Democratic Institutions
They argue that the electoral system shapes the incentives of politicians either to toe the party line or to use their influence over public policies to cultivate a personal following. Confirming a long-standing intuition among students of political parties, the incentives for politicians to campaign on a broad party platform depends on the ability of parties to control politicians’ re-election chances. The best-known means to accomplish such control is a closed party list system where a candidate’s rank on the list determines that candidate’s likelihood of re-election. By contrast, in systems where candidates are chosen through primaries, such as elections to the US Congress, political parties cannot directly control who gets on the ballot and politicians have a stronger incentive to pursue their own agendas and appeal to location-specific interests. Carey and Shugart do not fully spell out the implications of their argument for economic policy, but an obvious hypothesis is that systems that encourage politicians to cultivate a personal following will lead to a targeting of public money to local projects and narrowly defined groups. An example is the pre-1994 Japanese system where candidates from the same party competed against each other for a single non-transferable vote, and therefore produced strong incentives for politicians to spend on their districts while ignoring the public interest. The consequence was a highly fragmented fiscal policy (Cox and Thies 1998; Rosenbluth and Thies 2001). This resonates with Lowi’s notion of distributive politics, but in the Carey–Shugart framework the nature of the policy-making process is a function of the political system, which varies across time and space.
Put another way, Carey and Shugart help explain differences in countries with respect to their use of policies across Lowi’s categories. In Persson and Tabellini’s (1999, 2000) account, the critical institutional feature is the electoral formula. In majoritarian systems, they argue, if middle-class swing voters are concentrated in particular districts, parties have an incentive to ignore completely other districts that are leaning one way or the other ideologically. These districts are “safe” and therefore not worth fighting over. Similarly to Dixit and Londregan’s model, money therefore flows to swing votes in middle-class districts. In single-district PR systems, by contrast, there are no safe districts so politicians cannot ignore the loss of support among other groups by concentrating transfers on the middle class. The result is greater dispersion of spending across classes (or more spending on broad public goods). PR systems, however, tend to spend more on both transfers and public goods. To explain this Persson and Tabellini point to a “second-order effect” of PR, namely that PR systems tend to have more parties and be ruled by multiparty governments (see Ordeshook and Shvetsova 1994; Neto and Cox 1997 for the evidence).
If each party wants to spend on its own group (so that the space is multidimensional), this can lead to a common-pool problem with excessive spending as a result (see also Bawn and Rosenbluth 2002; Crepaz 1998). Again, Lowi’s distributive politics is here linked to an institutional feature that varies across countries. Electoral systems also appear to be systematically related to class politics. Iversen and Soskice (2006) argue that when parties representing different classes have to form coalitions to govern, as is typically the case under PR, the center and left have an incentive to get together to soak the rich. This is not true in two-party majoritarian systems where parties are coalitions of classes. In this situation, middleclass voters can be soaked by the poor if the center-left party deviates from a median voter platform. Assuming that the right cannot engage in regressive redistribution, incomplete platform commitment therefore puts the median voter at risk and the center-left at an electoral disadvantage.
The implication is that partisanship and redistribution, class politics in Lowi’s scheme, systematically covary with electoral system. Another democratic institution that has generated intense scholarly scrutiny is federalism.Much of the research in this area originates with Brennan and Buchanan’s (1980) argument that competition between local governments for mobile sources of revenue undermines the ability of predatory governments to impose excessive taxation. Coupled with the potential ability of states to secede, which restricts the ability of central governments to exploit member states, federalism may also constitute a credible commitment to property rights—whatWeingast (1995) calls “marketpreserving federalism.” Viewed from the left, this logic suggests that federalism may undermine the welfare state and lead to under-provision of social welfare or a “race to the bottom” (Pierson 1995). A related argument is that federalism makes it harder to pass new legislation because it has to be ratified in two legislative assemblies. The implication is a status quo bias, which is sometimes argued to have slowed the expansion of the welfare state (Huber, Ragin, and Stephens 1993). But while federalism does appear to be associated with smaller governments, there is in fact a striking amount of variance across federalist states (Obinger, Leibfried, and Castles 2005). Swiss and US federalism seems to be linked to low spending, but this is not true of German or Austrian federalism. To account for this variation, Rodden (2003, 2005) has proposed to distinguish between federalist systems with different fiscal institutions. If local spending is locally financed, tax competition puts a damper on spending, but if local spending is financed through central or intergovernmental grants, local politicians have little incentive to contain spending.
This argument applies more generally to political systems where there is a division of labor between local governments and the center (a unitary system like Sweden, for example, exhibits a lot of local policy autonomy). Revenue-sharing may be seen as a source of common-pool problems (or Lowi’s distributive politics), or it may be seen as a method to reduce the power of those with mobile assets and empowering governments to pursue redistribution. Whatever the normative perspective, if there are two different types of federalism, one with a “soft” and one with a “hard” budget constraint, then a key issue is why some governments have adopted hard budget constraints while others have not (Wibbels 2003;Wildasin, this volume). Alternatively, there is a wealth of interesting variation in revenuesharing between local governments and states that seems critical to fiscal policy, yet is not treated as an object of explanation. There is a literature on fiscal discipline and its consequences (see von Hagen and Harden 1995 and the von Hagen chapter in this volume), but budgetary rules are for the most part treated as exogenous. By anchoring model assumptions in the rich details of actual political institutions, the new institutionalist literature enables the coupling of formal reasoning with the realism of inductive research. It reduces the indeterminism of democratic policymaking and suggests promising ways to endogenize partisanship, coalition formation, and styles of policy-making. Our understanding of fiscal policies and distributive outcomes has been greatly advanced in the process.
But by highlighting the critical importance of institutional detail, one cannot help but wonder if the real task is not the explanation of the institutions themselves. I return to this question below. But first I turn to another successful branch of institutionalism that focuses on modern capitalism as an economic system..
