The Third Wave: Inside the Numbers
6 Dec 2013
By Jørgen Møller and Svend-Erik Skaaning
Beginning in the mid-1970s, and especially during the decade after the 1989–91 breakdown of communist regimes, the world saw a remarkable rise in the number of democracies. Recently, however, the trend has slowed. According to Freedom House’s annual Freedom in the World survey, the last seven years (2006–2012) were notable for democratic setbacks rather than democratic gains. This seven-year slump has breathed new life into two closely intertwined debates about the “third wave” of democratization.[1] Each debate belongs in turn to a broader development within democratization studies, a development that Thomas Carothers has called a shift from democratic optimism to democratic pessimism. [2]
The first debate is about the probability of a reverse wave—the potential for which first received attention in the mid-1990s, as scholars highlighted a series of threats to further democratization.[3] The recent deteriorations have rekindled these arguments. [4] Factors thought to be conducive to a reverse wave include the global financial crisis that began in 2008, the wars in Iraq and Afghanistan, the economic success of authoritarian China, and a general dissatisfaction with regime performance in new democracies. The second debate concerns the spread of a “gray zone” between democracy and autocracy. Scholars have pointed out that in many cases political openings, rather than producing genuine democracies, have yielded only “hybrid regimes,” characterized by a mix of autocratic and democratic features. Accordingly, one should not expect too large a democratic dividend from the political change that swept the world after the end of the Cold War. Most countries that have experienced regime change during the third wave are—so the argument goes—lingering in a “diminished subtype” of democracy, such as illiberal democracy, or have developed some version of electoral autocracy.
In what follows, we use a typology that we recently presented in these pages to reappraise these debates by mapping developments across the world’s political regimes between 1972 and 2012.[5] Our typology recognizes four types of democracy: minimalist democracy, electoral democracy, polyarchy, and liberal democracy. Minimalist democracy requires only that political competition for leadership take place via regular elections with uncertain outcomes. Electoral democracy goes beyond this by demanding higher levels of electoral integrity. The notion of polyarchy adds to this respect for the freedoms of speech and association. And liberal democracy adds the rule of law, understood as equality before and under the law.[6]
Finally, with regard to the residual autocratic regimes, we distinguish between closed autocracies and electoral autocracies. The latter have multiparty elections, but they are not competitive enough for the regime to qualify as democratic.[7] The electoral autocracy category (which we also refer to as multiparty autocracy) thus also serves to place in relief the defining attributes of minimalist democracy. While both regime types exhibit multiple parties and elections, there is a crucial difference: In electoral autocracies, there is no uncertainty about who will win office, whereas in minimalist democracies electoral outcomes are in fact open to opposition victories. Here, we invoke what might be termed “Schumpeter’s razor”—whether “individuals acquire the power to decide by means of a competitive struggle for the people’s vote” [8]— in order to differentiate the “thinnest” meaningful sense of democracy from electoral autocracies that formally have parliaments, elections, and multiple parties.
Shedding new light on political-regime developments during the third wave requires more than simply drawing distinctions between different kinds of democracy and autocracy, however; we must also determine whether or not global dynamics are reflected at the regional level. There are reasons to believe that regional patterns of regime change differ from one another in systematic ways due to factors such as intraregional diffusion. To understand this, we first present a global overview and then, using Freedom House’s regional divisions of the world, revisit developments in the Americas and the Asia-Pacific region; in Eastern Europe (a term that we use in place of Freedom House’s “Central and Eastern Europe/former Soviet Union”) and sub-Saharan Africa; and in the Middle East and North Africa.[9]
The purchase of this twofold disaggregation is reflected in Larry Diamond’s recent discussion of a potential “coming wave” of democratization. According to Diamond, East Asia is the region most likely to be the scene of a “new lift to global democratic prospects,” in terms of both
the number of democracies and their quality. East Asia more than the Middle East, writes Diamond, “is better positioned to increase the number of liberal and sustainable democracies.”[10] We can obviously only make such determinations about ongoing patterns of regime developments by both scoping down to the regional level and distinguishing between minimalist and maximalist types of democracy. This is what we set out to do in this essay. Based on our empirical surveys, we are able to draw some conclusions about the current dynamics of democratization—most notably, about the extent to which a general democratic Zeitgeist is safeguarding democracy at the global level.
Global Trends
As recently as 1978, only 49 of the world’s then 158 states (31 percent) were democracies; by 2012, 117 of 195 countries (60 percent) were democracies. Figure 1 documents this remarkable change in the ratio between democracies and autocracies, which confirms the notion of a third wave of democratization. But the figure also shows that the gradual increase before 1989 pales next to the dramatic spike that took place especially in 1990–93 and came to an end around 1995. That is not to say that after the mid-1990s there was a complete absence of further democratic gains. An important movement of countries within the set of democracies occurred in the early 2000s, when the number of liberal democracies grew substantially at a time when the number of democracies overall had reached a standstill.
What does Figure 1 tell us about an ongoing democratic rollback and about the increasing importance of the so-called gray zone? The overall
number of democracies has in fact declined since 2005, though so far the dip has been slight (from 123 to 117). Indeed, the overall number of democracies has been remarkably stable since the mid-1990s. Next, our overview confirms the widespread impression that minimalist democracies and multiparty autocracies—the two regime types in the gray zone—have proliferated since 1989. According to our account, only nine countries were minimalist democracies in 1988; in 2012, the number was thirty. The profusion of multiparty autocracies is equally striking. This regime type is not a brand-new phenomenon; there were multi-party autocracies in the 1970s and 1980s, but after the 1989–91 rupture the number of multiparty autocracies exploded—from 21 in 1988 to 56 today. In contrast, during the same period the number of closed autocracies decreased from 76 to 22. In the late 1980s, the number of closed autocracies gradually decreased; then, with the end of the Cold War, it sharply and abruptly fell, and went on to experience a further gradual decline during the 2000s.
These developments obviously testify to the growth of the gray zone. It is worth noting, however, that minimalist democracy and multiparty autocracy were not the regime types highlighted by the democratic pessimists of the 1990s. Rather, those scholars were concerned about the emergence of political regimes that combined free and fair elections with other illiberal features, such as violations of freedom of speech, problems with regard to the rule of law, or the absence of horizontal accountability—in other words, according to our typology, electoral democracies. [11] Except during some of the 1990s, we find relatively few instances of this type, and in recent years the category has almost become extinct.
The Americas and the Asia-Pacific. The global dynamics presented above subsume a diverse array of regional developments. The two regions where developments most closely match the global trends are the Americas and the Asia-Pacific countries. Beginning in the late 1970s, the number of democracies gradually increased in both regions, leveling off in the mid-2000s in the Americas and the mid-1990s in the Asia-Pacific (see Figures 2 and 3).
The Americas represent the third-wave region par excellence. Needless to say, it is Latin America specifically that has driven the changes. In this region, a gradual increase in the number of democracies characterized the entire period from the late 1970s—beginning with democratic transitions in Ecuador in 1979, Peru in 1980, and Argentina in 1983—up to 2005. By the mid-1990s, the steady stream of new democracies had almost dried up, but an important development marked this period of overall stagnation: the movement of countries from thinner to thicker types of democracy—in particular, the rising number of liberal democracies, including Chile and Uruguay. Finally, there was an aggregate decline in the number of democracies after 2005. This decline stems solely from a decrease in instances of thinner types of democracy due to negative developments in countries such as Venezuela, Honduras, and Nicaragua that moved toward multiparty autocracy.
Developments in the Asia-Pacific region have been fairly similar to those in Latin America: a gradual increase in the number of liberal democracies (some of the Pacific’s small island states), a relatively stable number of polyarchies (including South Korea and Taiwan), and a notable increase in the number of minimalist democracies (such as Bangladesh, the Philippines, and Thailand) that have tended to move
in and out of the set of democracies over the period scrutinized. Yet the Asia-Pacific also differs from Latin America in key ways. Notably, in the Asia-Pacific democratic progress began from a somewhat lower base; the post-1989 increase was more abrupt; and today multiparty autocracies and closed autocracies both are more prevalent than in Latin America. Finally, in the Asia-Pacific region we find no evidence of a general democratic decline after 2005.
Eastern Europe and Sub-Saharan Africa. In contrast to Latin America and Asia, neither Eastern Europe nor sub-Saharan Africa was affected by the third wave prior to 1989. As demonstrated in Figures 4 and 5, there were no democracies in the Eastern bloc in the 1970s and 1980s, while the number in sub-Saharan Africa was negligible. The breakdown of communist regimes was the “big bang” that inaugurated a period of political change in Eastern Europe (reflecting the Cold War dividing lines, this region includes all former Soviet countries). After a few years, the majority of the postcommunist countries had democratized, although to different degrees. Whereas the East European countries that are today EU members moved on quickly to the thicker types of democracy, the new democracies in the Balkans and Caucasus remain only minimalist, and the five Central Asian countries never made it into the set of democracies at all.
This critical juncture meant that—in terms of political regimes— Eastern Europe went from being the world’s most homogeneous region to its most heterogeneous, now representing a distribution of regime forms quite similar to the global range. One aspect of this development is that the overall number of democracies and autocracies has been remarkably stable since the mid-1990s. Yet within each of these overarching categories, there have been important developments. First, our
survey of regime types shows a proliferation of multiparty autocracies. Second, an even more conspicuous trend is the democratic deepening that characterized both the 1990s and 2000s, during which time a number of countries first become polyarchies and then liberal democracies.
The development in sub-Saharan Africa is in many ways related to these trends. Here, the “big bang” took place a couple of years after the end of the Cold War—and it was more geographically scattered and temporally spread out. Nevertheless, the democratization of the early 1990s transformed a region dominated by closed autocracies into one where almost half the countries were democracies by the mid-1990s. Conspicuously absent in sub-Saharan Africa, however, are genuine instances of liberal democracy—only Cape Verde qualifies, though there are a number of polyarchies. Rather, it has been the minimalist democracies (countries such as Liberia, Malawi, and Tanzania) and especially multiparty autocracies that have made headway, thereby rendering sub-Saharan Africa the gray-zone region par excellence. Most striking here is the observation that close to half of all sub-Saharan African countries are presently multiparty autocracies and that this proportion has been on the rise since 2007. What is more, recent democratic breakdowns in countries such as the Central African Republic, Guinea-Bissau, Madagascar, and Mali have led to a significant decrease in the number of democracies in the region.
The Middle East and North Africa. Finally, we come to the one region that the third wave seemed to pass by: the Middle East and North Africa (MENA). [12] Over the entire period of our study, this region is— with Israel as the only major exception—consistently autocratic (see Figure 6). As such, it did not differ much from sub-Saharan Africa and
the areas dominated by the Soviet Union in the 1970s and 1980s. But since the end of the Cold War, the MENA region has been reduced to an autocratic island in an increasingly democratic sea. To be sure, following the “Arab Spring” Tunisia and Libya have become minimalist democracies according to the Freedom House figures. But it remains to be seen whether the overt challenges to autocratic rulers are evidence that the global process of democratization has finally reached this region. So far, the only broad change in the region that can be documented is a proliferation of multiparty autocracies since 1992. This development indicates that the third wave has in fact had some impact, even in the one region where it did not produce democracies as such.
Different Measures, Similar Results?
Students of democracy know well that different measures often lend support to different conclusions. Therefore, we have reexamined the mapping of political regimes in two ways. First, we have used a different measure to distinguish between electoral autocracy and closed autocracy—namely, an indicator on electoral competitiveness (Legislative and Executive Indices of Electoral Competitiveness, or LEIEC) from the World Bank’s Database of Political Institutions.[13] This indicator identifies cases where the largest party officially received less than 75 percent of the vote: It is a more demanding indicator of electoral competitiveness than the one that we used above, which asks only whether multiple parties hold seats in the national legislature. It is therefore unsurprising that we find fewer electoral autocracies throughout the period if we employ the LEIEC indicator. That said, the trends are virtually identical, varying only in degree. We take this as a corroboration of our findings.
Second, we have employed the subcategory scores from the Freedom in the World survey for the period 2005–12 to distinguish among the four types of democracy. This was the procedure used when we presented our typology in these pages in January 2013, and it ensures a higher correspondence between our definition of the democratic regime types and the empirical measurement. The only reason that we have not used it consistently here is the lack of disaggregated scores for the years prior to 2005. It is heartening that the results derived by this method do not alter the general patterns reported above.
What Does It All Mean?
We have documented a small decline in the overall number of democracies in recent years. One way to put this development in perspective is to revisit the earlier waves that Samuel P. Huntington identified. Huntington wrote that the first, almost century-long wave of democratization began in 1828, with the extension of suffrage to most white men in the United States. This wave subsided with Mussolini’s rise in 1922, which marked the beginning of a reverse wave that lasted until 1942. A second wave of democratization commenced the following year and came to a close in 1962. Another reverse wave occurred between 1958 and 1975, while the third wave of democratization began in 1974.
Huntington’s wave metaphor is both appealing and easy to grasp. But subsequent reappraisals have questioned his descriptive findings. [14] Whereas the democratic increases associated with the three waves consistently show up in the data, the same is not true of the reverse waves. Although we can identify a first reverse wave in the interwar years, when democracy was opposed by totalitarian movements, the magnitude of this ebb tide was less salient than Huntington’s account indicates. The occurrence of a second reverse wave, inaugurated by the Western decolonization of Africa in the 1960s, is simply not supported by the data. Rather, the period between the early 1960s and early 1970s is an example of what Renske Doorenspleet has aptly termed “trendless fluctuation.”[15]
Based on these historical patterns alone, there is reason to doubt the likelihood of a new reverse wave of democracy. Democracy simply has not tended to wax and wane in such a consistent manner. The period after 1989 has been historically exceptional due to the speed and frequency of transitions to democracy. Similar developments in earlier periods were much messier, with significant fluctuation around a long-term, usually positive trend. So what can we infer from this pattern? If anything, it is that a new period in which democracy is challenged is more likely to yield a general standstill characterized by plenty of back-and-forth movement toward and away from liberal democracy than it is to produce a major democratic rollback.
Recall here the circumstances of the one genuine reverse wave: the dark years between the two world wars. This period was from the outset marked in Europe’s collective consciousness by the insidious political effects of the Great War. At the same time, the new democracies that formed in its wake, many of which had no prior democratic experience, were soon put under great stress first by the postwar inflationary crisis of the early 1920s and then by the Great Depression. Finally, at that time the democratic idea had some ideologically potent adversaries—most notably, communism, fascism, and Nazism. It was against this backdrop that democracy genuinely came under attack.
The current challenges to democracy pale in comparison to those of the interwar years. Not only is our current economic crisis less severe than was the Great Depression, but there is no contemporary equivalent to the traumatic experience of the First World War. Moreover, there are today no serious challengers to democracy as a regime form.[16] The democratic Zeitgeist, though less ebullient than in it was just after the Cold War ended, still reigns. Surveys show that most people in almost every country worldwide prefer democracy over other types of regimes.[17] One indication of this widespread preference for democracy is that the number of autocracies holding multiparty elections has increased even since 2005—that is, in the context of a democratic standstill (at best) or decline (at worst). This development reflects how hard it is for authoritarian rulers to avoid at least paying lip service to democratic ideals.
The empirical overviews presented in this essay show the value of looking at regime types and regional developments in a disaggregated way. First, although Huntington had already identified the third wave of democratization by 1991, the major democratic increases did not take place until after the Cold War’s end in 1989–91. This is visible in the global overview alone, but our illustrations of regional developments also reveal that the democratizations of the 1970s and 1980s left three of the five world regions unaffected. Only after 1989 did the wave reach Eastern Europe and sub-Saharan Africa, and only very recently have we seen serious challenges to autocratic rule in the Middle East and North Africa. These developments tend to support the claim that a fourth wave of democratization, qualitatively and quantitatively different from the third, should be teased out and dated to the breakdown of communist regimes and the dissolution of the Soviet Union.[18]
Second, even though most regions surely have a gray zone that has come to include more countries over time, this zone is far more salient in some regions, such as sub-Saharan Africa, than in others, such as the Americas, where many former autocracies have developed into polyarchies and liberal democracies. Furthermore, rather than being populated mainly by electoral democracies that combine free and fair elections with disrespect for liberal rights, these gray zones are home mostly to the lower-quality versions of electoral contestation—what we have termed minimalist democracies and electoral autocracies—that have proliferated in the post–Cold War era.
Third, the period between the mid-1990s and the mid-2000s has been characterized by two contradictory trends: stagnation in the total number of democracies and a simultaneous deepening of democracy, as evidenced by the growing number of liberal democracies.
What about the prospects for a new reverse wave? Does the post-2005 dip represent the first ripples of a turning tide? Any wavelike movement presents to the analyst a moving target, but our analysis shows that the decrease has so far been a marginal one and, furthermore, one that has not affected every region.[19] In fact, only the Americas and sub-Saharan Africa have registered declines. In two other regions, the Asia-Pacific and Eastern Europe, there has been no similar democratic decline since 2005. And in the Middle East and North Africa, the very limited and very recent changes actually point in the opposite direction.
Needless to say, the absence to date of a significant reversal does not prove that one will not occur at some point. But the historical comparison with the one genuine reverse wave—that of the interwar period— makes us more rather than less doubtful that a real rollback could occur today. That the third wave is relatively robust in the aggregate also seems to be the most recent conclusion of some of the authors who, in the wake of the 2008 financial crisis, warned about a possible rollback or turning of the tide.[20] One way of reading the evidence is to note that the modest decrease since 2005 shows just how strong the commitment to democracy is, considering all the stresses that it has had to bear. If we link these observations to the historical patterns, much of the evidence indicates that we have now entered a period of aggregate standstill, marked by frequent fluctuations in both directions.
Notes
[1] Samuel P. Huntington, The Third Wave: Democratization in the Late Twentieth Century (Norman: University of Oklahoma Press, 1991).
[2] Thomas Carothers, “How Democracies Emerge: The ‘Sequencing’ Fallacy,” Journal of Democracy 18 (January 2007): 12–27.
[3] See, for example, Larry Diamond, “Is the Third Wave Over?” Journal of Democracy 7 (July 1996): 20–37; Larry Diamond, “Is Pakistan the (Reverse) Wave of the Future?” Journal of Democracy 11 (July 2000): 91–106; Fareed Zakaria, “The Rise of Illiberal Democracy,” Foreign Affairs 76 (November–December 1997): 22–43.
[4] Arch Puddington, “The 2007 Freedom House Survey: Is the Tide Turning?” Journal of Democracy 19 (April 2008): 61–73; Larry Diamond, “The Democratic Rollback: The Resurgence of the Predatory State,” Foreign Affairs 87 (March–April 2008): 36–48; Paul Krugman, “Depression and Democracy,” New York Times, 12 December 2011.
[5] A table classifying the world’s regimes in 2012 by this typology may be found at www.journalofdemocracy.org/articles/supplemental-material; a full overview of regime type by country year is available at www.ps.au.dk/dedere.
[6] Jørgen Møller and Svend-Erik Skaaning, “Regime Types and Democratic Sequencing,” Journal of Democracy 24 (January 2013): 142–55; Jørgen Møller and Svend-Erik Skaaning, Requisites of Democracy: Conceptualization, Measurement, and Explanation (London: Routledge, 2011).
[7] We use the Freedom House (FH) annual lists of electoral democracies to distinguish democracies from autocracies. This list goes back only to 1989, however, so for the years 1972–88, we use the dataset on democracies presented in Carles Boix, Sebastian Rosato, and Michael Miller, “A Complete Dataset of Political Regimes, 1800– 2007,” Comparative Political Studies (forthcoming). Both measures capture the defining components of what we term minimalist democracy. Regarding the democracies, we consider those scoring worse than 2 on FH’s political-rights index to be minimalist democracies. Among the democracies earning political-rights scores of 2 or better, we use the score on the FH civil-liberties index to make further distinctions: Those earning civil-liberties scores worse than 2 are considered electoral democracies; those scoring 2 are considered polyarchies; and those scoring 1 are considered liberal democracies. To distinguish between the two kinds of autocracies, we use the lparty indicator from the “Democracy-Dictatorship” dataset provided by José A. Cheibub, Jennifer Gandhi, and James Vreeland in “Democracy and Dictatorship Revisited,” Public Choice 143 (April 2010): 67–101. Countries with multiple parties in the national legislature (scoring 2) are considered multiparty autocracies, while those with only the ruling party represented in the legislature (scoring 1) or no legislature at all (scoring 0) are considered closed autocracies. For the years 2009–12 and a few additional missing values, we have updated this indicator ourselves. This measurement procedure is imperfect; for the time being, however, a lack of appropriate, disaggregated indicators places serious limits on achieving alignment between concepts and measures.
[8] Joseph A. Schumpeter, Capitalism, Socialism and Democracy (London: Unwin University Books, 1974 [1942]), 269. See also Møller and Skaaning, “Regime Types and Democratic Sequencing.”
[9] Our numbers are based on the absolute number of countries situated in each regime form. An alternative would be to use the proportion of countries, but that could be misleading because a spike in the total number of countries would appear as a reverse wave insofar as most new states begin their lives as autocracies.
[10] Larry Diamond, “The Coming Wave,” Journal of Democracy 23 (January 2012): 5.
[11] Guillermo O’Donnell, “Delegative Democracy,” Journal of Democracy 5 (January 1994): 55–69; Zakaria, “The Rise of Illiberal Democracy”; Larry Diamond, Developing Democracy: Toward Consolidation (Baltimore: Johns Hopkins University Press, 1999).
[12] According to Freedom House’s regional distinctions, Turkey is part of Western Europe rather than the Middle East and North Africa.
[13] Philip E. Keefer, “Database of Political Institutions 2012 (updated Jan. 2013),” World Bank, Research on Macroeconomics and Growth, http://tinyurl.com/fs5bg.
[14] [14] See Renske Doorenspleet, “Reassessing the Three Waves of Democratization,” World Politics 52 (April 2000): 384–406; Jørgen Møller and Svend-Erik Skaaning, Democracy and Democratization in Comparative Perspective: Conceptions, Conjunctures, Causes, and Consequences (London: Routledge, 2013), ch. 5.
[15] Doorenspleet, “Reassessing the Three Waves of Democratization,” 400–401.
[16] A partial exception might be political Islamism, which has at least formulated an elaborate ideological alternative to democracy.
[17] Yun-han Chu et al., “Public Opinion and Democratic Legitimacy,” Journal of Democracy 19 (April 2008): 74–87; Pippa Norris, Democratic Deficit: Critical Citizens Revisited (New York: Cambridge University Press, 2011).
[18] See, for example, Renske Doorenspleet, Democratic Transitions: Exploring the Structural Sources of the Fourth Wave (Boulder, Colo.: Lynne Rienner, 2005); Michael McFaul, “The Fourth Wave of Democracy and Dictatorship: Noncooperative Transitions in the Postcommunist World,” World Politics 54 (January 2002): 212–44.
[19] And if we apply the Polity IV data (another widely used democracy measure), we do not find a decrease at all. According to this dataset, the number of democracies in the 2000s is higher than in the 1990s, and there are as many democracies as ever today. See Møller and Skaaning, Democracy and Democratization in Comparative Perspective, 87.
[20] As already mentioned, the overall stability disguises a lot of fluctuation. A large share of third-wave democracies have regressed or even broken down at some point. But many have redemocratized, and some autocracies have given way to genuinely new democracies, thereby preventing a rollback in the aggregate. For warnings of a democratic rollback, see Arch Puddington, “The Freedom House Survey for 2009: The Erosion Accelerates,” Journal of Democracy 21 (April 2010): 136–50; Larry Diamond, “The Impact of the Economic Crisis: Why Democracies Survive” Journal of Democracy 22 (January 2011): 17–30.