualize political legitimacy in combining an external normative with an internal empirical perspective, collect data for both and compare the relationship between these two dimensions of political legitimacy in an internationally comparative framework. Using data from Integrated Values Surveys and Varieties of Democracy we find that both dimensions are closely linked in general, discover important deviations from this pattern in the case of citizens' performance evaluations, pinpoint a particular group of hybrid cases that either lost internal or external legitimacy while collecting positive evaluations on the other dimension, and discuss the consequences of these findings for regime stability and future research on political legitimacy.

Introduction

In another contribution to this Frontiers in Political Science Topic, Wiesner and Harfst (2022) present a genealogy of political science research on political legitimacy. Based on a literature review they argue that political legitimacy is still an “essentially contested concept” and localize the roots of this contestation in the coexistence of two research traditions, one predominantly empirical and the other characterized by a strong normative stance. Wiesner and Harfst (2022) argue that the divide between these two traditions can be bridged in a two-dimensional conception of political legitimacy. Following the empirical tradition in legitimacy research that mainly draws on survey data, they propose an “internal” dimension that is based on citizen evaluations of their political system and an “external” dimension that is inspired by normative considerations on political legitimacy.

In this contribution we argue that this two-dimensional approach can be translated into empirical research by linking two related literatures: one institutionalist perspective on regime quality and another inspired by political culture research on regime support. We propose a conceptually sound integration of these two research traditions and demonstrate empirically that this combination yields new and fruitful insights into political regimes' legitimacy. Most importantly, the combination of an internal and an external dimension—which results in a matrix distinguishing four types of political legitimacy—allows us to identify regimes where political culture and institutionalist measures of legitimacy diverge. We argue that those mixed cases which are neither fully legitimate nor fully illegitimate have not been systematically identified so far but merit further academic attention. In order to pinpoint such hybrid cases, we use institutionalist data from the Varieties of Democracy (V-Dem) project and survey data form the Integrated Values Survey (IVS). This allows us to examine internal and external legitimacy in a sample covering democracies as well as autocratic regimes and to compare cases across all world regions and cultural spheres. We thus capitalize on the availability of easily accessible data to test the implications of a two-dimensional approach to political legitimacy. This allows us to derive some innovative analytical ideas from their combination, and to evaluate the potential of this two-dimensional approach for further research. Our contribution thus adds to both the political culture and the institutionalist literatures on political legitimacy and generates novel insights into the interplay between these two traditions and their respective measurement instruments. Notably, the introduction of a two-dimensional concept of legitimacy allows for the systematic identification of hybrid or mixed cases which qualify as legitimate on one but not on the other dimension.

In the following we first outline in more detail the two-dimensional approach to political legitimacy introduced above. We then link this approach to existing political culture and institutionalist accounts of legitimacy. While the political culture tradition brings the tools to operationalise the internal dimension of political legitimacy, institutional measures of regime quality allow us to investigate the external dimension of political legitimacy. The following section presents the conceptual foundations of the two-dimensional approach to legitimacy. We then identify data sources, propose variables and indices, and implement a mainly descriptive analysis. This enables us to draw a two-dimensional map of political legitimacy covering democracies as well as autocratic regimes. We find that internal and external measures of legitimacy diverge especially in the case of specific support for political institutions. This has implications for the functioning and sustainability of regimes, in particular in the case of democracies that depend on their citizens' support.

Conceptualizing political legitimacy in two dimensions

Wiesner and Harfst (2022) propose to conceptualize political legitimacy in two dimensions. Building on a genealogy of the last 50 years of research on political legitimacy, they argue that legitimacy is conceptually contested between an empirical and a normative research tradition. In the following, we thus define political legitimacy as a two-dimensional concept. Our point of departure for the subsequent analyses is the distinction between an internal and an external dimension of political legitimacy. The internal perspective on legitimacy focusses on citizens' beliefs and attitudes, the external one concentrates on expert judgements. The combination of an internal and an external axis results in four ideal types of political legitimacy (see Table 1): beyond fully legitimate and fully illegitimate systems, there are subjectively legitimate systems that can claim internal legitimacy but lack external legitimacy, and formally legitimate systems that can build on external legitimacy but lack internal support among their citizens.

Table 1

Table 1. Four ideal types of political legitimacy.

Compared to one-dimensional approaches, conceptualizing political legitimacy as a combination of an external normative and an internal empirical perspective has an added value on three fronts: it first allows us to examine political legitimacy of both democratic and non-democratic regimes; second, the distinction of four types of political legitimacy enables us to identify the presence as well as the absence of political legitimacy in terms of institutionalist as well as political culture terms. Finally, and most importantly, we can also show whether and where external and internal legitimacy diverge.

The internal dimension of political legitimacy

The internal dimension of political legitimacy focusses on empirical assessments of legitimacy based on citizens' individual values and attitudes. This older research strand on political legitimacy is anchored in the political culture tradition of political science. It has flourished in the aftermath of World War II as an attempt to explain the breakdown of many European democracies in the 1920th and 1930th (for an overview see Wiesner and Harfst, 2022). It is inspired by the discipline's surge for an analytical foundation of its concepts and analyses and inscribed into comparative political science's reorientation away from purely descriptive institutionalism toward causal reasoning (Macridis and Cox, 1953). Lipset (1959) spelled out the empirical implications of this approach to research on political legitimacy and defined legitimacy as the degree to which a political system's values fit with those of its citizens, thus emphasizing the individual foundations of the concept. In close resemblance to Easton's (1953, 1975) notion of system support, Lipset claimed that political legitimacy, alongside with economic performance, is a crucial prerequisite of system stability. In a similar way, Almond and Verba (1963) operationalise a “civic culture” supportive of democracy in terms of individual political attitudes accessible through survey research. Since such values and attitudes are carried by individual citizens in different countries and are thus shaped by their perception and evaluation of these regimes, Wiesner and Harfst (2022) consider them as “internal” evaluations.

The political culture tradition of legitimacy research thus frames legitimacy in terms of system support. Most analyses in this tradition follow Easton's (1965, 1975) distinction between citizens' diffuse and specific support for their regime. Diffuse and specific support can be regarded as two extreme points on a continuum (Norris, 2011a, p. 24). The most specific form of support focusses on citizens' evaluations of individual political leaders, is highly performance related, determined by regimes' policy outputs and thus subject to short term changes. The other extreme point of the continuum taps into citizen's deeply rooted and stable values and principles regarding political regimes and their fundamental values. Easton suggested that specific and diffuse support are related to each other. First of all, diffuse support “forms a reservoir of favorable attitudes or good will that helps members to accept or tolerate outputs to which they are opposed or the effect of which they see as damaging to their wants” (Easton, 1965, p. 273). These favorable attitudes, however, do not arise in a vacuum. They are nourished by positive experiences with the regime's performance and its outputs. Easton expects positive “spillover effects from evaluations of a series of outputs and of performance over a long period of time” (Easton, 1975, p. 446). The relationship, however, can also be negative in the case of bad performance: if “little improvement in outputs occur, it will be impossible to prevent dissatisfaction from shifting toward the regime” (Easton, 1965, p. 231). And indeed, Claassen and Magalhães (2022) show that negative performance experiences first undermine specific regime support and then spill over to principled support.

The internal dimension of political legitimacy is thus characterized by citizens' support for a broad range of objects, ranging from trust in individual actors and institutions to value orientations regarding the best form of government. Based on citizens' individual regime evaluations, it thus entails specific as well as diffuse regime support. Building on results from survey research, it can be labeled as empirical in the sense that it does not subject these attitudes and values to a normative judgement.

The external dimension of political legitimacy

By contrast, the external dimension of political legitimacy emphasizes a normative perspective. It is rooted in an account of legitimacy that has developed in reaction to the empirical-analytical approach and highlights the importance of critical political theory. Its proponents (e.g., Habermas, 1976, 1984; Schaar, 1984) stipulate a normative perspective on legitimacy, blame the empirical-analytical approach's socially affirmative positions, and demand to conceptualize legitimacy beyond citizens beliefs. These critics suggest regime evaluations based on normative criteria. Given that normative evaluations are based on over-individual and internationally applicable fundamental considerations on regime principles, and thus exist independently from given political systems, Wiesner and Harfst (2022) consider this normative approach as the “external” dimension of political legitimacy. The question that arises with the definition of external, normative criteria, however, is which criteria are applied and how we assess their successful implementation. We might, for example, require a regime to regularly hold free and fair elections in order to classify it as legitimate. But we could also extend the list of normative criteria and include citizens' equality and the rule of law as additional criteria. Other normative standards to evaluate regime legitimacy might include social justice, environmental performance or human development. In principle, the external dimension of legitimacy thus remains open for any (normative) definition that researchers might want to implement. The choice of the external dimension's definition depends on researchers' normative judgement and the research interest at hand. The only requirement is that these definitions are made explicit, are open to critique, and remain susceptible for re-definition.

The empirical literature that comes closest to such a normatively inspired research on political legitimacy is rooted in institutionalist accounts of regime quality. It normatively posits liberal democracy as the most desirable form of government and then introduces institutions, rights, and freedoms that implement democracy. Historically, it departed from a minimalist version of democracy and emphasizes the importance of free and fair elections. It then evolved and included more and more additional elements. Recent contributions in the field like the Democracy Barometer (DB) (Bühlmann et al., 2012) and the Varieties of Democracy Project (V-Dem) (Coppedge et al., 2011) now propose highly refined conceptions of regime quality. The Democracy Barometer, for example, evaluates a regime's equality, freedom and control dimensions. V-Dem advances an even more detailed account of political institutions, covering electoral, liberal, participatory, deliberative, and egalitarian elements. The broad inclusion of different elements allows DB and V-Dem to paint a nuanced picture of contemporary and—in the case of V-Dem—historic political regimes around the world. It enables researchers to examine correlates of democracy (or its absence) that might vary across these dimensions. The existence of such highly differentiated measurements for regime quality also makes clear that—like in the case of regime support—we are again confronted with a continuum, ranging from fully fledged autocratic rule to democratic regimes (Merkel, 2004).

Institutionalist measures provide a normative element to our external dimension of political legitimacy since every institutionalist regime evaluation compares regimes against a minimum standard of institutional settings that need to be met. In that sense, our external dimension of political legitimacy as we conceive it here, uses a certain level of democracy as a benchmark that regimes need to fulfill in order to be classified as legitimate.

Integrating political culture and institutionalist approaches to political legitimacy research

Integration of the two vantage points of research on political legitimacy is a relatively recent development in the literature on political legitimacy. Coming from an institutionalist and political culture tradition, respectively, Weatherford (1992) and Beetham (2013 #2070) both advance similar solutions to such an integration. Beetham (2013) puts forth a threefold matrix of political legitimacy that includes the rules that govern a polity, the justifiability of these rules in terms of citizens' beliefs, and citizens' expressed consent. Beetham's conception thus includes an institutionalist element—namely a regimes' constitutional and legal foundations—, as well as empirical components—namely citizens' beliefs and consent. In a similar vein, Weatherford (1992) proposes to supplement political legitimacy's micro perspective emphasizing individual beliefs with regimes' macro features at the system level. Both authors thus blend institutional arrangements and citizens' individual attitudes in order to bridge the empirical-normative divide in research on political legitimacy. More recently, also Mayne and Geissel (2016) invite researchers to combine individual level attitudinal data and the institutionalist account of regime quality in order to draw a more adequate picture of political legitimacy.

Fuchs and Roller (2018) as well as Pickel et al. (2016) follow this invitation and engage in an empirical operationalisation of the proposed link between institutionalist and political culture approaches. Pickel et al. (2016, p. 652) conclude that “individual-level evaluations [...] do indeed provide a complementary perspective” to institutional accounts. However, both analyses limit their empirical examination to European democracies. Furthermore, both analyses are constrained by a comparison of rankings of states. In order to further develop these proposals to integrate institutionalist and political culture approaches, additional conceptual and empirical work is needed.

Against this backdrop we develop a proposal that tackles the conceptual challenges of such an integrated perspective on political legitimacy via the proposed two-dimensional approach to political legitimacy. Compared to the work cited in the previous paragraph, it can include democracies as well as non-democracies into the analysis. Furthermore, it is able to discern between the legitimacy of institutional arrangements that can be evaluated against normative criteria and citizens attitudes vis-a-vis the regime. This allows to draw a fuller picture of political legitimacy. Most importantly, the two-dimensional approach enables us to identify hybrid regimes which either lack citizen support or a positive evaluation by institutionalist measures of regime quality. We label such ambiguous cases “subjectively” and “formally legitimate” regimes (see Table 1).

Two lines of research that place hybrid cases with divergent classifications in the internal and external dimensions of legitimacy at the center stage come to mind at this point: works that detect dissatisfaction among citizens in democracies and analyses that find high levels of democratic values in autocracies. The former line of research notes widespread principled support for democracy combined with dissatisfaction with the way democracy performs (Klingemann, 1999; Dalton, 2004). Norris (2011b) interprets this discrepancy positively. She claims that “critical citizens” improve the quality of the democratic process. They take democracy by the word and, while being principally convinced by the advantages of democracy, engage into criticizing the specific performance of institutions and actors. This critical potential within democracies has been anticipated by Eastons analytical distinction between diffuse support for regime principles and specific satisfaction with the performance of its actors and institutions. Klingemann (2014) shows that Norris' optimistic view on “dissatisfied democrats” is well founded empirically, at least in European states. He suggests that “dissatisfied democrats might mobilize political pressures for democratic innovation and reform by addressing new issues and opening up the political process” (Klingemann, 2014, p. 138). The broad preference for democratic principles is rooted in a set of supportive civic attitudes, even if support becomes more instrumental and performance oriented. In particular, compared to satisfied democrats, dissatisfied democrats are characterized by low levels of confidence in order institutions and the parliament. High shares of dissatisfied democrats would lead to a regime classification as “formally legitimate.”

The latter line of research identifies broad support for democratic values even in autocratic regimes. While early accounts of the phenomenon optimistically interpreted support for democratic values in autocracies as potential for democratic reform (e.g., Mishler and Rose, 2002), more recent approaches advance two more pessimistic interpretations. Either respondents surveyed in autocratic regimes mistrust the survey organization, fear negative consequences and report false preferences in order to align their answers to the perceived expectations of the regime. Tannenberg (2022) finds “self-censorship to be a severe issue in most autocracies” he examines. Respondents avoid socially undesirable answers to sensible questions tapping into regime preferences and the performance of institutions and actors. The other reason for a high level of support for democracy in autocratic regimes, noted for example by Norris (2011b), might be a misunderstanding of the concept of democracy among the citizens in these regimes. This results in “authoritarian notions of democracy” (Kirsch and Welzel, 2018). Support for democracy then coexists with an absence of liberal attitudes and emancipative values as well as a preference for army rule, unchecked power of religious authorities and uncritical obedience to rulers. Both preference falsification and authoritarian notions of democracy would lead to a regime classification as “subjectively legitimate.” In the following section, and in order to systematically identify and describe such hybrid cases, we show the distribution of countries among the four types of our two-dimensional map of political legitimacy.

Two dimensional measurement of legitimacy: empirical perspectives

In the preceding section, the internal and external dimensions of political legitimacy have been tied to the political culture and the institutionalist tradition of research on regime quality. In the political culture tradition, measures of individual values and attitudes are used to determine the degree of system support among the citizens of a regime. In the institutionalist literature, macro indicators or expert evaluations are used to determine a regime's quality.

Turning to individual level measures of regime support first, we can note that survey researchers have proposed a variety of items to measure the different levels of support along the continuum between specific and diffuse support. Given the multidimensionality of citizens' regime support, it is clear that an examination of the internal dimension of political legitimacy should, at best, include indicators covering different levels of this continuum. Also, since problems at the specific end of the support continuum can have long term consequences for principled regime support among citizens, it is important to look at both when we examine political legitimacy. We begin with the most diffuse form of support and first present citizens' principled democratic values. At this end of the continuum of support, we include a certain degree of normative judgements in that we insist on democratic values. This corresponds with the external measure that we present below and helps us to examine whether there are differences between support for democracy and support for individual institutions like parties, parliament, and government that need not be democratic.

Fuchs and Roller (2018) explore citizens' democratic values using measures of support for hierarchically ordered visions of democracy. Their hierarchical conceptualization of individual support for democratic values entails that citizens need to support the lower levels of democratic values before they can meaningfully support hierarchically superordinate democratic values. Logically, this hierarchy implies and “AND” relationship between the different levels of democratic support. In technical terms, this hierarchy is mapped by a dichotomous coding of support of each level of democratic values and a multiplication rule. This makes sure that rejection of a lower level of democratic values cannot result in support for a higher level value. For example, if a respondent does not adhere to fundamental democratic values, she cannot receive a positive value for the support of (hierarchically superordinate) electoral democracy since “fundamental democracy” ^* “free elections” always results in a value of 0 if “fundamental democracy” is coded 0. This also implies that support for the different levels of democratic values will progressively decrease from level to level since it is technically impossible that more respondents support higher than lower levels of democracy.

In the following analyses we only focus on one of the lowest levels of democratic values proposed by Fuchs and Roller (2018), support for electoral democracy. In restricting our examination to this value orientation we take advantage of the fact that most non-democratic regimes today regularly hold elections so that citizens are aware of the potentials (and potential violations) of free and fair electoral competition. In a first step, support for fundamental values of democracy is operationalised through a positive answer to the item “Having a democratic political system” AND rejection of at least one of the items “Having a strong leader” or “Army rule.” The resulting dichotomous variable “Fundamental Democratic Values” is then taken to calculate the “Electoral Democratic Values” index (“Fundamental Democratic Values” AND “People choose leaders in free elections” AND “Women have same rights as men”)¹. To construct our “Electoral Democratic Values” index, we use data from the Integrated Values Surveys (IVS) time series including all European Values Studies (EVS) (EVS, 2022) and World Values Surveys (WVS) (Haerpfer et al., 2022) from 1981 to 2022². This allows us to cover a broad range of countries and time points, including democracies and autocratic regimes from all over the world.

Turning to the specific end of the regime support continuum, the IVS time series provides two sets of indicators that tap into this concept. One asks for trust in different institutions, the other asks for respondents' “satisfaction with the way democracy develops” (SWD) in their country. While citizens' democratic values mark the diffuse end of the continuum, trust is the most specific form of support, and SWD falls somewhere in between these two. In the following, we examine trust in parties, parliament, and government but do not consider SWD because of the item's unclear position on the continuum between diffuse and specific support (see, for example, Linde and Ekman, 2003) and the difficulties of interpretation that result from this ambiguity.

Measures of regime quality that we use to operationalise the external dimension of legitimacy have a long measurement tradition using macro level indicators (Vanhanen, 1997; Marshall and Gurr, 2020; Repucci et al., 2021) but no consensus has emerged on how to best achieve reliable and valid results (Munck and Verkuilen, 2002; Müller and Pickel, 2007). However, more recent efforts like the Democracy Barometer (DB) (Bühlmann et al., 2012) or the Varieties of Democracy Project (V-Dem) (Coppedge et al., 2011) seem to have solved many of their predecessors' conceptual and measurement problems and can be used to draw an internationally comparable picture of regime quality. In the following analyses, we opt for the use of V-Dem's “Electoral Democracy Index” that taps into “the core value of making rulers responsive to citizens” (Coppedge et al., 2022, p. 43). It matches our individual level indicator for democratic values developed by Fuchs and Roller (2018) which was explicitly designed to transpose V-Dem's concept of electoral democracy to the level of individual values and it covers all the countries and time points for which IVS data is available.

Turning to the empirical investigation on the relationship between the two dimensions of political legitimacy, we first examine the internal dimension of legitimacy. It revolves around citizens' democratic values and regime evaluations and ranges between most specific and most diffuse forms of regime support. The IVS time series covers a total of 393 surveys in 112 countries from all over the world, thus including democracies and non-democracies in all world regions and across different cultural spheres. Data for electoral democratic values is available for a total of nearly 285,000 respondents in 163 surveys covering 87 countries.

The question that now arises is in how far these relationships differ among countries and whether there are notable differences between democratic and autocratic regimes. In order to examine country level differences, we aggregate IVS data at the country-year level and match it to V-Dem measures³ of regime characteristics (also available in country-year format) for the years when the IVS survey was fielded in the respective country. Like in the previous section on individual level data, we look at specific support (trust in parties, parliaments, and executives) as well as diffuse support that respondents express through support for electoral democratic values. In other words, in the following, we examine the link between external and internal measures of political legitimacy in order to assess whether the introduction of two dimensions into the analysis of political legitimacy provides any additional insights.

Using the “Electoral Democratic Values” as well as V-Dem's “Electoral Democracy Index” we concentrate on the most basic democratic values at the citizen level as well as their realization in different national settings as measured by V-Dem. We thereby do not impose unrealistic requirements neither on the enthusiasm for democracy among citizens nor on the realization of democratic ‘nice-to-haves' at the state level. We restrict our analysis to the most fundamental core of democracy in order to impose a fair test even for the weakest competitors. The IVS was fielded in countries as divers in democratic standards as Finland and Burkina Faso. This allows for cross-cultural comparisons of regime support and support for democracy. At the same time, we cannot expect that respondents countries that hardly fulfill or even fail most basic democratic standards will be intrigued by the desire for more deliberative elements or referenda in their national institutions. In order to examine whether the relationships differ between established democracies and all other states, we first perform an analyses on the full sample including all V-Dem and IVS countries for which data are available (up to 381 country-years). Two additional sets of analyses then differentiate between democratic and autocratic regimes. We impose the threshold between autocracy and democracy at a V-Dem “Electoral Democracy Index” level of 0.5. All states attaining this threshold are considered as democracies while all states falling below this threshold are coded as autocracies. Applying the threshold results in up to 111 country-year observations in autocratic and 270 observations in democratic states.

Let us first examine average support for institutions and democracy in democratic and autocratic states. Table 2 shows the respective share of respondents who support electoral democratic values or trust in parties, parliaments, and governments. In the case of democracies, a majority of respondents supports democratic principles while less than half of the respondents trust in parties, parliaments and governments. Governments garner the highest trust rates, parties the lowest and parliaments fall into an intermediate position. The general pattern of institutional trust in autocracies resembles the one in democracies, with governments taking the lead, followed by parliaments and parties being the least supported institution. However, the levels of support differ to a great degree from the ones in democracies. More than half of the respondents in autocratic regimes trust in governments (democracies: 38.7%). Also parliaments and parties enjoy considerably higher support rates in autocracies (parliaments: 47.8%; parties: 35.5%) than in democracies (parliaments: 36.4%; parties: 22.0%). At the same time, principled support for electoral democratic values is considerably lower in autocracies (37.1%) than in democracies (55.2%). We thus find higher levels of institutional trust in autocracies than in democracies. In particular, there is a positive bias toward support for executives in both autocracies and democracies, but considerably stronger in autocracies. This is combined with a notable divergence in support for basic values of democracy being nearly 20 points higher in democratic regimes. In other words: respondents in democracy support democratic values while those in autocracies support political institutions.

Table 2

Table 2. Specific and diffuse support in democratic and autocratic states.

When we now turn to relationships between our internal and external dimension measures, we first note that in the full sample (including both democratic and non-democratic states) internal democratic values and the external V-Dem measure of democratic quality are positively and strongly associated (see Table 3). Democratic values thus grow stronger with higher levels of institutionalized democracy. The three performance indicators that measure specific regime support, however, are negatively associated to V-Dem's electoral democracy index, indicating that higher levels of external legitimacy are associated to lower trust in institutions. Both findings mirror the results from Table 2.

Table 3

Table 3. Relationships between internal and external dimensions of political legitimacy.

Looking at differences in these associations between democratic and autocratic regimes, we observe notable discrepancies among these two groups. First, the external political legitimacy dimension is positively related to democratic values in democratic states but negatively (though not statistically significant) in autocracies. Increases in the institutional quality of democracy thus translate into higher levels of principled support for democratic values in democracies, but not in autocracies. The picture is somewhat different in the case of trust in parties, parliament and government. Here, we observe mostly negative associations with V-Dem's “Electoral Democracy” index across both autocracies and democracies. Differences emerge when we consider the significance levels. In the case of democratic states, only parties are evaluated significantly more negatively as V-Dem's electoral democracy score increases. The coefficients in the case of parliaments and governments are not statistically significant. When we look at autocratic states, all trust measures are significantly and negatively associated to the external democracy measure. Thus, higher levels of external legitimacy translate into lower levels of institutional trust in autocracies. In democracies, this is only true in the case of parties.

To summarize, we observe high levels of principled support for democratic values in democratic states that increases as external legitimacy increases. Levels of principled support are much lower in autocracies and there is no statistically significant relationship with external legitimacy. This result adds a nuance to the findings presented above where we noted that democratic values might exist in autocratic regimes. This may indeed be the case but increases in external legitimacy do not translate into growing support for democratic values. This might present a burden to democratization. Regarding levels of institutional trust in democracies and autocracies, our results show that respondents in autocracies express higher levels of trust than those in democracies. We can interpret this finding as evidence confirming a tendency of preference falsification in autocratic regimes where respondents refrain from criticizing the institutions that carry the regime. Interestingly, the association between our external measure of legitimacy and trust in institutions is negative, indicating that preference falsification might become less of an issue when regimes democratize and citizens become more willing to evaluate political institutions negatively. If this interpretation is correct, this would be good news for the development of a political culture allowing for critical views. At the same time, a critical political culture seems to focus on parties in democratic regimes, especially as they grow more legitimate externally. Critical citizens can be interpreted as a democratic achievement, especially if support for democratic values is high. However, the finding that dissatisfaction affects parliaments and—most notably—executives to a lesser extend than parties contradicts Klingemann's (2014) findings. This might be a warning signal for established democracies. Contrary to Klingemann's interpretation, dissatisfied democrats might not criticize government's performance (and engage in attempts to change policy outputs through elections) but become more and more at odds with a necessary element of representative democracy, the parties. This might hint to an increase of populist attitudes among citizens in democratic regimes who reject pluralist visions of democracy in favor of the expression of a general will of the people (Mudde, 2004, 2017).

In order to push the presentation closer to a view of political legitimacy as a two-dimensional concept, we move on to graphically represent these relationships (see Figures 1–4). In these graphs, the external dimension of legitimacy, measured by V-Dem's “Electoral Democracy” index is represented by the scatterplots' x-axis. The dashed gray line indicates the threshold between autocracies and democracies. Autocratic regimes fall below the threshold of an index value of 0.5 while democratic regimes are located on or above this line. The y-axis shows the internal dimension of legitimacy. It depicts the share of citizens in a given country that support democratic principles, and trust in parties, parliaments or governments. Comparable to the autocracy-democracy divide on the x-axis, the dashed gray line indicates a threshold. Below the line, <50% of IVS respondents express support for the respective value or institution. Above the line, more than half of the citizens declare diffuse or specific regime support. Thus, the four quadrants in each graph represent one of the four ideal types of legitimacy that result from the application of a two-dimensional concept. In quadrant I we find fully legitimate regimes; in quadrant II we find subjectively legitimate systems that can only claim to be supported by their citizens but have no external recognition of their regime qualities; in quadrant III fall fully illegitimate systems; finally, in quadrant IV are located formally legitimate regimes that can claim external approval of its status as democracies but lack internal support of their citizens.

Figure 1

Figure 1. External legitimacy and citizens' democratic values.

Figure 1 shows the relationship between the external measure of regime quality provided by V-Dem and one of our internal measures of political legitimacy, citizens' diffuse support for electoral democratic values. We observe the clearly positive relationship that we can expect given the coefficients reported in Tables 2, 3 above. Most countries fall in the quadrants I, III, and IV. In many of them—autocracies (quadrant III) as well as democracies (quadrant IV)—a majority of citizens does not support electoral democratic values. Only countries in quadrant I can be considered fully legitimate democracies according to our two-dimensional conception of legitimacy: institutions guarantee democratic standards and the majority of citizens share democratic values that support these institutions. They are internally and externally legitimate and can thus be considered as stable democracies.

When we turn to specific regime support and first inspect the relationship between our external measure of regime quality and trust in parties (see Figure 2), parliaments (see Figure 3), and governments (see Figure 4) we observe negative relationships in all cases. Given the results reported in Tables 2, 3, we expected to find this general pattern. It is most noteworthy that parties are regularly evaluated negatively. There are almost no cases where a majority of citizens living in externally legitimate states express trust in political parties. Interestingly, the level of parties' trustworthiness is particularly high in two one-party states that score lowest on the external V-Dem measure of regime quality, China and Vietnam. As we move on to trust in parliaments and governments, the overall share of citizens expressing positive evaluations of these institutions increases and with it the number of fully legitimate regimes. This finding mirrors our results presented above. Parties enjoy the least support among political institutions, even in regimes that can claim external legitimacy, qualifying as electoral democracies according to the V-Dem measure. Executives benefit from a prime in trust, in particular in autocratic regimes. Parliaments tend to be evaluated more positively in democracies.

Figure 2

Figure 2. External legitimacy and citizens' trust in parties.

Figure 3

Figure 3. External legitimacy and citizens' trust in parliament.

Figure 4

Figure 4. External legitimacy and citizens' trust in government.

With regard to the question whether a conceptualization in two dimensions increases our insights in the phenomenon of political legitimacy, it is important to have a particularly close look at those cases that figure in the two categories where internal and external evaluations of political legitimacy diverge (see Table 4). Subjectively and formally legitimate systems would have been wrongly classified as either legitimate or illegitimate, depending on whether an institutionalist or political culture approach had been used. In our two dimensional model, they can be flagged as debatable cases that might require (or allow for) additional research. In the case of diffuse support, about 34% of cases are classified as hybrid case. They are either formally (27.7%) or subjectively (6.5%) legitimate.

Table 4

Table 4. Classification of cases in two ideal types of hybrid regimes, in per cent.

First of all, these findings show that there is a widespread match between democratic values in the internal dimension and external regime quality evaluations. In other words: in more than 65% of all cases included in our analysis, people live in a regime that fits their most fundamental political values in terms of support for electoral democracy. However, the misfit is particularly large in countries that our V-Dem index on “Electoral Democracy” classifies as democratic. 27.7% of cases (as compared to 34.2% of all hybrid cases) fall into this group. This means that the observed mismatch between internal and external legitimacy in the field of democratic values can in large parts be traced back to country-years where citizens in democratic states lack diffuse support for electoral democracy in principle terms. This represents a severe problem for the sustainability of democracy in these states, especially if diffuse support is a “reservoir of favorable attitudes” (Easton, 1965, p. 273) toward the regime.

Looking at our three trust measures, the share of countries that cannot be classified in a clear-cut manner as either fully legitimate or illegitimate more than doubles and increases to around 70%. Again, in their vast majority, these hybrid cases fall into the group of formally legitimate regimes. There is a mismatch between an externally legitimate democratic regime that receives negative institution ratings from its citizens in the internal dimension. In parts, this mismatch might be explained by dissatisfied citizens' critical evaluations of institutional performance. Performance driven dissatisfaction in democracies should, however, mainly affect governments that citizens could replace in elections. But this is not the case. Governments in democracies receive more positive evaluations than parliaments and parties. If there is virtually no fully democratic state where a majority of citizens trusts in parties, this might become a problem for representative democracy that requires parties, especially opposition parties, and could be an expression of populist attitudes. In our externally illegitimate cases which do not qualify as democracies, institutional trust is also low and also affects parties more than parliaments and governments. However, the proportion of observations where these institutions receive positive evaluations regardless the regime's non-democratic character is relatively high. This might either be a sign of preference falsification or point to the relative stability of these regimes and their ability to organize (most often performance related) specific support.

The most problematic cases are those where democratic values as well as trust in all three institutions are low. In democratic regimes, institutional dissatisfaction cannot be bolstered by principled regime support. In autocracies, the lack of support for democratic values is a burden for the democratization of these regimes. In a final step of the analysis, we therefore concentrate on those cases that regularly figure among the formally and subjectively legitimate systems. This might help us to identify specific patterns that lead to such ambiguous classifications. When we now list all country-years that classify as hybrids in each of the possible combinations of one external and four internal legitimacy measures, we only observe three cases which are consistently classified as hybrids in the “subjectively legitimate” category (located in quadrant II in Figures 1–4): China 2007, Singapore 2020 and Vietnam 2006. China and Vietnam already came to our attention when we analyzed the discrepancies in external political legitimacy and the particularly positive evaluation of parties in these one-party states. These two autocratic regimes as well as Singapore seem to be particular well equipped to ensure citizens' commitment in terms of both diffuse and specific support.

Turning to the other group of hybrid cases, the ones consistently classified as “formally legitimate” (quadrant IV in Figures 1–4) we first note that this group is considerably larger than the “subjectively legitimate” one. Thirty country-years in 20 countries fall into this group⁴. Among these are eight Latin American states with 14 observations. The remaining hybrid cases scatter more or less all around the world. About half of the countries and observations in the category of “formally legitimate” states thus seems to be clustered in one particular world region. The example of Brazil might be telling in terms of the potential consequences of a period of formal legitimacy. In 2018, the last year for which data is available for Brazil, Jair Bolsonaro was able to win the presidential election after more than a decade of formal legitimacy. His campaign was based on an anti-democratic and populist platform. Once in office, he openly opposed Congress and the courts and at the end of his term he fueled doubts about the electoral process that brought him out of office in October 2022. Thus, formally legitimate democracies face a real risk of collapsing—but they also have the strength to overcome such authoritarian threats.

All in all, our data points to the fact that there are quite a few cases which fall into the ambiguous categories of formally or subjectively legitimate regimes which either lack internal or external legitimacy. While there is a relatively high consistency in the approval rates for diffuse electoral democratic values and the corresponding V-Dem evaluations, the discrepancies are more numerous in specific support for individual institutions. The widespread match between internal and external legitimacy in the case of principled democratic attitudes is good news. It indicates that citizens have a realistic view on the democratic quality of their regime (also see Brunkert, 2022). Furthermore, democratic values as a long term attitude which are in line with external measures of regime quality are a good sign since they show that there is a reservoir of democratic values in (formally) democratic states that they can build on to sustain the regime.

The more pronounced deviations between the specific forms of internal support and external measures first of all show the short term and performance driven character of these evaluations. Citizens take critical positions toward political institutions when they have deviating policy preferences or in times of crisis. However, the reservoir of principled regime support might suffer when bad performance evaluations spill over and erode diffuse support in the medium to long term. In this regard, the 20 countries in the group of formally legitimate states (see text footnote 4) that lack both specific and diffuse support are in particular danger. Even if we have not analyzed the data at hand over time, such negative spill overs seems to already have taken place in these countries so that it might become difficult to sustain democracy in the long run.

Conclusion

This contribution departed from the question how to best measure political legitimacy. We propose a conceptualization of political legitimacy along two dimensions, one internal and one external. The internal dimension ties in with the political culture tradition in legitimacy research and highlights the importance of citizens' values and beliefs for legitimate regimes. The external dimension is rooted in legitimacy research's institutionalist tradition and focuses on expert evaluations and the interpretation of official data to judge the legitimacy of a regime. The combination of an internal and an external dimension in our conceptualization of political legitimacy allows us to bridge the gap between empirical and normative elements in legitimacy research. The internal dimension builds on the numerous survey items that empirically minded researchers have devised and improved in the last decades. The external dimension allows for the integration of normative judgements on the quality of political regimes.

The added value of such a conceptualization of political legitimacy along two dimensions is to provide new and maybe unexpected insights especially on hybrid cases that fall into different categories on the two dimensions. Regimes that qualify as legitimate in one dimension but fail to do so in the other have not been systematically investigated so far. We rely on V-Dem's “Electoral Democracy” index to operationalise the external dimension and on items from IVS to measure the internal dimension of political legitimacy. We discover a close association between the external measure and principled regime support while there are impressive discrepancies between V-Dem's measure and citizens' specific, performance based support for particular institutions. This finding per se is not particularly astonishing given the different temporal patterns in specific and diffuse support. Specific support is inherently short term in nature and a function of a regimes' ability to deliver on citizens' demands. A drop in specific support is therefore not critical if the regime is able to improve its performance. If, however, specific support remains low for a longer period, this negative performance evaluation might spill over to principled regime support. In this respect, it is most noteworthy that our conceptualization of political legitimacy along two dimensions has identified 20 mainly Latin American countries which enjoy formal legitimacy only. In all these cases and in spite of a positive external judgement of the regime, all internal legitimacy measures point to the negative. In other words: a majority of citizens not only evaluates regime performance negatively but also rejects fundamental democratic standards. Given that this pattern prevails for more than a decade in some countries (e.g., Brazil and Mexico), democracy might well be in acute danger. Without a two dimensional concept of legitimacy along an internal and an external dimension, we would not have able to systematically detect this particular pattern of problematic and long lasting discrepancies between institutional measures and citizens' values, beliefs, and attitudes.

We therefore believe that a conceptualization of political legitimacy that includes internal as well as external evaluations is of fundamental importance for the assessment of political legitimacy. Since especially specific internal support can considerably deviate from external evaluations and since these negative evaluations of regime performance can spill over to principled diffuse support in the long run, surveying both the internal and the external dimension of political legitimacy is of paramount importance. Future research building on the findings presented here could systematically evaluate temporal patterns in the development of internal and external legitimacy and examine the reasons for observed mismatches between the two dimensions of political legitimacy.

Data availability statement

Publicly available datasets were analyzed in this study. Replication data and syntax is available via Harvard Dataverse at https://doi.org/10.7910/DVN/BPKBPD.

Author contributions

PH was responsible for text and quantitative analyses. CW was responsible for the text. All authors contributed to the article and approved the submitted version.

Acknowledgments

We acknowledge support by the Open Access Publication Funds of the University of Göttingen.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

1. ^For a detailed presentation of the democratic values indices and the thresholds used to dichotomise each item see Fuchs and Roller (2018, p. Table 1).

2. ^Given the translation problems for the “Army rule” item, Vietnam 2001 was removed from this and all the following analyses (cf. Meuleman et al., 2023). All analyses based on IVS data use IVS's weights defined in its variable S017. For details see the documentation of the IVS time series at https://europeanvaluesstudy.eu/?s=ivs, accessed December 5, 2023.

3. ^See https://www.v-dem.net/vdemds.html, accessed March 21, 2022.

4. ^Armenia 2018, 2021; Bolivia 2017; Brazil 2006, 2014, 2018; Chile 2018; Colombia 2012, 2018; Ecuador 2018; Guatemala 2020; Japan 2010; Maldives 2021; Mexico 2005, 2012, 2018; Mongolia 2020; Nigeria 2012, 2018; North Macedonia 2019; Peru 2012, 2018; Slovakia 2017; South Africa 2013; South Korea 2005, 2010; Tunisia 2013, 2019; Ukraine 2020; Zambia 2007.

References

Almond, G. A., and Verba, S. (1963). The Civic Culture: Political Attitudes and Democracy in Five Nations. Princeton, NJ: Princeton University Press. doi: 10.1515/9781400874569

Crossref Full Text | Google Scholar

Beetham, D. (2013). The Legitimation of Power. Basingstoke: Palgrave Macmillan. doi: 10.1007/978-1-137-36117-2

Crossref Full Text | Google Scholar

Brunkert, L. J. (2022). Overselling democracy-claiming legitimacy? The link between democratic pretention, notions of democracy and citizens' evaluations of regimes' democraticness. Front. Polit. Sci. 4, 880709. doi: 10.3389/fpos.2022.880709

Crossref Full Text | Google Scholar

Bühlmann, M., Merkel, W., Müller, L., and Weßels, B. (2012). The democracy barometer: a new instrument to measure the quality of democracy and its potential for comparative research. Eur. Polit. Sci. 11, 519–536. doi: 10.1057/eps.2011.46

Crossref Full Text | Google Scholar

Claassen, C., and Magalhães, P. C. (2022). Effective government and evaluations of democracy. Comp. Polit. Stud. 55, 869–894. doi: 10.1177/00104140211036042

Crossref Full Text | Google Scholar

Coppedge, M., Gerring, J., Altman, D., Bernhard, M., Fish, S., Hicken, A., et al. (2011). Conceptualizing and measuring democracy: a new approach. Perspect. Politics 9, 247–267. doi: 10.1017/S1537592711000880

Crossref Full Text | Google Scholar

Coppedge, M., Gerring, J., Knutsen, C. H., Lindberg, S. I., Teorell, J., Altman, D., et al. (2022). V-Dem Codebook v12. Gothenburg: Varieties of Democracy (V-Dem) Project.

Google Scholar

Dalton, R. J. (2004). Democratic Challenges, Democratic Choices. The Erosion of Political Support in Advanced Industrial Democracies. Oxford: Oxford University Press. doi: 10.1093/acprof:oso/9780199268436.001.0001

Crossref Full Text | Google Scholar

Easton, D. (1953). The Political System. An Inquiry into the State of Political Science. New York, NY: Knopf.

Google Scholar

Easton, D. (1965). A Systems Analysis of Political Life. New York, NY: Wiley.

Google Scholar

Easton, D. (1975). A re-assessment of the concept of political support. Br. J. Polit. Sci. 5, 435–457. doi: 10.1017/S0007123400008309

Crossref Full Text | Google Scholar

EVS (2022). EVS Trend File 1981-2017. Köln: GESIS. ZA7503 Datenfile Version 3.0.0.

Google Scholar

Fuchs, D., and Roller, E. (2018). Conceptualizing and measuring the quality of democracy: the citizens' perspective. Politics Gov. 6, 22–32. doi: 10.17645/pag.v6i1.1188

Crossref Full Text | Google Scholar

Habermas, J. (1976). “Legitimationsprobleme im modernen Staat,” in Politische Vierteljahresschrift, Special Issue 7, ed. P. G. Kielmansegg (Wiesbaden: Springer), 39–60. doi: 10.1007/978-3-322-88717-7_2

Crossref Full Text | Google Scholar

Habermas, J. (1984). “What does a legitimation crisis mean today? Legitimation problems in late capitalism,” in Legitimacy and the State, ed. W. E. Connolly (New York, NY: New York University Press), 134–155.

Google Scholar

Haerpfer, C., Inglehart, R., Moreno, A., Welzel, C., Kizilova, K., Diez-Medrano, J., et al. (2022). World Values Survey Trend File (1981-2022) Cross-National Data-Set. Madrid: JD Systems Institute and WVSA Secretariat.

Google Scholar

Kirsch, H., and Welzel, C. (2018). Democracy misunderstood: authoritarian notions of democracy around the globe. Soc. Forces 98, 59–92. doi: 10.1093/sf/soy114

Crossref Full Text | Google Scholar

Klingemann, H.-D. (1999). “Mapping political support in the 1990s: a global analysis,” in Critical Citizens: Global Support for Democratic Governance, ed. P. Norris (Oxford: Oxford University Press), 31–56. doi: 10.1093/0198295685.003.0002

Crossref Full Text | Google Scholar

Klingemann, H.-D. (2014). “Dissatisfied democrats: democratic maturation in old and new democracies,” in The Civic Culture Transformed: From Allegiant to Assertive Citizens, eds C. Welzel, and R. J. Dalton (Cambridge: Cambridge University Press), 116–157. doi: 10.1017/CBO9781139600002.010

Crossref Full Text | Google Scholar

Linde, J., and Ekman, J. (2003). Satisfaction with democracy: a note on a frequently used indicator in comparative politics. Eur. J. Polit. Res. 42, 391–408. doi: 10.1111/1475-6765.00089

Crossref Full Text | Google Scholar

Lipset, S. M. (1959). Some social requisites of democracy: economic development and political legitimacy. Am. Polit. Sci. Rev. 53, 69–105. doi: 10.2307/1951731

Crossref Full Text | Google Scholar

Macridis, R., and Cox, R. (1953). Seminar report. Am. Polit. Sci. Rev. 47, 641–657. doi: 10.2307/1952898

Crossref Full Text | Google Scholar

Marshall, M. G., and Gurr, T. R. (2020). Polity5. Political Regime Characteristics and Transitions, 1800-2018. Data Set Users' Manual. Vienna: Center for Systemic Peace.

Google Scholar

Mayne, Q., and Geissel, B. (2016). Putting the demos back into the concept of democratic quality. Inte. Polit. Sci. Rev. 37, 634–644. doi: 10.1177/0192512115616269

Crossref Full Text | Google Scholar

Merkel, W. (2004). Embedded and defective democracies. Democratization 11, 33–58. doi: 10.1080/13510340412331304598

Crossref Full Text | Google Scholar

Meuleman, B., Zółtak, T., Pokropek, A., Davidov, E., Muthén, B., Oberski, D. L., et al. (2023). Why measurement invariance is important in comparative research. a response to Welzel et al. (2021). Sociol. Methods Res. 52, 1401–1419. doi: 10.1177/00491241221091755

Crossref Full Text | Google Scholar

Mishler, W., and Rose, R. (2002). Learning and re-learning regime support: the dynamics of post-communist regimes. Eur. J. Polit. Res. 41, 5–36. doi: 10.1111/1475-6765.00002

Crossref Full Text | Google Scholar

Mudde, C. (2004). The populist zeitgeist. Gov. Oppos 39, 541–563. doi: 10.1111/j.1477-7053.2004.00135.x

Crossref Full Text | Google Scholar

Mudde, C. (2017). “Populism: an ideational approach,” in The Oxford Handbook of Populism, eds C. Rovira Kaltwasser, P. Taggart, P. O. Espejo, and P. Ostiguy (Oxford: Oxford University Press), 27–47. doi: 10.1093/oxfordhb/9780198803560.013.1

Crossref Full Text | Google Scholar

Müller, T., and Pickel, S. (2007). Wie lässt sich Demokratie am besten messen? Zur Konzeptqualität von Demokratie-Indizes. Polit. Vierteljahresschr. 48, 511–539. doi: 10.1007/s11615-007-0089-3

Crossref Full Text | Google Scholar

Munck, G. L., and Verkuilen, J. (2002). Conceptualizing and measuring democracy. Comp. Polit. Stud. 35, 5–34. doi: 10.1177/001041400203500101

Crossref Full Text | Google Scholar

Norris, P. (2011a). Cultural explanations of electoral reform: a policy cycle model. West Eur. Polit. 34, 531–550. doi: 10.1080/01402382.2011.555982

Crossref Full Text | Google Scholar

Norris, P. (2011b). Democratic Deficit: Critical Citizens Revisited. Cambridge: Cambridge University Press. doi: 10.1017/CBO9780511973383

Crossref Full Text | Google Scholar

Pickel, S., Breustedt, W., and Smolka, T. (2016). Measuring the quality of democracy: why include the citizens' perspective? Int. Polit. Sci. Rev. 37, 645–655. doi: 10.1177/0192512116641179

Crossref Full Text | Google Scholar

Repucci, S., O'Toole, S., Aaron, E., Buyon, N., Linzer, I., Meijer, D., et al. (2021). Freedom in the World 2020. The Annual Survey of Political Rights and Civil Liberties. Lanham: Rowman and Littlefield.

Google Scholar

Schaar, J. H. (1984). “Legitimacy in the modern state,” in Legitimacy and the State, ed. W. E. Connolly (New York, NY: New York University Press), 104–133.

Google Scholar

Tannenberg, M. (2022). The autocratic bias: self-censorship of regime support. Democratization 29, 591–610. doi: 10.1080/13510347.2021.1981867

Crossref Full Text | Google Scholar

Vanhanen, T. (1997). Prospects of Democracy: A Study of 172 Countries. New York, NY: Routledge.

Google Scholar

Weatherford, M. S. (1992). Measuring political legitimacy. Am. Polit. Sci. Rev. 86, 149–166. doi: 10.2307/1964021

Crossref Full Text | Google Scholar

Wiesner, C., and Harfst, P. (2022). Conceptualizing legitimacy: what to learn from the controversies related to an “essentially contested concept”. Front. Polit. Sci. 4:867756. doi: 10.3389/fpos.2022.867756

Crossref Full Text | Google Scholar