Ethnic Fractionalization, Governance and Loan Defaults in Africa
| DOI | http://doi.org/10.1111/obes.12152 |
| Author | Panicos Demetriades,Svetlana Andrianova,David Fielding,Badi H. Baltagi |
| Published date | 01 August 2017 |
| Date | 01 August 2017 |
435
©2017 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 79, 4 (2017) 0305–9049
doi: 10.1111/obes.12152
Ethnic Fractionalization, Governance and Loan
Defaults in Africa*
Svetlana Andrianova,†Badi H. Baltagi,†,‡ Panicos
Demetriades† and David Fielding§
†Department of Economics, Leicester University, University Road, Leicester, LE1 7RH, UK
(e-mail: sa153@le.ac.uk, pd28@le.ac.uk)
‡Department of Economics, Syracuse University, 110 Eggers Hall, Syracuse, NY 13244,
USA (email: bbaltagi@maxwell.syr.edu)
§Department of Economics, University of Otago, PO Box 56, Dunedin, 9054, New Zealand
(e-mail: david.fielding@otago.ac.nz)
Abstract
We present a theoretical model of an imperfectly competitive loans market that is suitable
for emerging economies in Africa. The model allows for variation in both the level of
contract enforcement (the quality of governance) and the degree of market segmentation
(the level of ethnic fractionalization). The model predicts a specific form of nonlinearity in
the effects of these variables on loan default. Empirical analysis using African panel data
for 110 individual banks in 28 countries over 2000–08 provides strong evidence for these
predictions. Our results have important implications for the conditions under which policy
reform will enhance financial development.
I. Introduction
In terms of financial development, Africa still lags behind other parts of the world.African
banks are deterred from lending in domestic markets by a lack of creditworthy borrowers,
and loan volumes are highly sensitiveto def ault rates (Andrianova et al., 2015). As a result,
many African banks are excessively liquid and channel an unusually large proportion of
domestic savings abroad, although there is substantial variation in banks’ default risk and
asset structure, both within and between countries (Honohan and Beck, 2007). At the
same time, firm and household surveys reveal endemic credit constraints. For this reason,
understanding the determinants of the rate of loan default is crucial in overcoming obstacles
to financial developmentin Africa. Our paper makes a first step in this direction by providing
both theory and evidence that shed new light on the factors behind the high rate of loan
default in many parts of Africa.
JEL Classification numbers: G21, O16.
*Weacknowledge the support of ESRC-DFID grant number ES/J009067/1 and thank two anonymous referees for
their comments. All errors are our own.
436 Bulletin
The focus of attention in both the theoretical and empirical parts of the paper is on
two key characteristics of the African banking sector. Firstly, there is a great deal of
variability across Africa in terms of both the level of corruption and the quality of contract
enforcement. This variability is revealed in indices of the quality of governance produced
by organizations such as the World Bank, Transparency International, and the Bureau van
Dijk. Daumont et al. (2004) argue that weak contract enforcement in Africa is due to a
number of factors, including excessively time-consuming and unwieldy legal procedures,
high litigation costs, a lack of appropriately qualified judges, and inadequacies in the
cadastral system that inhibit the identification of collateral property. The large variation
in the magnitude of these problems across Africa suggests that they may help explain
the observed variation in the incidence of loan default. Secondly, many African countries
are characterized by a high level of ethnic fractionalization (Easterly and Levine, 1997).
A lack of trust between different ethnic groups is likely to generate high inter-ethnic
transactions costs, which will lead to a high level of market segmentation (Aker et al.,
2010; Robinson, 2013). Existing studies of ethnic market segmentation have not focused
explicitly on financial markets, but there is no reason to suppose that financial markets are
any less susceptible to this problem than others. Even in countries with a large banking
sector, ethnic fractionalization is likely to create monopolistic competition, because banks
are differentiated by the ethnicity of their staff. African banks are therefore likely to face
unusually inelastic demand curves.
Our paper builds on Andrianova et al. (2015), who show that loan defaults are a major
factor inhibiting African banks lending when institutional quality is low, but do not give
any evidence on the causes of high loan default. Our paper presents an empirical analysisof
the causes of high default rates which is informed by a theoretical model of an imperfectly
competitive banking sector. Our model is a refinement of the model in Andrianova et al.
(2015), capturing the nature of market segmentation in a more realistic way.1
One feature of this type of model of imperfectly competitive banking which is not
spelled out in Andrianova et al. (2015) is that it entails some specific predictions for the
way in which the quality of contract enforcement and the degree of market segmentation
interact in generating a certain level of loan default. The theory implies that marginal
improvements in enforcement quality will reduce loan default rates only when the con-
tract enforcement problem is initially neither much more severe nor much less severe than
the market segmentation problem. This can help explain why in some circumstances im-
provements in enforcement quality have a large effect on loan default rates, but in other
circumstances there may be little or no effect.
In this paper, we draw out the predictions of the theoretical model regarding the in-
teraction between the contract enforcement and market segmentation problems, and then
present an empirical model of loan default rates in a panel of African banks which focuses
1In Andrianova et al. (2015), bank differentiation was captured using Salop’s ‘circular city’ framework. In the
model here, differentiation is captured using the ‘linear city’ framework of Hotelling (1929). Using a ‘linear city’
framework, the level of differentiation between two banks is captured by the distance between them in a single
dimension, and moving one bank further along this line unambiguously increases the level of differentiation. We
believe that this second feature, which is not present in ‘circular city’ models, is a more realistic characterization
of differentiation arising from ethnic fractionalization. The key nonlinearities predicted bythe ‘circular city’ model
are also predicted by the ‘linear city’ model, but the finding that the model’s predictions are robust with respect to
differences in the topology of bank differentiation is a result which is newto this paper.
©2017 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial