Banking privatization and market structure in Brazil: a dynamic structural analysis

• Author: Fabio Sanches,Daniel Silva Junior,Sorawoot Srisuma
• Date: 01 December 2018
• Published date: 01 December 2018
• DOI: http://doi.org/10.1111/1756-2171.12257

RAND Journal of Economics

Vol.49, No. 4, Winter 2018

pp. 936–963

Banking privatization and market structure

in Brazil: a dynamic structural analysis

Fabio Sanches∗

Daniel Silva Junior∗∗

and

Sorawoot Srisuma∗∗∗

This article examines the effects of bank privatization on the number of bank branches operating

in small isolated markets in Brazil. We estimate a dynamic game played between Brazilian public

and private banks. We find private banks compete with each other as expected. We also find

public banks generate positive spillovers for private banks. Our counterfactual study shows that

privatization substantially reduces the number of banks. The government can mitigate the effects

of privatization by providing subsidies to private banks. Our model predictssubsidy policies that

reduce operating costs are more cost-effective than entry costs for isolated markets in Brazil.

1. Introduction

The rise of the privatization of public, state-owned entities has been observed since the

1980s in many parts of the world—see Megginson and Netter (2001) for a survey on privatiza-

tion programs in several countries. The conventional argument for privatization is that private

companies are more efficient and able to deliver better products and services for final consumers.

Privatization also raises funding that can be used to finance other public areas of need. Privatiza-

tion is usually a contentious topic that generates political and social tensions. Depending on what

∗PUC-Rio; fmiessi@gmail.com.

∗∗City, University of London; danielsjunior@gmail.com.

∗∗∗University of Surrey; s.srisuma@surrey.ac.uk.

We are indebted to Martin Pesendorfer for his support and guidance during this project. We are also grateful to the

Co-Editor, Aviv Nevo, and three anonymous referees for comments and suggestions that helped improve the article.

We would like to thank Dimitri Szerman for the help with the data and for insightful comments on earlier versions of

this draft. Robinson Silva helped us organize the data. Wealso benefited from discussions with Juliano Assunc¸˜

ao, Tim

Besley,Francesco Caselli, Francisco Costa, Michael Dickstein, Gabriel Garber, Jason Garred, Matthew Gentry,Maitreesh

Ghatak, Joachim Groeger, Emmanuel Guerre, Bernardo Guimar ˜

aes, Panle Jia, Gabriel Madeira, Jo˜

ao Manoel Pinho de

Mello, Robert Miller, Mateusz Mysliwski, Ugo Panizza, Fabio Pinna, Bruno Rocha, Pasquale Schiraldi, Pedro Souza,

and Johannes Spinnewijn. Sanches gratefully acknowledges the financial support from CAPES (Brazilian Ministry

of Education) and Silva Junior gratefully acknowledges the support from CNPQ (Brazilian Ministry of Science and

Technology).The usual disclaimer applies.

936 C2018, The RAND Corporation.

SANCHES, SILVAJUNIOR AND SRISUMA / 937

is being privatized, there are several different, often subjective, aspects for evaluating the impact

of privatization.

This article studies the potential effects of privatizing public banks on small isolated markets

in Brazil. We quantify the potential impact of privatization by studying changes in the number

of bank branches.1The focus on the number of bank branches operating in small markets is of

economic relevance. Forinstance, Burgess and Pande (2005), Pascali (2012), and Bruhn and Love

(2014) have shown causal effectsof branch expansion into unbanked rural areas on the reduction

of poverty and local development. Additionally, anecdotal evidence suggests that the lack of bank

branches in isolated markets has nonnegligible economic effects in Brazil.2

Todo this, we estimate a dynamic entry game between Brazilian public and private banks. We

use data from isolated markets in Brazil for 1995–2010 to estimate the primitives of the game. We

then use the model to make predictions about the effects of privatization on the number of bank

branches. Our model predicts that the privatization of public banks would cause a significant

reduction in the total number of bank branches operating in these markets. One way that the

government can mitigate this problem is to provide subsidies, to incentivize banks to operate in

isolated markets. Our study demonstrates that subsidies to operating costs are more cost-effective

than subsidies to entry costs.

The driving forces behind the market structure of the banking sector are complex; see La

Port a, L´

opez-de-Silanes, and Shleifer (2002), Levy-Yeyati, Micco, and Panizza (2004, 2007). On

the one hand,a strand of the banking literature argues that public banks complement private banks.

Public banks focus their activities on segments of the market with high social (but low private)

returns, fostering financial development, and eventually providing conditions for the growth of

private banks. On the other hand,critics of public banks argue that they crowd-out more efficient,

more competitive private banks, thus slowing the development of the financial system.

Building on this literature, we first perform a descriptive analysis to look for evidence

whether public banks complement or crowd-out private banks in Brazil. We regress observed

activity decisions of private banks in each market/year on the number of private and public

competitors operating in the same market in the previous period. We find that the number of

private competitors in a given market reduces the probability of a private bank being active in

the same market, but the number of public banks increases this probability. This suggests that

though private banks compete with private banks, public banks appear to complement private

banks. Therefore, we expect that privatizationwould reduce the overall number of bank branches.

We then set up and estimate a dynamic entry game between Brazilian banks to precisely

quantify the effect from privatization. We assume private banks are profit maximizers. It is not

clear if public banks are necessarily profit maximizers—see Sapienza (2004), Micco, Panizza,

and Ya˜

nes (2007), Cole (2009), and Carvalho (2009). We therefore taketwo different approaches

to model public banks. In the first, we do not model the objective function of public banks. We

take their entry patterns in the data as an exogenous process. In the second, public banks are

modelled as profit maximizers analogously to private banks. We estimate the profit functions for

all profit-maximizing banks.

The model estimates of the competition/complementarity parameters for public and private

banks are qualitatively consistent with the descriptive study. For the private banks, we find the

entry and operating costs are large and account for a major share of their profit function. When we

1There are other important aspects to privatization in the banking sector. For instance, many studies focus on

the operating efficiency and profitability of banks following privatizations. See Clarke, Cull, and Shirley (2005) and

Megginson (2005) for surveys. Other authors focus on the effects of privatizationon economic g rowth.For example, see

Berkowitz, Hoekstra, and Schoors (2014) and the references therein.

2There have been several recent highly publicized bank robberies in isolated municipalities in Brazil, where the

robbers also destroyed the bank during the act of crime. Details of disruptions and effects on economic activities in

these cities have been reported, for examples, in (i) diariodonordeste.verdesmares.com.br/cadernos/regional/ataque-a-

bancos-em-choro-causa-prejuizos-ao-comercio-1.1398310 and (ii) g1.globo.com/sao-paulo/itapetininga-regiao/noticia/

2016/03/comerciantes-reclamam-de-prejuizos-com-explosoes-em-bancos.html.

C

The RAND Corporation 2018.