Fiscal policy and (re-)elections: a dangerous liaison for Latin America?

Sebastian Nieto-Parra, Javier Santiso 22 December 2009

a

A

The global financial crisis has placed fiscal policy squarely back at the centre of public policy debate. Studies show that the benefits of fiscal stimulus in emerging countries can be limited, with fiscal stimulus having a greater effect on GDP growth in high income countries than in developing countries (Ilzetzki, Mendoza and Vegh 2009). Emerging countries should therefore be cautious about implementing such fiscal packages (Reinhart and Reinhart 2009). These warnings may be particularly important to Latin America.

New research

In Latin America, political and fiscal cycles have historically been closely linked, often with fiscal deteriorations occurring prior to elections. Now is the time to revisit the topic of political budget cycles in Latin America. Starting with Chile and closing with Mexico and Venezuela in December 2012, no fewer than 16 presidential elections due to take place.

In recent work, we study these issues by measuring presidential elections that may in turn affect fiscal balances (Nieto-Parra and Santiso 2009). We ask the following questions:

  • To what extent do Latin American governments expand fiscal expenditures around elections to attract voters?
  • Do different types of electoral systems have different effects on fiscal policies around elections?
  • Is Latin American fiscal policy around elections maturing?

New evidence on fiscal policy and elections in Latin America

Elections matter and can affect economic policies. Specifically, some components of fiscal policy may be expanded around elections in order to attract voters. Politicians’ behaviour can also be influenced in political systems permitting re-election of officials. Incumbents can artificially expand economic activity during election years to improve their chances of re-election. Unfortunately for incumbents, however, empirical evidence shows that expansionary fiscal policy around elections does not always bring the “pay off” of re-election (Brender and Drazen 2008).

Our study of the period 1990-2006 suggests that in Latin American countries the average primary balance declines by close to 0.7% of GDP during an election year (Figure 1). Most of this movement is due to the expenditure component. Within this, it is current rather than capital expenditure that is most affected. By contrast, in OECD countries, observed changes in the primary balance and current expenditures during election years are less than 0.1% of GDP for either measure (Nieto-Parra and Santiso 2009). This difference between Latin American and OECD countries is all the more remarkable considering the relatively small size of government in Latin American democracies.

Figure 1. Impact of elections on fiscal policy in Latin American and OECD countries
 
(Changes in selected fiscal indicators, percentage of GDP)

Note: The impact of elections on fiscal policy is calculated as the difference between the fiscal variable (as percentage of GDP) during the election year and non-election years. Source: Nieto-Parra and Santiso (2009)

We note considerable differences across Latin American countries and across time. In Brazil, the impact of elections on primary expenditure as well as on capital expenditure prior to elections is higher in the election of 1998 than that of 2002. For the case of Venezuela, in the last two presidential elections, primary expenditure over GDP has increased more than in earlier elections. By contrast, in Chile, observations suggest that there is not an impact of elections on fiscal policy.

Political systems allowing re-election increase candidates’ propensity to use fiscal policy to attract votes

The current trend in election reform in Latin America is moving towards a system less restrictive to re-election. Venezuela in 1998, Dominican Republic in 2002, Colombia in 2005 and Ecuador in 2008 have adopted immediate re-election in recent years, as exists in some OECD presidential regimes such as the US. This will be a crucial factor in upcoming electoral cycles.

Immediate re-election can be associated with high incentives for incumbent candidates to increase the components of government expenditure that are observed by voters, who can then be influenced by these fiscal measures.

Additionally, the debate in some Latin American countries has focused on promoting immediate re-election, for example in Bolivia and Nicaragua, and even allowing re-election for more than two consecutive mandates as in Colombia and Venezuela. More importantly, the 2009 Honduran coup d'état to the president Manuel Zelaya occurred in the context of an ongoing dispute over a new constitution that would allow immediate re-election.

Immediate re-elections have a considerable impact on the expenditure side of the fiscal balance. Results show that immediate re-elections increase primary expenditure by more than 1.2% of GDP. This result contrasts with the case in which immediate re-election is not allowed (less than 0.2% of GDP). Moreover, the impact of immediate re-elections on primary expenditure is more than twice the impact in the case of elections as a whole (Nieto-Parra and Santiso 2009). The findings suggest that incumbent candidates in Latin American democracies considerably increase the components of government expenditure which may be observed by voters. The opposite is true when there is non-immediate re-election and when re-election is forbidden.

Is Latin American fiscal policy around elections maturing?

The year 2006 set a record for presidential elections in Latin America – more than in 1994, with 10 presidential run-offs and over 80% of the region's population heading to the polls. But primary surpluses tended to grow rather than shrink. Part of this result can be ascribed to the more forgiving conditions of high real GDP growth. Certainly, spending restraint was not driving low deficits. Many of those same countries witnessed some increases in primary expenditure as a share of GDP.

By comparing the 2005-2006 electoral cycle with prior electoral cycles, we note a slight improvement of fiscal management around elections. Prior to 2005, primary expenditures increased significantly while for 2005-2006 the rise is much less pronounced on average. Similarly primary fiscal balances tend to be negative prior to 2005 and positive in more recent years (Figure 2). Despite this, even today we find that Latin American countries still increase public expenditure during elections, showing that fiscal discipline around elections remains an issue for the region (Nieto Parra and Santiso 2009).

Figure 2. Changes in current expenditure and primary balance prior 2005 and for the year 2005-2006 (Percentage of GDP)

 

Note: The impact of 2005 and 2006 elections on fiscal policy is calculated as the difference between the fiscal variable (as a proportion of GDP) during the election year and prior non-election years. The impact of elections prior to 2005 on fiscal policy is calculated as the difference between the average of the fiscal variable during elections and the average of this variable for non-elections years. Source: Nieto –Parra and Santiso (2009).

Conclusion

Presidential elections are associated with much greater changes to the major components of fiscal policy in Latin America than in high-income countries. Most of this movement is due to the expenditure component and within this it is current expenditure that is most affected. Our results also suggest that political systems allowing immediate re-election in Latin America have a considerable impact on the expenditure side of the fiscal balance. Finally, by comparing the 2005-2006 electoral cycle with respect to prior electoral cycles, we note a slight improvement of fiscal management around elections.

Several policy implications can be derived from our findings.

  • First, fiscal rules focusing on the stability of the major components of fiscal balance may be an appropriate policy measure for Latin American countries.
  • Second, in the current debate on electoral systems in some Latin American countries, the argument for immediate re-election systems should be tempered by acknowledging the risks of fiscal deterioration around election years such systems appear to cause.
  • Third, special bodies at the Ministries of Finances aiming to look at a long-term perspective on fiscal policy could be established. This would serve to protect the independence of fiscal policy from short-term incentives of politicians.

The policy options put forward may serve to mitigate the potential fiscal disruptions ahead of elections, which are, after all, regular and normal events in democracies.

References

Brender Adi and Allan Drazen (2008), “How Do Budget Deficits and Economic Growth Affect Reelection Prospects? Evidence from a Large Panel of Countries”, American Economic Review, Vol. 98, pp. 2203-2250.

Nieto-Parra, Sebastian and Javier Santiso (2009), “Revisiting Political Budget Cycles in Latin America”, Development Centre Working Paper No. 281, OECD, Paris.

Ilzetzki Ethan, Enrique Mendoza and Carlos Vegh (2009), “How big are fiscal multipliers? New evidence from new data”, VoxEU.org, 1 October.

Reinhart, Carmen and Vincent Reinhart (2009), “Is there scope for fiscal stimulus for debt-intolerant countries”, VoxEU.org, 22 August.

a

A

Topics:  Politics and economics

Tags:  elections, fiscal policy, government spending

Economist at the OECD Development Centre

Professor of Economics at ESADE Business School

Events