Political economy.

Alesina, Alberto F.

Political Economy

Economists usually look at how policy should be conducted in a world free of institutions in which a benevolent social planner maximizes society's welfare. Such normative models are essential for understanding the benchmark case of a planner who faces a "representative individual"; however, they do not fully explain the many apparent departures from first-best policies in the real world.

My research focuses on positive aspects of economic policy and, more specifically, on the interaction between politics and macroeconomics. It recognizes first that there are no neutral social planners; politicians respond to ideological and opportunistic incentives and pressures and are constrained by the institutions in which they operate. Second, individuals and groups have conflicting policy preferences, in particular because of the distributional consequences of different policies. Economic policies emerge from the resolution of these conflicts of interests through the political process.

Political Business Cycles

In a 1987 paper, I consider the consequences for business cycles of electoral uncertainty.(1) I develop a model in which two political parties with different preferences for inflation and unemployment alternate in office. Economic agents are rational and are aware of the differences in the policy preferences of the two parties, but they cannot anticipate post-election policy perfectly, since they are not certain about which party will be elected.

Because of this uncertainty, the rate of inflation in the post-electoral period cannot be anticipated perfectly before elections. If nominal contracts must be signed, and they are not synchronized perfectly with the elections, then the model predicts the level of economic activity will deviate from its "natural" level after the elections. When the unemployment-averse party (left) is elected, there will be rapid growth and rising inflation. After expectations adjust to the new government, economic activity returns to its natural level. Inflation remains high, because an unemployment-averse government finds it difficult to commit credibly to a policy of low inflation. The opposite outcome will occur when the inflation-averse party (right) is elected. After an early downturn, economic activity returns to its natural level with low inflation. Thus, the implications of this model are quite difficult from the traditional business cycle model, in which every government has the same opportunistic preferences and creates rapid growth before each election.

My empirical research on the United States has supported my 1987 model.(2) Two other papers show that this correlation between elections and economic fluctuations is common in many other OECD democracies as well.(3) In particular, the pattern is stronger in countries in which political changes between right and left are relatively unambiguous, as opposed to countries with large center-left coalition governments with frequent early elections and coalition collapses.

An important question raised by these findings is why political parties remain polarized, rather than converging to the preferences of the median voter. In a series of papers, I address the issue of the degree of policy convergence in a two-party election.(4) These papers develop electoral models in which two parties face each other in a series of elections. Both parties want to win per se, but they have different preferences over policies. In general, the parties will have to trade off between the ideologically preferred policies and more "moderate" policies that will increase their chances of electoral victory.

The credibility of pre-electoral policy announcements also becomes an important issue. Politicians face different incentives before and after elections. Before, they want to appeal to the largest possible electorate, and thus they would like to converge toward the median voter preferences. Once in office, though, they may feel freer to pursue their ideological objectives. The behavior of policymakers once in office may be affected by reputational considerations, too: fear of losing "reputation" in the eyes of the voters may constrain the policies each government chooses. In general, reputational considerations will reduce but not eliminate the political polarization in two-party systems.

Divided Government and the Mid-Term Voting Cycle

In American elections, the party holding the presidency always loses seats in mid-term congressional elections relative to those won in the previous presidential election year. This sometimes leads to a "divided government" in which the president's party does not have a majority in the House and/or in the Senate. In a recent paper with Howard Rosenthal, I suggest that voters choosing between two relatively polarized parties opt to counterbalance the president by leaning toward the other party in the legislative elections.(5) In this way, voters in the middle of the political spectrum can bring about "moderate" policy outcomes.

Part of this balancing also occurs in the congressional election held at the same time as the presidential election, but the voters have a second chance to moderate in mid-term elections. In mid-term, the voters already know who is in the White House and can choose the optimal amount of balancing; in presidential election years, on the contrary, the congressional vote in cast under uncertainty about which party will win the presidency.

In a related paper, Rosenthal and I develop a macro-economic model in which the two parties differ in their evaluation of the relative costs of inflation and unemployment.(6) A newly elected Democratic administration follows expansionary policies leading to a temporary high growth rate. In mid-term congressional elections, the voters fear excessive inflation and react by turning toward the Republican Party, which gains representation.

The opposite occurs if the Republican candidate wins the presidential elections. This model, which performs quite well on postwar U.S. data, implies correlations between macroeconomic variables and voting behavior that are quite different from those of the traditional economic model of voting; in that model, voters simply reward the incumbent if the economy is doing well before the elections. Our approach also explains why traditional voting models do not perform as well for congressional elections as they do for presidential elections. More generally, our model suggests that voter myopia does not explain observed voting patterns in the United States.

The Political Economy of Public Debt

In the last two decades, many industrial nations have experienced large increases in government debt. In several countries (such as Belgium, Italy, and Ireland) the debt-to-GNP ratio is around 100 percent. In the United States, the debt-to-GNP ratio rose sharply in the 1980s after a steady decline beginning after World War II. In two papers, Guido Tabellini and I ask why so many governments have been prone to peacetime deficits. We also explore the wide variance of debt levels in different countries with relatively similar economic conditions.(7)

We model government debt as the legacy that each government leaves to its successors. In a polarized political system, each government may be replaced by an opponent with very different preferences regarding the desired distribution of the tax burden, and the level of and the composition of spending. By manipulating the size of government debt, each government can affect the state of the economy and thus the set of policy options available to its successors. This strategic interaction of the current government, supported by the majority of voters, with future governments, possibly supported by different majorities, may lead to deficits in excess of the first-best optimum.

One crucial testable implication of this view is that deficits should be higher in countries and time periods with greater political polarization; that is, when the ideological distance between alternating governments is wide. A second implication is that when the current government is not likely to win the next election, deficits should be higher; the party in power has a stronger incentive to overspend and manipulate the budget to influence the choices of future governments.

In LDCs, political polarization and uncertainty may lead to accumulation of public external debt and private external assets (capital flight) at the same time.(8) Suppose that a government allied to business interests is in office. External public borrowing occurs as a way of financing redistribution toward the groups supporting the government. The burden of the debt is left to future governments, which may be of the type supported by different groups, such as the "workers." At the same time, the possibility of a change in regime in favor of the workers' party interests implies a risk of expropriatory taxes on capital: this political risk leads to capital flight. We show that the parties supporting the interests of the capital owners and the workers would follow different policies regarding the imposition of capital controls and the choice between debt default and fiscal adjustment. The empirical evidence of several indebted LDCs is consistent with the predictions of this model.

Delayed Stabilizations

Countries often delay the adoption of stabilization programs, even when they are unavoidable given the suboptimality and/or instability of current policies. Delays are particularly inefficient when, as is often the case, the lengthier wait will increase the cost of the stabilization. For instance, explosive budget deficits financed by inflation (degenerating into hyperinflations) often are allowed to continue for prolonged periods, even though it is well understood that sooner or later taxes will have to be raised to eliminate the deficit. Sometimes it appears that "things have to get worse before they get better."

Allan Drazen and I explain delays in the adoption of stabilization programs as the result of distributional conflicts over the allocation of the fiscal burden.(9) Delays occur because different groups attempt to shelter themselves from the burden of taxation: each group, by opposing stabilization programs unfavorable to them, hopes that in the future other groups will bear a higher share of the burden. We model the stabilization process as a "war of attrition" in which each group tries to wait out the other. The more the expected distribution of the burden of taxation after the stabilization, the longer stabilizations will tend to be delayed. An uneven distribution of the tax burden is more likely in politically polarized countries.

Delays also hinge on the uncertainty about which group has sufficient economic and political resources to oppose policies unfavorable to them. When this uncertainty is removed and a group is forced to "given in," stabilizations occur: the political consolidation of the "winners" leads to the imposition of a disproportionate burden of taxation on the "losers" of the war of attrition.

In an earlier paper, I study several interwar episodes of debt adjustment policies in more detail.(10) These episodes also suggest a war of attrition. In France, Germany, and Italy, all of which experienced a period of extreme political instability and fragmentation after World War I, the fiscal adjustment was delayed, leading to bursts of inflation. In these countries, stabilization policies were implemented only when a political consolidation occurred. On the contrary, in interwar England, the relatively stable political situation with the Conservative Party in control led to an immediate postwar fiscal stabilization.

This research on "delays" is applicable to more than simply stabilization. It suggests a general politicoeconomic explanation of why efficient policies are postponed, even if everybody knows that sooner or later they will have to be adopted, and even if the more time is wasted, the higher the aggregate costs of the policy will be.

In summary, I believe that the development and testing of positive models of policy, with particular attention to political and distribitional aspects, is crucial. First, these models help us to understand empirical observations that are clearly inconsistent with models based on the social planner assumption. Second, they provide insights on how to set up institutions and policy rules to achieve the best outcome, given the policymakers' ideological and opportunistic incentives, and given the underlying conflicts of interest of different individuals.

(1)A. F. Alesina, "Macroeconomic Policy in a Two-Party System as a Repeated Game," Quarterly Journal of Economics 102 (August 1987), pp. 651-678. (2)A. F. Alesina and J. D. Sachs, "Political Parties and the Business Cycle in the United States, 1948-84," NBER Working Paper No. 1940, June 1986, and Journal of Money, Credit and Banking 20, 1 (February 1988), pp. 63-82; and A. F. Alesina, "Macroeconomics and Politics," in S. Fischer, ed., NBER Macroeconomics Annual 1988, Vol. 3. Cambridge, MA: The MIT Press, 1988, pp. 13-52. (3)A. F. Alesina, "Politics and Business Cycles in Industrial Democracies," Economic Policy 8, pp. 59-78; and A. F. Alesina and N. Roubini, "Political Cycles: Evidence from OECD Economies," manuscript, 1990. (4)A. F. Alesina, "Credibility and Policy Convergence in a Two-Party System with Rational Voters," American Economic Review 78, 4 (September 1988), pp. 796-806; A. F. Alesina and S. Spear, "An Overlapping-Generations Model of Political Competition," NBER Working Paper No. 2354, August 1987, and Journal of Public Economics 37 (December 1988), pp. 359-379; and A. F. Alesina and A. Cukierman, "The Politics of Ambiguity," NBER Working Paper No. 2468, December 1987. (5)A. F. Alesina and H. Rosenthal, "Moderating Elections," NBER Working Paper No. 3072, August 1989. (6)A. F. Alesina and H. Rosenthal, "Ideological Cycles in Congressional Elections and the Macroeconomy," NBER Working Paper No. 2706, September 1988, and American Political Science Review, (June 1989). (7)A. F. Alesina and G. Tabellini, "A Positive Model of Budget Deficits and Government Debt," NBER Working Paper No. 2308, July 1987, and Review of Economic Studies, forthcoming, and "Voting on the Budget Deficit," NBER Working Paper No. 2759, November 1988, and American Economic Review, forthcoming. (8)A. F. Alesina and G. Tabellini, "External Debt, Capital Flight, and Political Risk," NBER Working Paper No. 2610, June 1988, and Journal of International Economics, (November 1989). (9)A. F. Alesina and A Drazen, "Why Are Stabilizations Delayed?" NBER WorkingPaper No. 3053, August 1989. (10)A. F. Alesina, "The End of Large Public Debts," in F. Giavazzi and Spaventa, eds., Surviving with a High Public Debt: The Italian Experience, Cambridge, England: Cambridge University Press, 1988.