Abstract
Do ideas matter in politics anymore? Have we seen the end of ideology? Is globalization forcing advanced capitalist democracies to “race to the bottom” by paring welfare states to the bone to prevent business from fleeing to countries that spend less on welfare programs and keep taxes low accordingly? These are some of the questions Alexander Horn tackles in Government Ideology, Economic Pressure, and Risk Privatization: How Economic Worldviews Shape Social Policy Choices in Times of Crisis, his meticulous and sophisticated statistical analysis of eighteen OECD countries over the last forty years. But the core questions driving his analysis are these: Have governments in these countries pursued welfare state retrenchment? And if they have, why?
Horn explains that there are three major views on these core questions. First is an argument about the persistence of partisan influence. For example, power resource and political partisanship theories argue that right-wing governments pursue retrenchment while left-wing governments resist it. This is because each party represents the interests of their chief constituencies: capital and labor, respectively. In other words, governments pursue policies that reflect the interests and therefore the ideological dispositions of their supporters.
Second, and in stark contrast, are arguments about the marginalization of politics. Here ideology doesn’t matter. On the one hand, this is because welfare programs tend to foster constituents that benefit from and therefore defend them. As a result, there is little reason to expect any sort of welfare state retrenchment, partisan or otherwise. This is an argument about path dependence and the absence of ideological effects. On the other hand, even if retrenchment does occur, it still has nothing to do with ideological partisanship but rather the functional imperatives of globalization and economic crisis.
Third, a reversed partisanship argument stipulates that welfare state retrenchment has occurred but thanks to left-wing governments rather than right-wing governments, and for two reasons. One is that leftist parties have gradually shifted in more ideologically conservative directions. The other is that leftist parties have been seen traditionally as defenders of the welfare state, so their arguments for retrenchment are more legitimate politically than whatever justifications right-wing parties offer. Put differently, if left-wing parties say we need to trim welfare programs, then it must be true.
Horn reviews these theories in detail and provides insightful critiques of them all, raising serious questions about each on theoretical and methodological grounds. For instance, conventional measures of left, center, and right party ideologies are static, but these ideologies may have changed significantly since the 1970s. Moreover, party ideologies vary depending on which policy areas we are talking about. In short, traditional ways of mapping generic left-right party distinctions are inappropriate for testing theories of welfare state retrenchment per se. This is what Horn calls the “independent variable problem.” But he also sees a “dependent variable problem.” The ways in which people measure welfare state retrenchment often focus on only one policy dimension, such as expenditure levels. He says that a multi-dimensional approach is necessary to get the complete picture. Horn takes significant steps toward resolving both problems. What does he find?
To begin with, Horn develops time-variant measures of the normative and causal dimensions of political party ideologies either favoring or opposing welfare state retrenchment. He also operationalizes retrenchment along two dimensions. One is welfare “generosity” (a metric involving welfare benefits net of taxes); the other is “conditionality” (a metric involving benefit entitlements and eligibility). He uses these measures throughout the book, focusing specifically on unemployment insurance as his proxy for welfare effort. He deploys these measures in a variety of statistical models testing hypotheses derived from the core theories.
His results are convincing. Over time, parties and governments tended to shift ideologically toward the right, increasingly favoring welfare state retrenchment. This was true both within and across countries. There was also a tendency toward convergence in how this played out across countries. However, these results were modest at best. Strong differences remained across countries and among parties within countries. But, shifting patterns aside, did ideology affect retrenchment?
The answer is yes, depending on how things were measured. Using most of the traditional—and overly general—measures of party ideology, Horn finds little evidence that ideology affects either welfare generosity or conditionality. But when it comes to his more precise—that is, welfare specific—ideological measures, ideology appeared to have significant effects. Notably, governments with strong anti-welfare state ideologies are significantly more likely to be associated with cutting unemployment insurance schemes. However, governments with strong pro-welfare state ideologies are not apparently inclined to expand unemployment insurance. Importantly, although economic conditions (but not measures of globalization) were significantly and consistently associated with retrenchment, they did not wash out the effects of anti-welfare state ideology. In fact, Horn concludes that the presence of a strong anti-welfare state ideology amplifies the effect of economic conditions. This supports his theory about cognitive framing, which suggests that how governments deal with economic problems is influenced by the ideological lens through which they interpret these problems and devise solutions for them. To further illustrate the point, he sketches a few qualitative case studies to augment the quantitative analysis—Helmut Kohl’s Germany, Margaret Thatcher’s Britain, and Ingvar Carlsson’s Sweden.
I was impressed by how even-handed Horn was in dealing with the various theories in play. And the data seemed to be consistent with his arguments. However, I wondered from time to time whether he had overplayed his hand. That is, in the end he concludes that ideas and ideology matter. Fine. I agree. But the ideology that mattered was only the anti-welfare state type, not the pro-welfare state type. Furthermore, the most consistent finding in all his models was that economic conditions mattered too. Economic conditions were associated in statistically significant ways with welfare retrenchment in every single model he offered (Tables 6.3 and 6.4). One way we could interpret these results is to say that not all ideologies are created equal in terms of their effects, but that the effects of economic conditions are at least of equal importance to those of ideology.
This book began its life as a doctoral dissertation and still reads a bit like that. But it has a lot going for it, and I would recommend it to comparative welfare state scholars and other researchers interested in globalization, the Varieties of Capitalism, and the political impact of ideas and ideology on policymaking. The discussion and critiques of different theoretical traditions and methodological approaches is excellent. The suggestions for conceptualizing and improving the measurement of party ideologies is especially useful. For these reasons alone, it is worth reading.