How People View Democracy:

Findings from Public Opinion Surveys in Four Regions

 

 

 By Larry Diamond

 

Presentation to the Stanford Seminar on Democratization.

 

January 11, 2001

 

 


Background:  The Growth of Comparative Survey Research on Public Opinion in New Democracies

 

            Over the past decade, one of the most important developments in the comparative study of democratization has been the growth of a major line of research on public opinion in democracies.  In one sense, this marks a resurgence of an earlier tradition of political culture research that began with the famous Almond and Verba book, The Civic Culture.   Building on this and many more recent studies of how individuals evalute, relate to, and participate in their political systems, the new wave of public opinion research has a number of practical and theoretical purposes.  One practical purpose is to generate a barometer of how mass publics evaluate the performance of their democratic systems, and to what extent they support democracy as a form of government.  These periodic measures of public opinion generate important raw empirical data for political scientists trying to assess the quality and stability of democracy in different countries (especially those that have recently experienced transitions or confront serious challenges to stability).  They also can provide important information for national and international policymakers about the public’s response to policy initiatives and institutional reforms, as well as directions and priorities for future reform. 

            At the same time, public opinion survey data is an invaluable source of information for assessing and extending theories of democratic consolidation.  One crucial dimension of consolidation involves norms and beliefs about the legitimacy of democracy, both in principle and as it is embodied in a particular regime.[1]  While partial inferences about regime legitimacy can be drawn from mass-level behavior, and from expressions of public sentiment in the mass media and elsewhere, there is no way that legitimacy can be adequately assessed and comprehended without nationally representative, public opinion survey data.  The additional value of such surveys is that they enable us to analyze the correlates and determinants of support for democracy, how legitimacy is shaped by socioeconomic status, by party and ideological orientations, by evaluations of economic and political performance, and a variety of other factors.  In short, public opinion surveys are opening up an unprecedented analytic window onto the study of the dynamics of democratic regimes, especially what fosters democratic consolidation as opposed to stagnation, instability, or even breakdown.

“Barometer” is increasingly the term that is used to describe these periodic (even annual) surveys.  For the countries of post-communist Europe, there is a “New Democracies Barometer” that has been administered by Richard Rose and his colleagues region-wide (in ten countries) five times since 1991.  There is also a “New Russia Barometer” and “A New Baltic Barometer” and a “New Korea Barometer” that employ many of the same questions measuring perceptions of and responses to both political and economic change.[2]For some time, a “Eurobarometer” has periodically measured political attitudes, values, and behaviors in the established democracies of western (including southern) Europe.  Many of those questions were incorporated into a “Latinobarometro,” which in early 2000 completed its fifth survey since 1995, now covering 17 countries in Latin America, from Mexico to the Southern cone.   This project is coordinated by Marta Lagos in Santiago, Chile.  The model of a regional barometer of attitudes toward political and economic change and performance in new democracies was then taken to Africa by Michael Bratton.  He has worked with a team of African researchers that includes Robert Mattes, based in South Africa at Idasa (which has a substantial Public Opinion Service that has been surveying attitudes and values in South Africa since the transition to democracy there).  The  resulting “Afrobarometer” has been administered or is being prepared for administration in roughly a dozen African countries, and if funding is available, the number of countries will be increased in the next few years.

            Ironically, despite its relative wealth and social science expertise, Asia has been the region of the developing world that has seen the least progress toward the development of a regional barometer of public opinion in new democracies.  This, however, is changing.  In July of 2000, social scientists from eight Asian countries gathered in Taiwan with leaders of the other regional barometers (Rose, Bratton, and Lagos) to begin designing the first systematically comparative regional survey of attitudes and values toward democracy in Asia.  This first “East Asian Barometer” of attitudes and values toward democracy and the performance of democratic regimes aims to administer a common core survey in the second half of this year in six East Asian democracies—Taiwan, Korea, Japan, the Philippines, Thailand, and Indonesia.  Much of the core survey will also be administered in Hong Kong and mainland China.  Ultimately, it is hoped that a “South Asian Barometer” can be joined to this project.[3]  The East Asian barometer is administered from Taiwan by a team led by professors Hu Fu and Yun-han Chu, with support from Taiwan’s ministry of education.

            This recent (post-1990) wave of comparative research on public opinion in (mainly) new or developing democracies is unprecedented in several respects.  First, there has never been such a comprehensive effort to measure how people view democracy and regime performance in so many emerging political systems.  Second, the regional surveys have repeated at fairly frequent intervals (typically every second or third year or so, but sometimes annually) many of the same precise questions, which generates an exceptional time series of data.  Third, within each region, the barometers have standardized the questionnaire wording and taken considerable pains to ensure a common meaning through “back translation” and preliminary testing of questions in the field.  This enables rigorous comparison of public attitudes and values across different countries with a region, and even pooling of the data across countries for broader regional assessments.  In almost every country, the surveys are also based on nationally representative random samples of at least 1000 respondents, which permits inferences within a relatively small margin of error about the overall state of public opinion, and which provides enough cases to analyze a variety of sub-groups.

To date, the problem has been the disarticulation among the different regional (and in some cases, isolated country) surveys.  Different regional surveys have used very different types of questions, or similar questions but different response formats, to measure common underlying concepts, such as legitimacy (or support for democracy), satisfaction with the performance of democracy, efficacy, institutional trust, and social capital.  Public opinion about politics and regimes (and no doubt, other matters) is highly sensitive to the way a question is worded, and even to the type of response structure that is offered.  (For example, levels of trust in different institutions will appear lower, possibly by a considerable margin, when respondents are offered a mid-point on an odd-numbered scale, signifying neither trust nor distrust, than when they have an even-numbered scale that forces them to express either trust or distrust).[4]  To overcome this handicap toward globally comparative research on new democracies, the coordinators of the different regional barometers are increasingly employing questions from the other regional surveys and seeking ways to standardize questionnaire wording and format on a number of themes.  Thus, while this endeavor will not have the central coordination and standardization of the World Values Survey, which now administers a common survey in more than 50 countries around the world, it will soon have a comparable reach in number of countries (some 50 or more overall).  And it will generate numerous opportunities for meaningful and rigorous comparative analysis across regions.

The remainder of this essay will examine the levels and trends in public opinion about democracy in three regions—postcommunist Europe, Latin America, and Africa—plus two countries, Korea and Taiwan, where recent surveys represent the beginning of a fourth regional barometer. This discussion is based on a cluster of four articles (one on each of these regional groupings) published in the January 2001 Journal of Democracy under the title, “How People View Democracy,”[5] as well as supplementary data and analysis from the different regional and country survey projects.[6]

 

 

LEGITIMACY AND SUPPORT FOR DEMOCRACY

 

 

            In terms of its implications for democratic stability and consolidation, the core dimension of public opinion concerns legitimacy, or support for democracy.  Ideally, this should be measured both as a general principle and in specific reference to the system in place and to alternative systems that people could imagine.  To what extent do the publics in these countries appear committed to democracy as the best form of government for their society?

 

The Postcommunist States

 

Unfortunately, the data from postcommunist Europe are not comparable with those from the other regions, as different questions are employed.  Nevertheless, we can make two crucially important comparisons: over time, within each country and the region as a whole, and between postcommunist countries, especially Eastern Europe vs. the former Soviet countries.  We can begin by examining the extent to which the publics in these countries give a positing rating to the working of two different systems:  “the former communist regime” and the  “current system of governing with free elections and many parties.”  Because these two questions pertain to “how our system of government works,” they straddle the boundary between support for democracy and satisfaction with the way democracy works.  Thus, they are not purely a measure of regime support.  Nevertheless, they do illuminate the overall public view of democracy and provide one indication of system support.

            Two points merit emphasis in reviewing the data in Tables 1 and 2, on approval of the current system and the former communist one.  First, as is the case on so many different measures consistently over time, the publics in Central and Eastern Europe (CEE) are much more supportive of democracy than are the former Soviet publics, in this case Russia, Ukraine, and Belarus.  Whereas strong majorities, typically over 60 percent, in each of the seven CEE countries have approved of the way the current multiparty system works, the overall percentage in the post-Soviet heartland has never been much more than a third.  And the most recent figure has increased mainly because of the rise in Belarus, which is now a very authoritarian state, making it difficult to interpret just what it is that the Belarussian public is approving.  At the same time, retrospective approval of the previous communist system is much higher in the post-Soviet states than in the CEE ones—by an average of 30 percentage points in the most recent survey.  And the difference would be even greater if Hungary were set aside, given that its much softer “goulash” communism does not evoke the same repressive, authoritarian memories that the other systems do.  Overall, more than two-thirds of these post-Soviet publics look back favorably (perhaps wistfully) on the old Communist regimes.[7]

            The other point that stands out from these two tables is the recent downward trend in the CEE states in approval of the democratic system’s performance.  In most of these states, there was a perceptible dip between 1995 and 1998 in approval of the current regime.  This is likely due to protracted economic difficulties in many countries, such as Romania (where approval of the current economic system in 1998 was only 23 percent, compared to 40 percent in Hungary and Slovenia and 61 percent in Poland).  When people were asked to compare the “overall economic situation of your household” currently (in 1998) with that before the fall of communism, an average of 56% in the seven CEE states (about the same as Russia) but 79% in Belarus and 90% in Ukraine thought the past was better.[8]   The other factor that may be depressing approval is the widespread perception of corruption in the region.  “Across postcommunist Europe, Barometer surveys find an average of 72 percent believe that their new regime is more corrupt than its predecessor;” only six percent perceive a reduction in corruption, and in every postcommunist country surveyed, a majority of the public believes the national government is corrupt.[9]  As we will see, these proportions bear a striking resemblance to the views of Latin American publics and epitomize the general growth of cynicism about politics and politicians in all contemporary democracies.

Two others questions are particularly useful for evaluating public support for democracy in these countries.  One examines legitimacy in the essential comparative context: to what extent would people support an alternative, undemocratic regime?  Four alternative non-democratic regime types have been presented to the postcommunist publics in the New Democracies Barometers over the years: a return to Communist rule, rule by the army, closing down parliament in favor of “strong leader who can decide things quickly,” and a return to monarchy.[10]  Rejection of all of these authoritarian alternatives can be read as rather robust support for democracy as a form of government.  Such robust support is lacking in Russia, Belarus, and Ukraine, where, in 1998, an average of only 36 percent of the public opposed all authoritarian alternatives (Table 3).[11] Once again, the difference with the European postcommunist states is striking.  In the latter, two-thirds of the public oppose all authoritarian alternatives (as many as three-quarters do so in the Czech Republic and Slovenia).  There is not a single postcommunist European state, including Croatia and Yugoslavia (Serbia and Montenegro), where the return of Communist rule is opposed by less than seventy percent of the public.  By contrast, only half do so in Ukraine and 59 percent in Russia.[12]

            Another indication of regime support is rejection of the idea of suspending parliament and abolishing political parties.  This proposition has been steadily rejected by every postcommunist public save for Ukraine’s (Table 4).  However, support for democracy by this measure is once again much stronger in CEE, where across the five surveys it has consistently averaged over three-quarters of the public. 

As Richard Rose shows, the crucial divide among the postcommunist states is not purely between Europe as opposed to the former Soviet Union.  Rather it is between those states with levels of civil and political freedom (as assessed annually by Freedom House) roughly comparable to the European Union and those that lag well behind.  Among the ten countries rated “free” by Freedom House, an average of 63% reject all authoritarian alternatives.  Among the six others, only 43% do so (see Table 1 of Richard Rose’s article).[13]  A similarly large gap exists between the two sets of countries in approval of the current system (as measured by the question in Table 1).  Among the “free” countries, an average of 55% approve, while among the less democratic countries, only about one-third approve.  This is indicative of a more general problem.  “The great obstacle to the completion of democracy in postcommunist Europe is the absence of the rule of law.”[14]  This absence is particularly glaring in the less-than-free states, which are also the ones with the highest external perceptions of corruption, as measured by Transparency International. 

Thus, as Rose demonstrates, there are really two sharply diverging sets of postcommunist states emerging.  One set is in and of Europe, and looks to integration into a broader definition of what is Europe, through an expansion of the EU (and also NATO).  This set, which very much includes the three Baltic states, is more free and generally less corrupt.  The other set contains the states that are slipping backwards, away from democracy and the rule of law.  These are the more repressive and corrupt of the postcommunist states, where democracy is increasingly shallow and beleaguered, or gone altogether (in Belarus).  It is important to underscore that the problem is not simply a divergence between electoral democracy and the other dimensions of liberal democracy (such as freedom and the rule of law).  The growing weakness of liberty and accountability, and increasing centralization of power in countries such as Russia and Ukraine is exacting a heavy toll on the electoral dimension of democracy as well, raising doubts as to just how competitive and fair elections are.  With the recent defeat of the quasi-dictator Slobadan Milosevic in the 2000 presidential election, Serbia shows signs of moving away from this group of democratic laggards, toward Europe, as Croatia began doing two or three years before.  But the other post-Soviet electoral regimes, from Georgia and Azerbaijan to Kyrgyzstan, remain stuck to one degree or another in this ambiguous status between democracy and authoritarianism.  There is, of course, a third group of post-Soviet states, primarily in Central Asia, that no longer even seriously attempt to appear democratic.

 

Latin America

Within Latin America as well, there is significant variation in support for democracy.  The key measure used here—whether “democracy is preferable to any other kind of government” or sometimes “an authoritarian government can be preferable”—was borrowed from the Eurobarometer.  It has now been used in numerous surveys around the world, not only in Western Europe but also in Korea, Taiwan and sub-Saharan Africa.  Several features of the time series data for Latin America are noteworthy.[15]  First, only two small countries have levels of support for democracy comparable to Spain, Portugal, and Greece, and indicative of firm democratic consolidation.  These are Costa Rica and Uruguay, where support for democracy exceeds 80 percent and the willingness to entertain the authoritarian option is at 10 percent or less.  The only other country that consistently shows 70 percent-plus support for democracy is Argentina.  By this measure, some other countries have made progress or appear reasonably supportive, but others have regressed.  Particularly striking are the very low level of support for democracy in Brazil; the rather high levels of consideration for authoritarian rule in Mexico, Paraguay, Brazil, Colombia, and Venezuela; and the deterioration in support for democracy in Colombia (Table 5).

The above data are best analyzed in conjunction with responses to the question whether there can be democracy without a national congress.[16]  Uruguay, Argentina, and Costa Rica (now joined by El Salvador) have the most democratic responses to this question, but Chile and a few other countries also have quite democratic orientations, with two-thirds or so of the public saying there can be no democracy without a national congress.  The low appreciation of the importance of a Congress in Venezuela, and the precipitous decline since 1997 (from 51 to 29 percent) helps to illuminate the crisis of democracy in that country, where President (former Colonel) Hugo Chavez has appeared increasingly bent on constructing a personalistic, Peronist type of pseudodemocracy.  The low recognition of the need for a congress in Ecuador (29%) and Colombia (39%) also reflects the crisis of democracy in the Andean region.  As Lagos notes, the increase in the percentage of Colombians (from 38 to 45%) who think there can be democracy without a congress, “and the parallel rise in the number of Colombians who believe that there can be democracy without parties (from 38 to 46 percent), provide further evidence of growing disillusionment with established democratic institutions.”[17] Interestingly, in the one Andean country that has really had a quasi-dictator for most of the past decade, Peru, nearly two-thirds of the public have repeatedly embraced the importance for democracy of a national congress (Table 5).  Overall in Latin America, the belief that “without a National Congress there can be no democracy” has declined six points in the last three years (from 63 to 57 percent), without a corresponding drop (62 to 57 percent) in the belief that democracy requires political parties.

The malaise of democracy in the Andean region and in some other parts of Latin America is driven by dissatisfaction with the way democracy is working, and underlying that, concern about corruption, poor economic performance, and a general lack of responsiveness on the part of politicians.  One response that has remained virtually unchanged over the past four years in Latin America is the assessment of the trend with respect to corruption.  When asked whether it has increased a little or a lot, remained the same, or decreased a little or a lot in the last 12 months, an astonishing 75% of Latin Americans said in 2000 that corruption has increased a lot.  Another 10% said corruption has increased a little and less than five percent perceived any decline.[18]  Even in countries with strong support for democracy, such as Costa Rica, Uruguay, and Argentina, most people think corruption has increased a lot.  Interestingly, the country with the least extreme perception was Venezuela the percentage believing that corruption has increased a lot declined from 94% in 1998 to 54% in 2000, with the coming to power of the populist Hugo Chavez.   As Richard Rose noted for the postcommunist states, much of this perception may be the ironic consequence of greater freedom to report political wrongdoing, and of press sensationalism.

As for economic performance, several countries remained mired in economic crisis, and even where the economy is growing overall (as in Brazil) severe problems of inequality leave huge swaths of the population excluded from its benefits.  The disgust with democratic politicians in Ecuador, where indigenous organizations and sections of the military joined to seize power briefly in January 2000, can be attributed at least in part to severe economic crisis.  In 1999, severe budgetary cuts to finance international debt service caused social spending to fall by half in real terms; the minimum salary fell by a quarter while unemployment doubled in a country that already had the third worst distribution of income in the region.[19]

The political malaise in Latin America is vividly conveyed by the trends in satisfaction with the way democracy is working in each country.  Between 1995 and 2000, satisfaction has declined from 62 to 18% in Brazil, from 71 to 35% in Chile; from 75 to 24% in Peru; and from 81 to 55% in Venezuela (where it sunk as low as 35% in 1998).  Overall, satisfaction with democracy in the region has risen from its nadir of 27% in 1996, but at an average of 37% it still remains quite low.  Only in three countries were majorities of the Latin American public satisfied with the working of democracy in 2000, and the extremely low levels of satisfaction in Brazil (18%), Ecuador (23%) and Paraguay (12%) are indicative of a crisis of democratic confidence.

 

East Asia and Africa

            The Latin American data needs to be appreciated in a comparative context.  On the one hand, support for democracy is, on average, higher in Latin America than in Korea and Taiwan, where the most recent data display a nearly identical average of around 54% (Table 5).  Particularly striking is the substantial proportion of Koreans (almost one-third in 1998 and 1999) who say that sometimes an authoritarian regime may be preferable.  On the other hand (as previously noted), the levels of support for democracy in most Latin American countries are well below those recorded for Spain, Portugal and Greece.  By around 1990, support for democracy in Western Europe varied between 75 and 92% on this standardized question of whether democracy is always the best form of government.[20]  Moreover, while democratic support appears lower in Korea and Taiwan, satisfaction is higher, particularly in Taiwan where three in five say they are satisfied (Table 7).

            The data from Korea and Taiwan paint an ambiguous picture.  Neither democracy is in danger of collapse.  Indeed, each system is considerably more secure than the democracies of the Andean region, for example.  But as we see in Table 5, support for democracy in Korea has declined significantly in the wake of the financial crisis and a string of high-profile corruption scandals.  On some other measures as well that are unique to these two countries, such as the 10-point scale of democratic suitability, support for democracy has declined perceptibly in the last three years in Korea, and is outpaced by the level of support for democracy in Taiwan.  Koreans are torn.  They reject authoritarian alternatives about as often as the more democratic postcommunist publics, yet when forced to weigh democracy against economic security, their commitment wavers much more.  And about seven in ten Koreans seem willing to tolerate illegal actions by the president in a crisis.

            Even more striking by way of comparison are the African data.  Of the five African countries which Bratton and Mattes report on in their Journal of Democracy article, four of them show levels of support for democracy (as “preferable to any other form of government”) higher than the Latin American average, and at 57% Namibia is only slightly lower.  In the context of its long history of stable democratic functioning without interruption, the 82% support level in Botswana is evidence of democratic consolidation.  The same level (81%) in Nigeria more likely reflects the broad revulsion with predatory military rule and the euphoria of the transition back to civilian, democratic government that had been completed only about six months after the survey was conducted.  The 71% figure in Zimbabwe may also be a statement of opposition to Robert Mugabe’s repressive and corrupt regime, but it helps to explain the strength of the democratic opposition. The even higher figure in Ghana (76%) can now be read as a harbinger of the extraordinary opposition victory in the December 2000 presidential election.  Even among those who say that democracy is “not working,” “democracy is supported by large majorities in Zimbabwe (74%) and Botswana (65%) as well as by substantial majorities in Malawi (59%) and Zambia (54%).”[21]

            The high levels of support for democracy in these African countries cannot be dismissed as deriving from a vacuous or specious understanding of democracy.  As Bratton and Mattes show, large majorities in each country (overall, three-quarters) are able to attach a meaning to democracy, and most of these (about seven in ten) define democracy in terms of political procedures, not substantive outcomes.[22]  Indeed, “popular African conceptions of democracy are, perhaps unexpectedly, quite liberal,” with the open-ended responses citing “civil liberties and personal freedoms more frequently than any other meanings (34 percent).”[23]  Where democracy has been in place by far the longest, in Botswana, the identification of democracy with civil and political rights is most frequent (55%). 

            The Afrobarometer is the first regional survey to employ both the standard support for democracy measure and the innovative item—support for specific authoritarian alternatives—developed by Richard Rose for the study of postcommunist states.  On this measure as well, large majorities in five of the six countries reject authoritarian alternatives such as military and one-party rule, but such opposition is somewhat weaker in Namibia.  Interestingly, except in Zimbabwe—where there has been no democracy for some time—satisfaction with the way democracy works is much higher (from 54 to 84%) than in Latin America. And this may be one reason one why support for democracy is so high.  Bratton and Mattes find that in Africa (as elsewhere) “popular support for democracy has a strong instrumental component.  Citizens extend support to a democratic regime in good part because they are satisfied with its performance in delivering desired goods and services.”[24]  Yet most Zimbabweans support democracy (59%) even though they are dissatisfied with the way it works—in fact, because they perceive it to be largely lacking, and appreciate its importance.  And overall, one in five Africans surveyed support democracy despite dissatisfaction with its performance.  Moreover, asked if they would prefer democracy or a “strong leader who does not have to bother with elections” when “democracy does not work,” substantial majorities in several countries still choose democracy as “always best.”[25]

 

Comparative Trends in Institutional Trust

Everywhere, democratic publics are less satisfied with the performance of democracy than they are committed to it as the best (or lesser evil) form of government.  And some democracies, notably Italy, have mass publics that are chronically cynical and dissatisfied.  However, cynicism is one thing in a consolidated democracy; it may affect the character and quality of democracy, but not necessarily its viability.  Where democracy is still a young and somewhat fragile plant, lacking other attributes of consolidation, chronic dissatisfaction and alienation can be an obstacle to consolidation.  Indeed, it may well be the case that different dimensions of cynicism--namely, dissatisfaction, alienation (low system efficacy), and institutional distrust--inhibit the growth of democratic legitimacy.

One of the most widely tested and theoretically important dimensions of system evaluation is public trust or confidence in political institutions.  The data presented here are not entirely standardized, in that the Taiwan survey used an 11-point scale while the Korea survey used a four-point Likert scale (the most common response structure for this item across countries).  In his surveys of the new democracies of postcommunist Europe, Richard Rose used a 7-point scale.  The Taiwan and postcommunist data are thus the most comparable, since both have odd-numbered scales with a neutral mid-point.   For the most part, citizens in Korea and Taiwan appear to have middling levels of trust in democratic institutions.  Levels of trust in Taiwan vary only modestly across institutions, ranging from about 40 to 55%.  In Korea, they range much more widely: while almost three-quarters of Koreans trust the military (on which they rely heavily for defense against an ever-threatening North Korean regime), only one in five trust their respective legislative bodies (Table 8).  The overall result is similar, however. 

While levels of trust in public institutions appear quite modest in Korea and Taiwan, they are better, or at least no worse, than in the post-Communist states.  Trust in their legislative body is just as low in the post-Communist world as in Korea, but Koreans generally have more trust in other public institutions.  One may argue, however, that the post-Communist states represent a low standard of comparison, given the cynicism about the state and the party left from the Communist era.  Yet trust in the national legislature is also very low in Chile, Brazil, and Argentina, as well as Japan (Table 8).  And overall levels of trust in political institutions are not very high in West European democracies, either.  In fact, on several dimensions, the public in Taiwan is at least as trusting, if not more so.

 

Some Concluding Thoughts

 

            What do we learn from this growing accumulation of data?  I will not attempt here to provide any kind of comprehensive or definitive conclusions, but will offer just a few reflections.

            First, despite the extraordinary outpouring of data over the past decade, the comparative study of how mass publics in emerging democracies view and value their institutions is only now emerging into a more mature phase.  There has been a good deal of innovation—but still not enough—in the design of questions to capture complex and subtle distinctions in public attitudes and values.  The need for experimentation with new questions in order to probe deeper, or at least differently, is unfortunately in tension with the need for standardization longitudinally as well as cross-nationally, in order to facilitate comparison.  Standardization presses us to go with what we have, at the risk of some ossification of our understanding.  Experimentation and innovation promise either to generate new important new insights into how the public views democracy, or at least to enhance the reliability of the understanding we derive from our existing measures. The latter would not be a trivial achievement.  Both goals—standardization and innovation—are in fact vital if we are going to take the comparative study of democracy to a new level of sophistication and value added.  In principle, within the space of a questionnaire that can probably only accommodate about 100 questions overall before the typical respondent loses focus or breaks away, it should be possible to include both old questions and new ones about legitimacy, satisfaction, trust, and so on.  The problem is a practical one.  Funding a survey often means satisfying diverse constituencies, not just in the research community but among government agencies and commercial enterprises that may be funding all or part of the survey.  If a survey is to become a true barometer, it must be repeated often, and that will require raising funds from sponsors who have other interests—for example in topical issues and pressing policy questions such as drug trafficking and poverty.  These may sometimes generate valuable independent variables whose association with the democracy measures is well worth exploring, but inevitably, it reduces space available for democracy questions.  Probably the only alternative is to rotate some questions in and out over time.  The collaborative team that is now designing the East Asia Barometer has faced many painful trade-offs as a result of the diversity of intellectual items on the research agenda and the need to encompass a sufficiently wide range of issues to appeal to potential funding sources in some countries.  No doubt, the questionnaire that has been agreed upon for the first survey will be revised in the future if indeed a true periodic “barometer” can be funded.

            A related point has been made at some length in my book and I will not belabor it here.  But it is worth reiterating that we cannot generate reliable cross-national comparisons unless the same questionnaire wording and format is used.  We now have two different strategies, for example, for measuring legitimacy. One is to ask directly whether democracy is the best form of government. Another is to inquire indirectly by assessing support for nondemocratic alternatives.  Our understanding would advance much more if each regional comparative survey employed both strategies, as the Korea Barometer began to do some years ago and the Afrobarometer now does.  Legitimacy is the irreducible phenomenon that must be measured in any attempt to understand democratic progress, and the cost of two or three additional questions on the subject is well worth bearing.  Experimentation on this dimension is particularly necessary and welcome.

            Both the Korea Barometer and the Afrobarometer have devised new questions to get at one of the most subtle and elusive issues in the study of democratic regime legitimacy.  When can support for democracy be assumed to be intrinsic rather than instrumental or simply the response that seems to be appropriate and expected by society, and by the interviewer?  How can we press respondents to reveal their true feelings, fears, anxieties, and inhibitions about regime alternatives?  One way is to pose for them plausible hypotheticals, and then, by carefully finding neutral language, to give them normative space to give a nondemocratic response if that is what they feel.  This is what the New Democracies Barometer does in posing the regime alternatives.  But we can and should press further by pitting competing values against one another, as the World Values Survey has done for many years.  Thus Doh Chull Shin has asked, in various ways, respondents to indicate which value is more important, economic development or democracy.  The tension between freedom and order also needs more exploration. 

The Afrobarometer employs a question that Robert Mattes and his team first developed for their surveys of South Africa.

Sometimes democracy does not work.  When this happens, some people say that we need a strong leader who does not have to bother with elections.  Others say that even when things don’t work, democracy is always best.  What do you think? Which statement do you agree with most: Need strong leader, or Democracy is always best?

The new information this question generates is quite significant.  While 83% of the public in Botswana say democracy is preferable to any other kind of government, in response to the standard question, support for democracy when it “does not work” declines to 65%.  Support for democracy under this scenario also declines in most other African countries surveyed.  Yet this question may give us a better, more revealing indication of the rock-bottom support for democracy than does the more abstract item that has been used for some time.  And it may also bring some surprises.  By the standard measure, 71% of Zimbabweans support democracy, but the figure actually rises slightly (to 74%) when the question invokes the prospect of a strong leader who does not bother with elections (Table 9).[26]  This may be because Zimbabweans by now have had their fill of a strong, abusive leader such as Robert Mugabe.

            Substantively, two conclusions merit mention.  One is that the road to democratic consolidation is longer and more complicated than was often blithely assumed a decade ago when many of these new democracies were taking shape.  There is no imminent threat to democracy in Korea and Taiwan, and no democracy has ever broke down in a society even approaching their level of per capita income.  However, the survey data demonstrate that at the level of mass public beliefs, values, and evaluations, democracy has yet to be consolidated in either country, and indeed has probably moved in the reverse direction in Korea.  One need only look to the political crises and stalemates in each country, the inability to find a sustainable working relationship between executive and legislature, government and opposition, to find confirmation of the cautionary tone of our survey findings.  Elsewhere, most of Latin America is stuck somewhere between democratic stability and crisis.  In Central and Eastern Europe it is mainly an external factor—the enormous gravitational pull of the EU, with its clear political conditionality—that is propelling these systems toward consolidation, though not without some signs of regression, anxiety, and institutional deformity.  In a number of countries, public opinion is much more supportive of democracy than are the elites or the institutional actors.  This is particularly so in Africa, where, with the exception of Botswana, the high levels of public support for democracy do not indicate consolidation.  Yet these levels do constitute a positive sign, and in Ghana they may well herald a new era of sustainable democracy.

            Second, and finally, the more recent evidence from the studies and data analyzed here continues to point toward the crucial importance of regime performance in fostering or stunting the growth of democratic legitimacy.  We are still a very long ways from being able to determine very clearly and satisfactorily what generates sustainable support for democracy.  But the recent data add to our sense that both economic and political performance matter.  A sharp economic downturn can diminish support for democracy when misrule appears to be the cause, at least in part.  And people will blame the government for bad economic times, whether it deserves that blame or not. The key is whether they will go on to blame the system of government.  The capacity for throwing the incumbents out has eclipsed that judgement for some time in many countries.  However, the declines in democratic support in Colombia, Venezuela, and elsewhere Latin America suggest that if economic difficulties go on for long enough—and especially if they are deemed to derive from broader problems of governance, such as corruption and abuse of power—the system of government will begin to be blamed. 

            This raises again the relationship between political performance and democratic legitimation.[27]  Causal analysis of the Korea and Taiwan data suggests that political factors are more important than economic ones in shaping support for democracy.  The more satisfied people are with the way democracy works, the more they trust key political institutions, and in particular the more they perceive the system to be democratic, the more likely they are to support democracy.  The African survey data as well contains some indications that perceptions of the degree to which the system is functioning democratically may be important in affecting support for the political system.  And Rose and his colleagues have shown the perception of greater freedom to be a significant determinant of support for postcommunist democracies, while political factors in general outweigh economic ones.[28]

            Freedom is important but it is not sufficient.  If democracies are to be seen to deliver on the political promise of democracy, they must deliver a rule of law in a second sense as well, not just permitting citizens to express themselves and live their lives, but also restraining government officials from abusing power.  We do not have nearly enough data specifically measuring how citizens perceive corruption and what factors shape their perceptions.  The data we have demonstrate widespread and growing cynicism about corruption in public life, even though actual levels of corruption in some countries may well be lower than when the press was less free to report it and the institutions of investigation and horizontal accountability less active.  If democracy is going to be consolidated at the level of mass public beliefs and values, there will need to be much more dramatic progress in controlling corruption in politics and government and improving responsiveness and accountability more generally.  The growing accumulation of public opinion data only reinforces this fundamental point: the challenge of democratic consolidation remains substantially one of providing effective democratic governance.


Table 1

Support for the Current Regime in Postcommunist States

 

Percentage of the public expressing approval

Country

1991

1992

1993

1995

1996

1998

Czech Republic

71

71

78

77

 

56

Slovakia

50

58

62

61

 

50

Hungary

57

43

51

50

 

53

Poland

52

56

69

76

 

66

Slovenia

49

68

55

66

 

51

Bulgaria

64

55

58

66

 

58

Romania

69

68

60

61

 

55

CEE average

59

60

61

65

 

57

Russia

 

14

36

26

28     38

36

Belarus

 

35

29

35

 

48

Ukraine

 

25

24

33

 

22

RBC average

 

25

30

31

 

35

Sources:  William Mishler and Richard Rose, “Five Years after the Fall: Trajectories of Support for Democracy in Post-Communist Europe, Studies in Public Policy 298, Center for the Study of Public Policy, University of Strathclyde, 1998, Table2; Richard Rose and Christian Haerpfer, “New Democracies Barometer V: A 12-Nation Survey” Studies in Public Policy 306, pp. 49-50; and Richard Rose, “New Russia Barometer Trends Since 1992,” Studies IN Public Policy 320, Table 4.1.

 

Note:  CEE indicates the above seven countries of Central and Eastern Europe.  RBC indicates Russia, Belarus, and Ukraine.


Table 2

Support for the Previous Communist Regime

 

Percentage of the public expressing approval

Country

1991

1992

1993

1995

1996

1998

Czech Republic

23

29

23

24

 

31

Slovakia

44

48

50

52

 

46

Hungary

51

67

58

66

 

58

Poland

34

42

38

25

 

30

Slovenia

41

40

32

36

 

42

Bulgaria

30

41

51

58

 

43

Romania

26

35

33

28

 

33

CEE average

35

43

41

40

 

41

Russia

 

50

62

67

59     60

72

Belarus

 

60

64

77

 

60

Ukraine

 

55

55

75

 

82

RBC average

 

55

60

73

 

71

Sources: See Table 1.

 

Table 3

Reject all Authoritarian Alternatives

Country

1991

1992

1993

1995

1996

1998

Czech Republic

 

 

79

82

 

75

Slovakia

 

 

67

70

 

57

Hungary

 

 

72

69

 

68

Poland

 

 

57

63

 

67

Slovenia

 

 

58

68

 

78

Bulgaria

 

 

44

55

 

56

Romania

 

 

60

61

 

60

CEE average

 

 

62

67

 

66

Russia

 

 

 

39

58        46

39

Belarus

 

 

35

30

 

46

Ukraine

 

 

37

23

 

24

RBC average

 

 

 

 

 

36

Sources:  see Table 1.


 

Table 4

Reject Suspension of Parliament

 

                 Percentage of public disapproving suspension

Country

1991

1992

1993

1995

1996

1998

Czech Republic

88

78

82

75

 

79

Slovakia

85

81

76

78

 

77

Hungary

75

75

70

74

 

83

Poland

67

57

71

68

 

81

Slovenia

85

89

n.a.

81

 

74

Bulgaria

79

75

72

78

 

77

Romania

90

81

76

88

 

73

CEE average

81

77

75

79

 

78

Russia

 

 

n.a.

61

66    65

63

Belarus

 

 

57

60

 

72

Ukraine

 

 

44

39

 

55

RBC average

 

 

51

53

 

63

            Sources:  See Table 1


Table 5

Democratic Legitimacy

Responses in Percentages

 

Country, Year

Democracy is Always Preferable

Sometimes Authoritarianism is Preferable

It Doesn’t Matter to People Like  Me

Taiwan 1998

54

12

17

Korea, 1999

            1998

            1997

            1996

55

54

69

65                        

30

31

20

                  17

15

15

11

               10

Spain    1995

             1992

             1985

79

78

70

9

9

10

8

7

9

Portugal 1992

              1985

83

61

9

9

4

7

Greece   1992

              1985

91

87

4

5

3

6

Latin America average            

                2000

                1996

 

60

61

 

17

17

 

17

17

Costa Rica 2000

                   1996

83

80

7

6

 

Uruguay  2000

                1996

84

80

9

9

6

8

Argentina 2000

                 1996

71

71

16

15

11

11

Chile        2000

                 1996

57

54

19

19

22

23

Brazil       2000

                 1996

39

50

24

24

28

21

Paraguay  2000

                 1996

48

59

39

26

13

13

Venezuela 2000

                  1996

61

62

24

19

10

13

Colombia  2000

                  1996

50

60

23

20

20

18

Sources for comparative data:  Marta Lagos, “The Latinobarometro: Media and Political Attitudes in South America.”  Paper presented to the 1996 Annual Meeting of the American Political Science Association, August 29-September 1, San Francisco; Informe de Prense: Latinobarometro 1999/2000, Santiago, Mayo 2000; José Ramón Montero, Richard Gunther, and Mariano Torcal, “Democracy in Spain: Legitimacy, Discontent, and Disaffection.”  Estudio/Working Paper 1997/100, June 1997, Center for Advanced Study in the Social Sciences, Juan March Institute, Madrid, Spain.   Additional data on Latin America were provided directly by Marta Lagos.


Table 6

Democratic Support and Satisfaction in Latin America

 

 

No Democracy without a National Congress  (percent agreeing)

Satisfaction with Democracy (percentages)

South America and Mexico

1997

2000

1996

1998

2000

Argentina

74

70

34

49

46

Bolivia

62

54

25

34

22

Brazil

50

47

20

27

18

Columbia

46

39

16

24

27

Chile

73

68

27

32

35

Ecuador

45

29

34

33

23

Mexico

65

64

11

21

36

Paraguay

58

61

22

24

12

Peru

64

64

28

17

24

Uruguay

74

79

52

68

69

Venezuela

51

29

30

35

55

Average

60

54

27

34

35

Central America

 

 

 

 

 

Costa Rica

81

72

51

54

61

El Salvador

73

72

26

48

27

Guatamala

57

54

16

57

36

Honduras

73

67

20

37

43

Nicaragua

75

68

23

27

17

Panama

52

42

28

34

47

Average

68

63

57

43

39

LA Average

63

57

27

37

37

           Source:  Latinobarometer.


Table 7

Satisfaction with the Way Democracy Works

 

                                      Percent satisfied with the Way Democracy Works*

Country

1985-91

1993

1995

1996

1997

1998

1999-

2000

Korea

 

 

 

 

49

46

45

Taiwan

 

 

 

60

 

 

59

 

EC avg., 1976-1991

EU avg., 1997-1999

57

 

 

 

 

 

 

53

 

Spain

58

39

41

57

 

 

 

Portugal 

63

54

40

 

 

 

 

Greece  

56

45

28

 

 

 

 

Latin America Avg.

 

 

 

27

41

37

37

Uruguay                  

 

 

 

52

54

68

69

Argentina                

 

 

 

34

42

49

46

Chile                       

 

 

 

27

37

32

35

Brazil                       

 

 

 

20

23

27

18

Venezuela               

 

 

 

30

35

35

55

Colombia                 

 

 

 

16

40

24

27

Peru                         

 

 

 

28

21

17

24

Costa Rica               

 

 

 

51

68

54

61

Eastern Europe avg.

 

 

 

36

 

 

 

Czech Republic 1991

35

53

 

41

 

 

 

Hungary            1991

34

29

 

22

 

 

 

Poland               1991

35

26

 

48

 

 

 

Romania

 

 

 

52

 

 

 

Bulgaria         

 

 

 

6

 

 

 

Slovakia

 

 

 

21

 

 

 

Russia                 

 

 

 

8

 

 

 

Sources:  see Table 1.  Also, Dieter Fuchs, Giovanna Guidorossi, and Palle Svensson, “Support for the Democratic System,” in Hans-Dieter Klingemann and Dieter Fuchs, eds., Citizens and the State (New York: Oxford University Press, 1995), p. 341, table 11.4; Larry Diamond, Developing Democracy: Toward Consolidation (Baltimore: Johns Hopkins University Press, 1999), Table 5.4, pp. 180-181, and Richard Rose and Christian Haerpfer, “New Democracies Barometer V: A 12-Nation Survey,” Centre for the Study of Public Policy, University of Strathclyde, Studies in Public Policy, no. 306, pp. 50-51; Gabor Toka, “Political Support in East-Central Europe,” in Klingemann and Fuchs, Citizens and the State, pp. 364-365, Table 12.3; Hans-Dieter Klingemann, “Mapping Support in the 1990s: A Global Analysis,” in Pippa Norris, ed., Critical Citizens: Global Support for Democratic Governance (Oxford: Oxford University Press, 1999), Table 2.10, p. 50.

 

* Surveys in Korea (1997 and 1999) used a ten-point scale (1-10) of democratic satisfaction.  The 1999 Taiwan survey used an 11-point scale (0-10) of democratic satisfaction.  Responses of 6 or higher (with 10 or 11 being highest) indicate satisfaction.  All other surveys (including Taiwan, 1996) used a four-point Likert scale, and the percentages indicated those who are “somewhat” (or “fairly”) or “very” satisfied.


Table 8

Trust or Confidence in Institutions

 

                        Percent Expressing Trust or Confidence in Specific Institutions

Country

Courts

Police

Military

Legisla-ture

Civil Service

Parties

 

Press

Presi-dent

Korea 1997

57

42

72

22

45

20

 

 

Taiwan 1998

43

49

53

42

48

 

45

54

EC 14-state avg,  1981

        1990                                      

 

64

57

 

75

74

 

58

50

 

50

48

 

45

44

 

n.a.

n.a.

 

33

35

 

Post-Communist

States (11) avg*, 1997-98

29

30

49

22

26

13

35

44

Czech Repub

25

29

31

15

27

15

48

60

Hungary

39

35

40

25

32

11

42

53

Poland

30

32

53

25

28

9

42

40

Slovenia

29

34

34

20

34

11

42

45

Romania

40

41

76

31

50

19

46

53

Bulgaria

19

27

54

21

18

13

27

70

Russia

24

18

34

13

n.a.

7

22

14

South Africa

 

 

 

60

 

 

 

 

Chile

 

 

 

38

 

 

 

 

Brazil

 

 

 

34

 

 

 

 

Japan

 

 

 

27

 

 

 

 

Argentina

 

 

 

15

 

 

 

 

 

Sources:  for Korea, 1997 Korea Democracy Barometer Survey; for Taiwan, 1998 National Chengchi University Survey; for the post-Communist states, Richard Rose and Christian Haerpfer, “New Democracies Barometer V: A 12-Nation Survey,” Studies in Public Policy No. 206, Centre for the Study of Public Policy, University of Strathclyde, 1999, pp 59-62; for Europe, Ola Listhaug and Matti Wiberg, “Confidence in Political and Private Institutions,” in Citizens and the State, pp. 304-305, Table 10.1; Klingemann, “Mapping Political Support in the 1990s,” Table 2.11, 51.

 


 

 

Table 9

Support for Democracy in Africa, Two Measures

 

Question

Botswana

Zimbabwe

Zambia

Malawi

Lesotho

Namibia

Democracy is preferable to any other kind of government

82

71

74

66

39

58

Democracy is always best, even when things don’t work

65

74

54

59

34

43

 

Source: Robert Mattes, Michael Bratton, Yul Derek Davids, and Cherrel Africa, “Public Opinion and the Consolidation of Democracy in Southern Africa: An Initial Review of Key Findings of the Southern African Democracy Barometer,” The Afrobarometer Series, no. 1, July 2000, pp. 12-13.



[1] For conceptual depictions of democratic consolidation in this way, see Juan J. Linz and Alfred Stepan, Problems of Democratic Transition and Consolidation: Southern Europe, South America and Post-Communist Europe (Baltimore: Johns Hopkins University Press, 1996), and Larry Diamond, Developing Democracy: Toward Consolidation (Baltimore: Johns Hopkins University Press, 1999).

 

[2] For details and to order specific reports, see http://www.cspp.strath.ac.uk//

[3]For more information on this survey, see the news item, “Surveying Value Change in East Asia,” in the Journal of Democracy 11, no. 4 (October 2000), p. 189, and visit http://eacsurvey.law.ntu.edu.tw.  The comparative survey will cover ten areas of public opinion: the meaning of democracy, democratic legitimacy, regime evaluation and democratic satisfaction, economic evaluations, trust in institutions, efficacy and systemresponsiveness, attitudes toward tradition, democratic vs. authoritarian values, social capital, political participation, electoral mobilization, and partisanship and psychological involvement politics.

[4]  For evidence and analysis of the impact of differences in questionnaire wording and design, see Larry Diamond, Developing Democracy: Toward Consolidation (Baltimore: Johns Hopkins University Press, 1999), chapter 5, and Yun-han Chu, Larry Diamond, and Doh Chull Shin, “Growth and Equivocation in Support for Democracy: Korea and Taiwan in Comparative Perspective,” paper presented to the 2000 Annual Meeting of the American Political Science Association, Washington, D.C., August 31-September 3.  A revised version of this paper will appear in the series, “Studies in Public Policy, distributed by the Center for the Study of Public Policy of the University of Strathclyde.

[5] The specific articles are Richard Rose, “A Diverging Europe,” Michael Bratton and Robert Mattes, “African’s Surprising Universalism,” Yun-han Chu , Larry Diamond, and Doh Chull Shin, “Halting Progress in Korea and Taiwan,” and Marta Lagos, “Between Stability and Crisis in Latin America,” Journal of Democracy 12, no. 1 (2001), pp. 93-45.

[6] I am grateful to Richard Rose and Marta Lagos for providing me with additional data, as well as a number of the papers from their survey projects, and to Michael Bratton and Robert Mattes for providing me with additional papers from the Afrobarometer project.  I would also like to thank my co-authors Yun-han Chu and Doh Chull Shin.  Some of the data in this paper is drawn from our collaborative work, particularly our paper “Growth and Equivocation in Support for Democracy.”

[7] In particular, “In Ukraine the longer the country has been independent the greater the number who would welcome a return to Communist rule.”  SPP # 308.

[8] Studies in Public Policy no . 306, p. 47.

[9] “A Diverging Europe,” p. 101.

[10]  This omnibus question asks whether people agree that “most important decisions about the economy should be made by experts and not the government and parliament,” but majorities can—and consistenly do—agree with this proposition without this necessarily indicating lack of support for democracy. In fact, probably majorities in most Western democracies would agree with this statement.  I therefore delete this last item.

[11] In two more recent Russian surveys, the proportion rejecting all authoritarian alternatives rose to 45% in January 2000 but then slipped back to 37% in April 2000, shortly after Putin was elected.  Data are from “New Russia Barometers” VIII and IX, provided by Richard Rose and Neil Munro.  Of all the postcommunist states, only in Ukraine and Belarus does a majority endorse strong-man rule as an alterntive to democracy.  SPP #308.

[12]   SPP no. 306, p. 53.

[13] “A Diverging Europe,” Table 1, p. 99.  Data is lacking for Estonia, and the percentage is quite low (40) in Lithuania even though it is “free.”  The percentage is quite high (76) in Croatia even though at the time of the survey it was only partly free. 

[14] “A Diverging Europe,” p. 94.

[15] For the country data on this measure for the years 1996, 1998, and 2000, see Table 1 of Marta Lagos, “Between Stability and Crisis in Latin America,” Journal of Democracy 12, no. 1 (2001): 139.

[16] Responses were yes or no.  There was a similar question about whether there can be a democracy without political parties, but the national responses to these two questions were virtually identical.   The two questions were asked only in the 1997 and 2000 surveys.

[17] Ibid, p. 140.  The percentage believing democracy can work without parties rose even more sharply in Ecuador, from 47 to 59% between 1997 and 2000.

[18] Data provided by Marta Lagos from the Latinobarometro.  The percentages saying corruption has “increased a lot” were 75% in 1996, 79% in 1997 and again in 1998, and 75% in 2000.

[19] Jose Antonio Lucero, “Ecuador: Democracy in Hard Times,” forthcoming in the Journal of Democracy, April 2001.

[20]  However, it was not until more than a decade after its transition that Portugal exhibited a level of support for democracy comparable to the other established democracies of Europe.  See Chu, Diamond, and Shin, “Growth and Equivocation.”  For additional comparative data and discussion, see Diamond, Developing Democracy, pp. 174-182.

[21] Robert Mattes, Michael Bratton, Yul Derek Davids, and Cherrel Africa, “Public Opinion and the Consolidation of Democracy in Southern Africa: An Initial Review of Key Findings of the Southern African Democracy Barometer,” The Afrobarometer Series, no. 1, July 2000, p. 12.  The overall level of support for democracy in Zambia (which is not covered in the Journal of Democracy article) is 74%.

[22] Bratton and Mattes, “Africa’s Surprising Universalism,” p. 109.

[23] Ibid, p. 110.

[24] Ibid, p. 119.

[25] “Public Opinion and the Consolidation of Democracy in Southern Africa,” p. 13.  The support level under that condition is 65% in Botswana and 74% in Zimbabwe, but 43% in Namibia.

[26] Mattes et al., “Public Opinion and the Consolidation of Democracy in Southern Africa,” pp. 12-13.

[27] For an earlier summary discussion of the evidence, see Diamond, Developing Democracy, pp. 192-205.

[28] Richard Rose, William Mishler, and Christian Haerpfer, Democracy and Its Alternatives: Understanding Postcommunist Societies (Baltimore: Johns Hopkins University Press, 1999).