Wednesday, 2 July 2014

Political economy of debts and deficits (3): Hypotheses explaining debt accumulation

Continuing with the overview of the politics behind debt and deficits, in this post I present the dominant hypotheses aimed at explaining the vast debt accumulation on the past 40 years in many industrial economies. 

The debt dynamics in the past few decades has prompted a new stream of research that focused on political economy of public deficits and debt. A useful overview of this literature can be found in two papers by Alberto Alesina and Roberto Perotti (1994) "The Political Economy of Budget Deficits", and Alesina and Perotti (1996) "Budget Deficits and Budget Institutions".

The following six groups of models (presented in the above-mentioned papers) try to explain the political economy of the budget process:
  1. Models that focus on opportunistic behavior of policy-makers and the naivety of voters, who are subject to fiscal illusion; 
  2. Models of intergenerational redistribution; 
  3. Models of the public debt as a strategic variable; 
  4. Models of distributional conflicts; 
  5. Models of geographically dispersed interests; and 
  6. Models emphasizing the role of budget institutions. 
1. Models of opportunistic behavior and the fiscal illusion

The theory of opportunistic behavior of politicians and voters suffering from illusion suggests that politicians are inclined to increase spending more than taxes close to election time, so as to cure favors with voters who do not understand that increased taxation would have to follow. Voters overestimate the benefits of current expenditures and underestimate future tax burdens. Opportunistic politicians who seek to be elected will take advantage of this by running deficits in pre-election time to boost output and please the voters.

This is linked to models of political business cycles (originally presented in Nordhaus, 1975), where politicians will tend to create expansions before the elections (deficits spending to increase GDP growth above normal and lower unemployment) only to see a recession coming immediately after as an effect of tightening monetary and fiscal policy. Incumbents are reappointed if growth is high and unemployment is low in the election year. Voters are retrospective (backward looking) and highly discount the past (meaning that recent events are more important to them than past events – in other words, they will react more to events that happened in the election year than in the beginning of the incumbent’s mandate, 4 years ago). An empirical implication is that election years should be characterized by above normal growth and below normal unemployment with a moderate increase of inflation, whereas after the election the opposite will occur. The long term effect is higher debt and possibly higher inflation. However empirical evidence has failed to find any substantial support for this kind of behavior in the economy vis-à-vis the electoral cycle. A famous paper is Peltzman’s (1992) “Voters as fiscal conservatives” where he found that US voters tend to punish politicians which engage into deficit spending.

The traditional opportunistic models were later expanded by introducing rational expectations (Rogoff, 1990; Persson & Tabellini, 1990), where voters are no longer retrospective but forward-looking, and have introduced the government competence parameter (thus introducing the adverse selection problem).

Opportunistic models have also been contrasted with the so-called partisan models (Hibbs, 1977) where it is said that unemployment is persistently higher with right-wing parties in power and lower for left-wing ones, while inflation is permanently higher under left-wing governments and lower for right-wing ones.

One possible weakness of this theory is that voters aren't always retrospective, and don't always underestimate the tax implications of spending decisions (i.e. they don't always succumb to the fiscal illusion, over time they learn and accommodate). However empirically it has been shown in many economic voting models that voters are in fact much more retrospective (past-performance oriented) than prospective (expectations of future performance), and far more myopic, leading them to persistently succumb to the fiscal illusion and overestimate deficit-spending in good economic times.

The crucial problem with the theory is the failure to explain the cross-country differences. How come in some countries voters persistently suffer under the fiscal illusion while in others they don't? Are they smarter or better informed? Not likely. Different tax systems and different fiscal rules can explain the cross-country differences. This also implies that opportunist behavior of politicians and voter fiscal illusion aren't the only reasons behind high debt. They can perhaps explain it to some extent but different institutional settings are much more precise in explaining why some countries will face high debts and deficits and why others will not.

2. Models of intergenerational redistribution

The models of intergenerational re-distribution start from the fact that only the currently living people vote. The current generation can be selfish and leave debt to new generations without any repercussions against high debts. The only constraining factor for the selfish generation is the increased risk of default, which might affect their pension plans. For example, if the current generation is reckless and doesn't punish politicians which accumulate unsustainable debt levels, by the time they become retired the risk of government default will affect their pensions (either via austerity measures or via a literal default where the government simply won’t have any more money to service its existing liabilities).

In conventional macroeconomics, debt and pay-as-you-go social security are identical policies. Preferences for social security depend on age-earning profiles and the population growth rate, as well as intergenerational altruism. The degree of altruism varies across households. Richer households leave money for their future generations (they bequest) while poorer households are unable to do so – we can call them bequest-constrained. Voters from richer households are concerned only with the general equilibrium effects of fiscal policy and deficits and not how it redistributes across generations. But the bequest-constrained voters favor a budget deficit, because it allows them to do something they cannot do privately: redistribute resources toward themselves. In a median voter equilibrium, the size of the budget deficit depends on the efficiency effects and the number of bequest-constrained voters (from poorer households).

The politicians seize this opportunity and act to only increase the borrowing category to finance their deficits and rising government spending. Taxes are not a viable option since they risk no re-election, while cutting spending is arguably even worse for the politicians as this directly affects particular groups which could, as an effect, lower their support for the incumbent party.

However this too fails to explain the cross-country heterogeneity. In addition high public debts have often been both accumulated and sharply reduced during a lifetime of a single generation.

3. Models of the public debt as a strategic variable

In case of strong conflicts between two main political parties, public debt might become a strategic variable in the sense that one party might amass debt to constrain the actions of the other party.

This model (coined by Alesina and Tabellini, 1990) was applied primarily to US politics, characterized by a polarized electorate which is faced with only two viable options for holding office. In this model, we need to assume the existence of two parties with different preferences over public spending. The "defense" party (Republicans) runs large deficits during its term in office to prevent the "welfare" party (Democrats) from increasing spending when it comes to the power. The larger polarization between the parties about the composition of government spending, the stronger might be the incentive to leave high debt to successors to deal with it. These effects might become stronger with the increased instability of governments. For example, Alesina and Tabellini suggest that more frequent changes of governments in OECD countries since early 1970s were one of the reasons for debt expansion (see Alesina and Tabellini (1990) "A Positive Theory of budget Deficits and Government Debt", Review of Economic Studies 57, pp. 403-14)

A similar model (Persson and Svensson, 1989“Why a Stubborn Conservative Would Run a Deficit: Policy with Time-Inconsistent Preferences”) is when the two parties are in disagreement not in the composition of spending but in size of spending. This model predicts that only the low spending party increases debt since by lowering taxes and issuing debt the low spender constrains future spending. The new party in power must amount to lowering the debt burden due to a costly threat of default.

There is also the following conundrum: parties can use debt strategically to win elections. If we assume that the left wing party is more prone to a default since it’s mostly the upper classes and the banks which are debt holders, the right wing party can increase debt so as to make more people debt holders, and as a result the left would lose support.

This hypothesis can explain the rising debt levels that started since the 70-ies as it is in this period that most Western governments experienced more frequent changes of governments in power. In addition by using political polarization as an explanatory variable we can also explain why in some countries debt levels are higher while in others they’re not.

Furthermore empirical evidence has found support for the Persson and Svensson model, where right wing governments issue more debt when facing a probability of a defeat (e.g. the Reagan deficit effects on the Clinton government), while left wing governments do the opposite.

In the next and final blog post on this topic, the final three hypothesis will be presented along with some concluding remarks and a short literature overview for those more interested in the topic. 

1 comment:

  1. thanks for this short discussion. I found it most useful. I need to find the second half of this blog post. I am particularly interested in what you have to say about budget institutions.

    ReplyDelete