Abstract. This paper examines the Market for loans from
banks to micro- and small enterprises in Trinidad and Tobago.
It tests for the presence of racial and gender discrimination.
It takes the reported refusal to grant loans to groups, when
all other indicators of credit-worthiness are taken into account,
as evidence of discrimination. The paper distinguishes between
Application and Denial Rates. It finds that, once all other
indicators of credit-worthiness are taken into account, neither
Application nor Denial rates differ significantly by gender.
Reported Denial rates are, however, higher for Africans
compared with other ethnic groups, implying the possible
presence of discrimination.
This paper examines the market for loans from
banks to small and micro enterprises in Trinidad
and Tobago. It examines whether there is evidence
that certain types of businesses are less likely to
receive loans than others- most notably businesses
owned by females and by those of African origin.
In short, is there discrimination by banks against
The paper views discrimination as occurring
when the distribution of loans is influenced by
factors not relevant to the transaction – in this case
race and gender. Its key analytic distinction is to
recognise that the distribution of loans is influ-
enced by applications received, as well as deci-
sions made by the bank. Confirming the results
of Raturi and Swamy (1999), it finds some
evidence of racial and gender differences in
Application and Denial Rates. Discrimination,
however, is only defined to occur when
race/gender influence Denial rates independently
of other risk indicators. Data limitations, however,
preclude an assessment of whether ethnicity or
gender influence borrower discouragement.
In contrast with Raturi and Swamy it finds, for
Africans, evidence of higher Denial rates, implying
the presence of discrimination. It finds no
evidence of higher Denial rates for females.
The paper is structured as follows. It begins
with the analytical framework for testing for dis-
crimination, distinguishing between Application
and Denial Rates, generating five hypotheses to be
tested. It then describes the data set to be used in
the analysis. The paper provides simple univariate
statistics on Loans, Application and Denial Rates,
and how these vary according to firm and owner
characteristics. The bulk of the analysis is devoted
to testing the hypotheses in a multi-variate frame-
2. Literature review and hypotheses derivation
Becker (1957) defined financial discrimination to
exist when loans, normally made at interest rate
r, are made to the discriminated group at r(1 +
where δ is the interest premium. However Stiglitz
and Weiss (1981) show financial institutions
favour outright loan refusal to high risk applicants,
rather than making loans at higher interest rates.
Discrimination may therefore also be reflected in
a refusal to grant loans (Denial) rather than an
To demonstrate the presence of discrimination
requires more than the observation that specific
groups (G) – which we shall take as Africans/
Females – are consistently less likely to have loans
than other groups. Only after all other Non-
Racial and Gender Discrimination in the
Micro Firms Credit Market?:
Evidence from Trinidad and Tobago
D. J. Storey
Small Business Economics 23: 401–422, 2004.
2004 Kluwer Academic Publishers. Printed in the Netherlands.
Final version accepted on December 11, 2002
Centre for Small and Medium Sized Enterprises (CSME)
University of Warwick
Coventry CV4 7AL