Payday lenders and economically distressed communities: A spatial analysis of financial predation

Abstract

The payday lending industry has grown considerably in recent years, reflecting both widespread economic insecurity and market neglect by the traditional banking sector. Outlets are now commonplace in many communities across America. Accused by many of predatory practices, payday lenders can be viewed as financial hazards in already economically distressed communities. Using a Geographic Information System (GIS) and associated statistical analyses, this paper examines the social ecology of payday lending along the Front Range communities of Colorado. Comparison of means and logistic regression results reveal how communities’ composition by class, occupation, race/ethnicity, nativity, age, and military affiliation affect their likelihood of hosting payday lending.

Acknowledgements

The authors wish to thank the anonymous reviewers of The Social Science Journal for their helpful comments and suggestions.

Notes

1 The Colorado Deferred Deposit Loan Act (“DDLA”) exempts payday lenders from Colorado usury laws, granting payday lenders the right to charge 20% on the first $300 loaned, plus 7.5% of any amount loaned in excess of $300 (Graves and Peterson, 2005). Payday lenders are asked not to renew loans more than once, but they may refinance the loans. While some cities around the country have taken action to limit payday lending, no such city-level ordinances currently exist among the cities comprising our sample.

2 Colorado's legislature attempted, unsuccessfully, to place limits on the industry in the spring of 2008. The proposed legislation would have set a maximum finance fee of $30 per year and an APR cap of 45% (Brown, 2008).

3 In general, our results are not sensitive to alternative specifications for our buffer.

4 Although the data for several variables are not normally distributed, non-parametric tests returned a similar pattern on findings. This is in part due to the large sample size of block groups in this study.

5 Analyses of this sort raise the prospect of spatial autocorrelation. While it can be reasonably argued that the spatial clustering of variables is what is of interest in studies such as this, it is also important to note that spatial dependence may result in findings that overstate the significance of multivariate relationships. The need to employ highly specialized techniques to deal with the issue in this kind of research remains a subject of methodological debate (Pastor et al., 2004). In fact, most environmental inequality studies that use a similar research design as this study ultimately side-step the issue for practical reasons, namely the sheer logistical obstacles associated with performing the necessary calculations on large data sets (Morello-Frosch et al., 2001; Pastor et al., 2004). Moreover, while such techniques may impact the significance level of individual variables, they often do not alter the overall pattern of results. Given these considerations, this study follows many other spatial analysis studies in using standard regression methods.