The electoral bias: the political economy of subnational transfers in Latin America

ABSTRACT

This article examines whether transfers to local governments in Brazil, Mexico, Colombia and Chile have been allocated to obtain electoral advantage. A large panel data set and fixed-effects estimations uncover two types of manipulations: grant fluctuations along the municipal election cycle and biases towards aligned municipalities. Notwithstanding, there are significant cross-country differences. In Brazil, Colombia and Chile, mayors aligned with the central government coalition systematically benefit, especially ahead of elections, whereas in Mexico, political budget cycles do not discriminate in terms of partisanship. These results point to institutional conditions and the nature of electoral competition shaping distributive politics.

KEYWORDS:

• distributive politics
• intergovernmental transfers
• electoral competition
• political budget cycles
• local governments
• Latin America

JEL:

• H72
• H77
• O54
• R58

ACKNOWLEDGEMENTS

A preliminary version of this paper was presented at a workshop at the University of Talca, Santiago de Chile, 28 March 2019. The authors thank Bernardo Lara, Alejandro Corvalán, Paulo Cox and the other participants at the workshop for helpful suggestions. The authors also thank the two anonymous referees for their accurate comments throughout the review process.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. The binomial electoral formula created districts that elected only two representatives, forced partisan lists to run only two candidates per district, and assigned both sits to the winning list only if they obtained twice the vote of the runner-up list (Valenzuela et al., 2018, p. 136).

2. States can choose from two formulas. The first includes weighted indicators in several dimensions and compares them with reference values to build a household-level poverty index. This index is aggregated at municipal and state levels and each state receives a proportional amount of the fund according to its share. Later, states redistribute funds based on the formula computed in each municipality. Alternatively, states can choose a simpler formula that entails four equally weighted components. Most states have opted for the latter formula.

3. Because in Chile the regional intendente is directly appointed by the president and represents its interests, there are no grounds for controlling this intermediate institutional level.

4. Since in Mexico mayors can only serve one mandate, there is no need to include this control.

5. Although GMM is attractive to handle regressors’ potential endogeneity, it also introduces discretion when treating instruments and lags.

6. We calculate this value as the difference between being from the coalition parties, or not: Value = Aligned Mayor + (Aligned Mayor × Y₀). In Chile: 0.119 + 0.102 = 0.221. In Colombia: 0.0762–0.0285 = 0.047.

7. Because in Mexico local polls are held in different years depending on the state, our model does not allow the identification of the effect of presidential elections.

Additional information

Funding

This research was supported by the Regional Studies Association (RSA) Membership Research Grant (MeRSA) 2017 awarded to Xabier Gainza for the Research Project ‘The Electoral Bias: Distributive Politics and Local Development in Brazil, Mexico, Chile and Colombia’, for the period 1 December 2017–31 May 2019.