The Political Consequences of the Great Depression and the Great Recession: Remarkably Similar

Who benefits politically from a sharp economic downturn: the left or the right? Among political commentators, the most common interpretation of the political consequences of Great Recession (the financial and economic crisis that began in 2007–2008) is that although deep crises are typically good for the left, this one has not been. In 2009, for example, the Economist, a right-of-center newspaper, noted in its coverage of the elections to the European Parliament that center-left parties “failed to capitalise on an economic crisis tailor-made for critics of the free market.” The following year, in the New York Review, Tony Judt, a left-of-center academic, considered it “striking” that “in a series of European elections following the financial meltdown, social democratic parties consistently did badly; notwithstanding the collapse of the market, they proved conspicuously unable to rise to the occasion.” More recently, in the spring of 2012, the Washington Post columnist E. J. Dionne observed that right-wing parties had won most of the elections in Europe since the crisis began, even if “[a] crisis of capitalism is supposed to create an opening for the political left.”11 There are many other examples of similar commentary. See, for example, the Times of London (2009) and the New York Times (2009). I will argue that the premise of this interpretation is wrong. I do not mean to say that center-left parties have done well in the Great Recession; they clearly have not. I mean to say that this turn of events is not very surprising, for at least in their early stages, deep economic crises are not good for the left. On the contrary, I suspect that they are harmful. My argument draws on two recent papers, Lindvall (2012a) and Lindvall (2012b), both of which compare the consequences of the Great Recession and the Great Depression in the twenty to twenty-one countries that were democracies when the Great Depression began in 1929.22 The following countries were democracies in 1929 and held at least one democratic election between the Wall Street Crash and the Second World War: Australia, Austria, Belgium, Canada, Costa Rica, Czechoslovakia (now the Czech Republic and Slovakia), Denmark, Estonia, France, Germany, Greece, Ireland, Latvia, the Netherlands, New Zealand, Norway, Sweden, Switzerland, the United Kingdom, and the United States. The main finding that comes out of these papers is that in this sample of countries, the political consequences of the Great Recession and the Great Depression were remarkably similar: in the first phase of both crises, right-wing parties did significantly better than left-wing parties; after approximately three years, left-wing parties began to recover. The fact that the Great Recession and the Great Depression were so similar – beginning as financial crises in the United States but soon spreading to Europe and developing into worldwide macroeconomic crises that involved sharp falls in demand, output, and trade – is a powerful reason to engage in a comparison of these two episodes.33 For a critical discussion of the “Great Depression analogy,” see Bordo and James (2010). For the purposes of this paper, however, there is an even better reason: explicitly or implicitly, those who claim that crises of capitalism are “supposed” to benefit the left typically have the Great Depression in mind. The most important political consequence of the Great Depression was of course the rise of right-wing authoritarianism in Germany, Austria, Central and Eastern Europe, and Latin America. But in countries that remained democratic, the Great Depression is instead associated with political realignments that favored the left, such as the “red-green” coalitions behind ascending political leaders such as Franklin D. Roosevelt (President of the United States in 1933–1945) and Per-Albin Hansson (Prime Minister of Sweden in 1932–1936 and 1936–1946) (Gourevitch 1986, chapter 4). Sheri Berman (2006) has even argued that of the three defining ideologies of the inter-war period – fascism, liberalism, and social democracy – social democracy eventually reigned supreme. It is important to remember, however, that Franklin D. Roosevelt and Per-Albin Hansson were both elected in the autumn of 1932, three years after the Great Wall Street Crash of 1929. As Lindvall (2012b) shows, between the Wall Street Crash and the French legislative elections in the spring of 1932 (excepting the Czechoslovakian election of 1929, which was held only days after the first big fall in stock prices on the New York Stock Exchange), there was a net swing to the right in Australia, Austria, Canada, Germany, Ireland, Latvia, Norway, and the United Kingdom, but the left only made significant gains in two of the countries in the sample: in New Zealand in late 1931 and in the mid-term elections in the United States in 1930 (in neither case were the left's advances sufficient for a clear majority). This historical comparison makes the poor showing of center-left parties in the early stages of the Great Recession seem a little less remarkable. As for the string of left-wing victories that began in 1932, the current crisis is again rather similar, for since the spring of 2011, the left has done fairly well in countries such as Ireland, Denmark, and, most recently, France. Moving from election results to government turnover, Lindvall (2012a), which compares the same twenty to twenty-one democracies in the 1930s and the 2000s, shows that in the first phase of both the Great Depression and the Great Recession – that is, in 1929–1931 and 2008–2010 – most changes in government in this sample represented moves to the right (in Australia, Austria, Canada, and the United Kingdom in the 1930s; in Germany, the Netherlands, New Zealand, Slovakia, and the United Kingdom in the 2000s), with only one or two exceptions (such as Greece and the United States in the 2000s). In the second phase of both crises, however, most changes in government were moves to the left (France, Norway, Sweden, and the United States in the 1930s; Denmark, France, Ireland, and Slovakia in the 2000s), with, again, very few exceptions (Estonia and Greece in the 1930s, both of which turned to authoritarianism soon thereafter; Greece in the 2000s). It seems possible to draw two preliminary conclusions from the evidence that I have just discussed. First of all, the Great Depression analogy suggests that rather than being surprising and unusual, the failures of the left in 2009–2010 were more or less what we should have expected, if we had taken history as our guide. It may well be that crises of capitalism create windows of opportunity for the left, but if so, those windows open slowly. Second, the aftermath of the Great Depression and the Great Recession were similar, suggesting that there might be something systematic and predictable about the effects of deep economic downturns on the changing political fortunes of left and right-wing parties. Is there something about the immediate aftermath of profound economic shocks that renders the political climate inhospitable for the left? And is there something about the long-term effects of deep economic downturns that renders the political climate inhospitable for the right? The second point is perhaps more controversial than the first. Most of the existing literature on the political effects of the Great Recession is concerned with economic voting. This means that the basic explanatory model is one where voters in general punish (incumbent) politicians and parties for bad economic times, or reward them for good times – that is, a model where ideologies and party politics play only minor roles (for a discussion, see Hanspeter Kriesi's contribution to this issue of the Swiss Political Science Review). For example, Bartels (2011), in an analysis of post-crisis elections, finds a great deal of support for a “simple model of retrospective voting,” in which citizens “tended to reward their governments when their economies grew robustly and to punish their governments when economic growth slowed.” Kriesi (2011) also finds evidence of economic voting across Europe in the Great Recession (although, as he also shows, the main drivers appear to have varied between Europe's regions).44 See also the September 2011 special issue of Electoral Studies on economic voting in the crisis (vol. 30, issue 3). Like my own analysis, these studies suggest that it was indeed wrong to expect the crisis to create openings for the political left, as so much political commentary has since the crisis began. But they also suggest that I am wrong to expect any kind of ideological or party-political pattern in post-crisis elections: all that matters in these studies is growth and unemployment, and who is in power when the crisis starts. In Lindvall (2012b), I show that in the restricted sample of countries that were democracies in the 1930s, the right-wing “bias” in elections held in the first phases of the Great Depression and the Great Recession remains under control for incumbency and recent growth, suggesting that although economic voting has powerful effects on election outcomes, there is variation in the data that this explanation cannot account for. In his contribution to this issue, Hanspeter Kriesi notes that he found no evidence of a swing to the right in early post-crisis elections in the Great Recession in the sample that he has studied (which includes more Central, Eastern, and Southern European countries than my sample, but no non-European countries). This suggests that it is important to investigate how sensitive my findings are to sample selection (another advance would be to test whether the argument can be supported by an analysis of micro-level data, which is available for the Great Recession, but not, of course, for the Great Depression). Assuming, for now, that the pattern identified in Lindvall (2012a) and Lindvall (2012b) can not be explained by economic voting only, what might explain such a pattern? In another recent paper (Lindvall, Martinsson, and Oscarsson 2012), my co-authors and I have suggested that it is necessary to pay attention to distributional politics as well as economic voting when we study the consequences of economic crises. Analyses of economic voting typically estimate the effect of either subjective economic evaluations or objective economic conditions (especially at the national level) on support for the incumbent party, or parties. Governments are most successful, in these models, when they maintain a high level of support in the population as a whole (or, alternatively, among swing voters), by proving themselves to be competent economic managers or otherwise worthy of reward. Contemporary scholarship in comparative political economy is based on a very different idea of politics. In this literature, politics is all about distributional conflict (among classes, income groups, sectors, regions, or around some other political cleavage or distinction), and since there are typically more than two relevant social groups, governing parties do not stay in power by signalling competence to the electorate as a whole; they stay in power by maintaining the distributive coalition that brought them to power, or by building a new one. From this perspective, the main question that politicians face in a downturn is therefore how they will be able to maintain a sufficiently large electoral coalition in hard times. I do not think that the first set of mechanisms completely dominate the second. Concerning the specific pattern observed in this paper – early successes for the right, a comeback for the left later on – my own hypothesis (Lindvall 2012b) is that in the first phase of a deep downturn, most pivotal middle class voters are not personally affected. This means that the mechanisms that authors such as Alt (1979, 223–226, chapter 14) and Durr (1993, 167) have identified dominate: Alt concluded in a study of economic decline in the United Kingdom that economic distress makes the policy preferences of most voters less “generous,” or “altruistic,” rendering the political climate less hospitable for left-wing parties, and Durr showed that voters in the United States tend to become more liberal in good times and more conservative in bad times (a result, Durr argues, of both altruistic and non-altruistic motivations). In the second phase of a deep crisis, however, these sorts of effects diminish, since the economic and social distress brought about by a prolonged crisis is typically more widespread, which should favor the formation of center-left electoral alliances and governing coalitions (on the basis of self-interest, not only altruism). But these are only conjectures. In my view, a lot more work remains to be done before we understand the political consequences of economic crises well.