Banks and Payday Lenders: Friends or Foes?

  • James R. Barth
  • Jitka Hilliard
  • John S. JaheraJr.
Article

Abstract

This paper investigates the geographic distribution of payday lenders and banks that operate throughout the United States. State-level data are used to indicate differences in the regulatory environment across the states. Given the different constrains on interest rates and other aspects of the payday loan products, we empirically examine the relationship between the number of payday lender stores and various demographic and economic characteristics. Our results indicate that number of stores is positively related to the percentage of African American population, the percentage of population that is aged 15 and under and the poverty rate. The number of stores is also negatively related to income per capita and educational levels.

Keywords

Payday lending Small loans Credit issues 

JEL

D18 G21 G23 

Introduction

Most people find it convenient, and in many cases essential, at one time or another to borrow money to cover a variety of expenditures. This is almost always the case for big ticket items like homes or automobiles. Of course, such borrowers are charged an interest rate, taking into account the costs and riskiness associated with a loan, to compensate the lender for the service provided. To prevent what some consider to be unreasonable or excessive interest rates, most states have established usury laws that set maximum rates that can be charged on specific types of consumer loans, usually rates that are less than 25 percent. Banks are the most heavily involved and widely known of all the different types of financial firms when it comes to offering such loans. They are facilitated in this regard by relatively recent changes in the law, which allow banks today to open or acquire branches anywhere they wish throughout the Unites States.

There are people who also borrow money but in quite small amounts and for exceedingly short periods of time. Instead of being charged an interest rate with recurring interest payments borrowers are charged a flat fee, such as $15 per $100 borrowed. The type of financial firms offering such loans are commonly known as payday lenders and the product offered is most commonly known as a payday loan.1 A payday loan is typically structured with a due date that coincides with the borrower’s next payday, which is most frequently 2 weeks. An individual obtaining a payday loan is required to provide a post-dated personal check to the lender or an authorization to electronically debit the person’s deposit account for the loan amount and associated fee. The borrower usually agrees to return to the store of the payday lender when the loan is due to make payment. If the borrower does not do so, the lender has the option of depositing the person’s check or initiating an electronic withdrawal from the person’s deposit account. To obtain a payday loan, an individual typically only needs a bank account and a job (i.e., a regular source and minimum level of income). In comparison to banks, payday lenders have one advantage as well as one disadvantage. The advantage is they are allowed to charge flat fees that when converted to interest rates always exceed the usury rate, while the disadvantage is payday lenders are restricted as to where they can open or acquire stores in the United States.

Payday lenders are frequently a source of controversy centering on the fees they charge and their typical customer base. Consider the allowable fees on payday loans in two states. A fee of $15 for a $100 loan to be repaid in 14 days, which is allowed in Indiana, is equivalent to an annual percentage rate of 390 percent. A larger fee of $75 for the same loan, which is permitted in Missouri, translates into an annual percentage rate of 1,950 percent.2 Some individuals consider the fact that persons borrowing money pay such high interest rates to be an outrage. This is no doubt a factor in the decision of some states to either explicitly prohibit this type of financial firm from offering such loans or to do so de facto by setting much lower interest rate caps on small loans. In addition, there is a concern by some that payday lenders may engage in so-called predatory lending by locating their stores in geographical areas with higher percentages of people in poverty, lower-income individuals, unemployed and less-educated individuals. These are the groups that may fall prey to the less scrupulous firms as well as suffer more of a hardship when confronted with the high interest rates associated with payday loans. There is a related concern that the same geographical areas consist of disproportionately high percentages of African Americans and Hispanics. Indeed, it is reported that Senator Sherrod Brown, at a recent Senate Banking Committee hearing, said “… he was concerned that payday companies are marketing their high-cost loans to the very people who can least afford them, much like predatory mortgage lenders did in the run up to the housing crisis.”3

The purpose of this paper is to examine in some detail the different business characteristics of the payday lenders that operate throughout the United States. The examination is based on state-level data to emphasize differences in the regulatory environment across the states that constrains the prices and other aspects of the loan products that the firms are allowed to offer. Importantly, an empirical analysis is conducted to determine the extent to which the number of stores operated by payday lenders in the different states are related to various demographic and economic characteristics of those states in an attempt to address the concerns noted above.4 Since banks also offer consumer loans, but mainly to a different clientele and on different terms, the analysis also examines whether there is an association between the location of the branches of banks and the location of the stores of payday lenders. Some banks do, however, offer deposit advances to customers, which are typically structured as short-term loans but without a predetermined repayment date. This product is only offered to existing customers and the bank can take action to be repaid by debiting incoming electronic deposits before paying the customer’s other transactions. To the extent that payday lenders and banks locate offices in the same geographical areas, there may be an opportunity for some individuals to switch from being customers of payday lenders to banks and thereby obtain lower-cost loans.5

The remainder of the paper proceeds as follows. In the next section, a review of selected related literature is provided. This is followed by an overview of the payday lending industry, emphasizing two somewhat unique issues that arise when studying this industry. The third section presents and discusses our approach to analyzing some of the determinants of the location and concentration of payday lending stores operating in the various states throughout the United States. The last section focuses on plans for future research on payday lending.

Selected Literature Review

Much of the existing literature provides results that reinforce the view that payday lending is indeed predatory by targeting economically struggling and less educated individuals in the United States. Of course, there are other studies that report benefits associated with payday lending, such as fewer and more costly bounced checks and bankruptcy filings. In a relatively early study, Stegman and Faris (2003) analyze a database of 142 (165) payday lenders operating 807 (902) outlets for the year 1999 (2000) in North Carolina. Their data shows that there were double-digit increases in the number and value of deferred deposit checks as well as the payday loan fees collected over the 2 years. During the same period, it is noted that net charge-offs increased by 54 percent reflecting the higher risk of such loans. Their results indicate that lower-income African Americans are more than twice as likely to have taken out a payday loan as White non-Hispanics. Interestingly, they find that Hispanics are less likely to utilize payday loans than other groups. Older individuals, however, were less likely to be found as customers of payday lenders than younger individuals. Furthermore, the results indicate that the number of banks and thrifts in a household’s neighborhood has a small but significantly negative effect on the use of payday lenders.

Morgan and Strain (2008) also perform an examination of payday lending, focusing on Georgia and North Carolina, two states that had banned such loans in 2004 and 2005, respectively. Based upon an analysis of data for returned checks at Federal Reserve processing centers from 1997 to 2007, complaints filed with the Federal Trade Commission (FTC) between 1997 and 2007, and bankruptcy fillings between 1998 and 2007, they found that compared with households in states where payday lending is permitted, households in Georgia have bounced more checks, complained more to the FTC about lenders and debt collectors, and filed for bankruptcy protection at a higher rate. In a related nationwide study, Morgan et al. (2012) find some evidence that bankruptcy rates decrease after payday loan bans, but at the same time complaints against lenders tend to increase. Moreover, the authors report that their most robust finding is that returned check numbers and overdraft fee income at depository institutions decline when payday credit supply expands.6

Rather than focus on an entire state or states, Gallmeyer and Roberts (2009) conduct a study of payday lenders in the Front Range area of Colorado. An analysis is then performed on the socio-demographic characteristics of the communities, as measured by median household income, the percent of the population falling substantially below the federal poverty line, and the labor force composition. The authors find that payday lenders are more likely to concentrate in neighborhoods that have lower income, moderate poverty and higher percentages of ethnic minorities, immigrants, young adults, elderly, military personnel, and those working in non-management or professional occupations.

In another study focusing on two states, Zinman (2010) examines some of the effects of restricting access to expensive credit using data from two phone surveys conducted in 2007 of 1,040 payday borrowers residing in Oregon and Washington. Oregon imposed a binding rate cap on such credit in that year, whereas the neighboring state of Washington did not. Zinman finds that access to payday loans declined in Oregon relative to Washington, while many borrowers in Oregon shifted into plausibly inferior substitutes. In a related and more recent study, Carrell and Zinman (2014) analyze the impact of payday loan access on three different measures of military job performance in 35 states that both allow and prohibit payday lending for the period 1995 to 2007. Their empirical results indicate that payday loan access adversely affects overall job performance, retention and readiness.

Combining household survey data and county-level data for 13 states, three of which prohibit payday lending, Melzer (2011) examines whether payday loan access mitigates financial distress, as some claim. His results indicate that access to payday lending stores leads to increased difficulty paying mortgage, rent and utilities bills as well as delaying needed health care. Morse (2011), like Melzer (2011), also examines whether payday lending exacerbates or mitigates financial distress. Specifically, he considers whether the adverse effects of natural disasters on home foreclosures and small property crimes are mitigated when individuals have access to payday lenders, His analysis is based on data at the zip-code level for California over the period 1996 to 2002. In contrast to Melzer, however, Morse finds that payday lenders offer a positive service to individuals facing unexpected financial distress. Bhutta (2014) uses zip-code business data to analyze the socioeconomic factors correlated with payday lender concentration. Unlike the two studies that find both positive and negative effects of payday loans on financial well-being, his empirical results indicate little connection in terms of such loans and credit scores.

Quite recently, the Consumer Financial Protection Bureau (CFPB) (2013, Burke et al. 2014) that was established by the Dodd-Frank Act in 2010 has devoted attention to payday lending. The focus of its two “white papers” is on the long-term use of short-term loans evidenced by a pattern of repeatedly rolling over or consistently re-borrowing by individuals. In the 2013 white paper, the CFPB found that the median amount borrowed was $350 with about a third of borrowers having six loans or fewer and a total dollar amount borrowed of $1,500 during the year-long period. In the 2014 white paper, using the same data as in the 2013 study, the CFPB found that approximately 80 percent of loans are renewed with another loan within 14 days.

Overview of the Payday Lending Industry

In our and other studies of payday lenders two important issues arise. First, one must identify the legal status of payday lenders in the different states as well as the regulatory environment in those states in which such firms are allowed to operate. There are 13 states and the District of Columbia that actually prohibit payday lenders. These states are Arizona, Arkansas, Connecticut, Georgia, Maine, Maryland, Massachusetts, New Jersey, New York, North Carolina, Pennsylvania, Vermont, and West Virginia, as shown in Fig. 1.
Fig. 1

States that prohibit payday lending

Three states set maximum payday loan rates based on a finance charge for a 14-day $100 loan that are far below the typical rates so as to discourage this type of product being offered within their borders. The states and their rates are as follows: Montana at 36 percent, New Hampshire also at 36 percent, and Ohio at 28 percent. At the other end of the spectrum, there are six states that set no limit on the rate that may be charged on payday loans. In short, the sky is the limit. These states are Delaware, Idaho, Nevada, South Dakota, Utah and Wisconsin. The remaining 28 states allowing payday lending explicitly specify that triple-digit rates may be charged.7 Among these states, Missouri specifies the highest maximum interest rate that may be charged at 1,950 percent.8 Figure 2 shows the fairly wide distribution of the rates of interest that may be charged by payday lenders in the states in which they are allowed to operate. There are also limits on the loan amount in all but three states: Oregon, Texas and Utah. The stated maximum loan amount that is the lowest is $300 and is found in both California and Montana, while the stated maximum loan amount that is the highest is $2,500 and is found in New Mexico.
Fig. 2

Distribution of maximum allowable interest rates by payday lenders

The most frequent loan amount limit is $500 and is found in 18 states. In addition to limits being placed on loan amounts, there are specified limits on the terms on loans in all but three states: Idaho, Nevada and South Dakota. Fifteen states specify a maximum loan term but at the same time do not specify a minimum, including California, Delaware, Hawaii, Iowa, Louisiana, Michigan, Minnesota, Montana, Nebraska, North Dakota, South Carolina, Tennessee, Utah, Washington, Wisconsin and Wyoming. Wisconsin specifies the longest allowable loan term at 90 days, whereas Florida, Kansas, New Hampshire and Texas all specify the shortest allowable loan term at 7 days. The most frequent maximum loan term that is specified is 31 days. Interestingly enough, Colorado specifies a minimum loan term of 6 months.

There are also regulatory limits on the number of loans that an individual may have outstanding at one time and the number of times a loan may be rolled over. There are 12 states that either do not specify or set a limit on the number of outstanding loans, including Alaska, Louisiana, Minnesota, Mississippi, Nevada, Oregon, South Carolina, South Dakota, Texas, Utah, Wisconsin and Wyoming. Some states do not limit the number of outstanding loans but instead limit the dollar amount outstanding at any one time, such as Alabama, Delaware and Idaho. The most common limits set by states are one or two loans outstanding at any one time. As regards rollovers, 22 states prohibit any rollovers at all. The other 11 states allow between one and four rollovers, with the exception that Kansas, Nevada and Utah do not specify a limit. In a study of the borrowing patterns of over 12 million loans in 30 states, Burke et al. (2014, p.4) found that over 80 percent of payday loans are rolled over or followed by another loan within 14 days.

The second issue that arises in studies of payday lenders is determining the number of firms in the different states. Unfortunately, there is no central database for such information nor is such information readily available from the various state regulatory authorities of payday lenders. Nonetheless, estimates by Stephens Inc. (2013) indicate that there were 18,273 payday lending stores in 2012. Moreover, a few fairly large firms play a major role in the industry. Advance America is the largest such firm in the United States and was acquired by Grupo Elektra, a company owned by Ricardo Salinas Pliego of Mexico, in 2012. Advance America has roughly 2,400 stores throughout the nation. However, these are not exclusively payday lenders, with some of the stores offering pawn services, check cashing and other services. As of mid-2014, we have only been able to identify the following firms as publicly-traded entities: Cash America International (CSH), QC Holdings (QCCO), EZCORP Inc. (EXPW), First Cash Financial Services (FCFS) and DFC Global (DLLR). All of these firms engage not only in payday lending but offer other short-term financial services, such as pawn lending and check cashing. Cash America International has more than 1,000 stores; QC Holdings has about 500 outlets, while EZCORP Inc. has about 900 U.S. outlets, with roughly 500 being financial service stores. DFC Global operates in a number of countries, with about 293 outlets in the United States. First Cash Financial has 309 U.S. stores and more than that number in Mexico.

To obtain the number of payday lending stores in the United States, one usually relies on a proxy measure for such firms. In this regard, we follow the approach of Bhutta (2014), who relies on two North American Industrial Classification System (NAICS) codes to capture payday lending firms. Specifically, these two codes include firms primarily engaged in making unsecured cash loans to consumers and in facilitating credit intermediation, including check cashing services and money order issuance services.9 These firms encompass nondepository consumer lending and other activities related to credit intermediation.

The number of payday lenders based upon these codes for each state is provided in Appendix Table 3. The total for all the states is 29,044. As may be seen, numbers are included even for the 14 states that prohibit payday lending. Of the total number of firms in the table, 3,952 or 13.6 percent are located in these 14 states. Georgia has the largest number of such firms at 1,208 or 30.6 percent for the total of these states. Upon checking with the regulatory authorities and the appropriate statutes in Georgia, we have concluded that almost all of the firms for this state listed in the table are industrial loan firms. Importantly, these firms are allowed to charge an interest rate and a fixed loan fee for a small loan that, based on our interpretation, enables them to charge a maximum interest rate of 218 percent, which helps explain the large number of firms in the table for Georgia. Excluding this state, the remaining 2,744 firms in the other 13 states that prohibit payday lending offer similar services as payday lenders. In any event, our empirical analysis below will take into account all the firms as well as only those firms operating in states that allow payday lending. For convenience we will use the term “payday lenders” for the firms throughout the remainder of the paper.10

The distribution of payday lenders by state is shown in Fig. 3. As may be seen, the largest number of payday lenders are found in California, Illinois and eight southeastern states, which includes Alabama, Florida, Georgia, Louisiana, Mississippi, South Carolina, Tennessee and Texas. Each of these states has more than 1,000 payday lending stores. The state that has the most payday lenders is Texas with 4,623, while the state that has the fewest is Vermont with four. Appendix Table 3 shows that the mean (median) number of payday lenders in states allowing such firms is 678 (432), whereas the mean (median) number of payday lenders in states prohibiting such firms is 282 (173). A t-test of the difference in means of the number of payday lenders (as well as the number per 10,000 people) between states allowing payday lending and those prohibiting it indicate that the former states have significantly more payday lenders, which is what one would expect.
Fig. 3

Number of payday lenders by state (Number is based on NAICS codes that also include other nondepository consumer lending such as check cashing services)

Nationwide, there are 97,670 bank’s branches. Texas has the largest number of banks and branches at 6,875, while Alaska has the fewest at 132 (Fig. 4). On average, there are roughly three bank branches per payday store for all the states. In every state, there are more bank branches than payday stores. South Carolina has the largest ratio of payday stores to bank branches at 0.94, or nearly one store per branch. Rather than compare the number of payday lender stores to the number of banks and bank branches, Figure 4 shows the relationship of stores to branches for states when each is expressed per 10,000 people. As may be seen, Mississippi has the highest number of payday lending stores on this basis, followed by Louisiana, South Carolina, Oklahoma, Alabama, Tennessee and New Mexico. All of these states have more than two stores per 10,000 people. Vermont has the smallest number of stores per 10,000 people, followed by Maine, New Hampshire, Arkansas, West Virginia and Alaska, with each having less than 0.15 stores per 10,000 people. In the case of banks, North Dakota has the largest number of banks and branches per 10,000 people, followed by Nebraska, South Dakota, Kansas, Iowa and Arkansas. All these states have more than five banks and bank branches per 10,000 people. Alaska has the smallest number of banks and branches per 10,000 people with California a close second, with both states being the only states having fewer than two banks and branches per 10,000 people.
Fig. 4

Number of payday lenders and bank branches per 10,000 people by state

Empirical Model and Results

To address the issue of the concentration of payday lending stores in states throughout the United States, we specify the following model with states being units of observation:
$$ \begin{array}{c}\hfill y=\alpha +{\beta}_1\left( bank\; branches\right)+{\beta}_2\left(\; financial\; factors\right)+{\beta}_3\left( demographic\; factors\right)+\hfill \\ {}+{\beta}_4\left( educational\; factors\right)+\varepsilon, \end{array} $$
(1)
where y is the number of payday lending stores, bank branches are the number of banks and bank branches, financial factors include income per capita, the poverty rate and the unemployment rate, demographic factors include the percentages of the population that are African American, Asian, Hispanic, aged 15 and under and aged 65 and over, educational factors include the percentages of the population that have a high school degree or higher and have a bachelor degrees or higher, and \( \varepsilon \) is a random error term.11
Before presenting and discussing the empirical results based upon the estimation of Eq. (1), the simple correlations among the variables used in our analysis are shown in Table 1. In this table, given the substantial variation in population among the different states, the focus is on the number of payday lending stores per capita. As may be seen, there is no significant correlation between the number of bank branches and the number of payday lending stores, suggesting they are neither friend nor foe. Perhaps not surprisingly, the number of payday lending stores is positively and significantly correlated with the percentages of the population that are African American and aged 15 and under (indicating a larger family size). It is also found that the correlations between the number of payday lending stores and the percentages of the population that have high school and bachelor degrees are significantly negative, which also does not seem surprising. Turning to the financial factors, there is a significantly negative correlation between the number of payday lending stores and income per capita, but a significant and positive correlation between the number of stores and the poverty rate. Again, neither of these correlations is surprising. At the same time, there is no significant correlation between the number of payday lending stores and the unemployment rate, which does seem surprising.
Table 1

Correlations between the number of banks and payday stores and selected demographic and financial characteristics at state level

State

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

Bank’s branches per capita (1)

1.00

            

Payday loan stores per capita (2)

0.002

1.00

           

% White (3)

0.37**

−0.24

1.00

          

% Black or African American (4)

−0.07

0.40**

−0.74**

1.00

         

% Asian (5)

−0.36*

−0.21

−0.52**

−0.10

1.00

        

% Hispanic (6)

−0.48**

0.08

−0.22

−0.11

0.15

1.00

       

15 years and under (7)

−0.15

0.40**

0.10

−0.18

−0.07

0.28*

1.00

      

% aged 65+ (8)

0.36*

−0.15

0.29*

−0.16

−0.04

−0.25

−0.60**

1.00

     

Income per capita (9)

−0.09

−0.51**

−0.33*

0.21

0.22

0.07

−0.48**

−0.19

1.00

    

Poverty rate (10)

−0.03

0.61**

−0.26

0.44**

−0.26

0.15

0.11

0.05

−0.57**

1.00

   

% High school degree (11)

0.27

−0.55**

0.43**

−0.50**

0.09

−0.38**

−0.12

0.04

0.36**

−0.77**

1.00

  

% Bachelor’s degree or higher (12)

−0.09

−0.46**

−0.28*

0.18

0.18

0.12

−0.39**

−0.22

.90**

−0.47**

0.40**

1.00

 

Unemployment rate (13)

−0.58**

0.20

−0.42**

0.49**

−0.02

0.29*

−0.06

−0.07

−0.06

0.50**

−0.57**

−0.06

1.00

*Denotes significance at 5 % level, **denotes significance at 10 % level.

In addition to the correlations just discussed, Table 1 provides the corresponding correlations for the number of bank branches and the same variables. Briefly, there are significantly positive correlations between the number of branches and both the percentage of the population that is white and aged 65 and over, which do not seem surprising. The correlations for both Asian and Hispanic, moreover, are both significantly negative. The only other significant correlation is the one for the number of branches and the unemployment rate, and it is negative. These correlations would seem to indicate that more bank branches are found in states with lower unemployment rates.12

Turning from the bivariate to the multivariate empirical results, the dependent variable employed is the number of payday loan stores per 10,000 people (Table 2). The ordinary least squares results indicate that the only significant explanatory variable is the percentage of the population that is African American, and its coefficient is positive. This indicates that there is on average a greater concentration of payday lender stores in those states with a higher concentration of African Americans.
Table 2

OLS and ridge regressions: Number of payday lenders per 10,000 people on selected demographic and financial characteristics at the state level

 

OLS

VIF

OLS

VIF

Ridge Regression

VIF

Constant

20.314

 

−0.782

 

13.670

 

(0.481)*

 

(0.685)

 

(0.042)

 

Bank’s branches per 10,000 people

−0.053

3.27

−0.049

2.596

−0.007

1.37

(0.707)

(0.693)

(0.937)

% Black or African American

0.043

4.22

0.036

2.241

0.029

1.34

(0.006)

(0.001)

(0.001)

% Asian

−0.005

1.35

−0.005

1.331

−0.004

0.97

(0.669)

(0.684)

(0.704)

% Hispanic or Latino origin

0.013

4.02

0.008

1.880

0.005

1.14

(0.436)

(0.471)

(0.552)

% Age under 15

0.092

8.59

0.122

1.531

0.100

1.30

(0.467)

(0.023)

(0.050)

% aged 65+

0.009

5.27

  

−0.005

1.26

(0.934)

  

(0.931)

Ln(Income per capita)

−1.972

22.34

  

−1.196

1.23

(0.428)

  

(0.046)

Poverty rate

0.031

10.63

0.085

2.925

0.046

1.47

(0.716)

(0.055)

(0.154)

% High school degree or higher

−0.007

5.17

  

−0.027

1.65

(0.901)

  

(0.403)

% Bachelor’s degree or higher

−0.015

7.16

−0.044

2.449

−0.016

1.59

(0.681)

(0.045)

(0.357)

Unemployment rate

−0.114

3.17

−0.113

3.014

−0.071

1.46

(0.133)

(0.117)

(0.170)

Adjusted R2

0.518

 

0.539

   

k

0

   

0.099

 

Root MSE

0.559

   

0.572

 

Numbers in parentheses are p values

However, the correlations discussed above and the variation inflation factors (VIFs) provided in the table indicate a high degree of multicollinearity among some of the variables, which can lead to the insignificance of variables. It was therefore decided to omit three of the collinear variables, with the results reported in column four of the table. In this case, not only is the African American variable significant, but also three other variables. The poverty rate enters with a significantly positive sign, which one might expect. Also, as one might expect, the percentage of the population that is 15 and under enters with a significantly positive sign, while the percentage of the population that has a Bachelor’s degree or higher enters with a significantly negative sign. As a final check a ridge estimation technique is employed to address the multicollinearity issue.13 The ridge estimation was used in an early work by Manage (1983). The results of this estimation are reported in column six of the table. In this case the difference as compared to dropping variables is that per capita income now enters with a significantly negative sign and the poverty rate and educational variables lose their significance.

Overall, the empirical results indicate the following: (1) in all three of the estimated equations, there is a significantly positive relationship between the number of payday lending stores and the percentage of the population that is African American; (2) in two of the estimated regressions the percentage of the population that is aged 15 and under (larger family size) also enters with a positive and significant coefficient; (3) income per capita is significant and negative with the ridge estimation; (4) the percentage of the population that has a Bachelor’s degree or higher is significantly negative, but only in the case when some collinear variables are dropped, and (5) the poverty rate is significantly positive.

Plans for Future Research

It is clear that the payday lending industry receives mixed reviews. Some believe that payday lenders prey on lower-income and less financially literate individuals, frequently African Americans and Hispanics, charging exorbitant interest rates for extremely short-term loans. Still others believe that these lenders cater to individuals who benefit by gaining access to otherwise unavailable short-term credit for unexpected needs, such as a medical emergency. Unfortunately, despite several fairly recent and careful empirical studies of payday lending, there has been no consensus reached as to whether there has been a net gain in welfare to borrowers. Given the importance of these particular financial firms to a significant segment of the population, there is always the need for more research to better understand their role in the financial system. This is especially important in view of fact that federal policymakers appear to be contemplating a tightening of regulations over the payday lending industry. Our results indicate this may indeed be appropriate to the extent that payday lenders are locating in areas simply to provide more and quite costly funds to African Americans as well as poorer and less educated individuals without any corresponding welfare benefits.

In terms of future research, our data and that of most other researchers is limited by no readily available database on just payday lenders and the various state regulations to which they are subjected. We are therefore in the process of collecting more detailed information on the state regulations governing payday lenders and their operations at the county and zip-code level. This effort is expected to enable us and others to more carefully and thoroughly assess the role of payday lenders in the consumer credit marketplace licensed payday lenders.

Footnotes

  1. 1.

    Payday lenders are also referred to as deferred deposit originators and their product as payday advances, cash advances, deferred deposits, among other terms.

  2. 2.

    The interest rates in both cases are calculated assuming the two loans are outstanding for a year and the fees are paid every 14 days. Of course, the rates are much higher if one assumes a new loan is taken out every 14 days and the same fees charged.

  3. 3.

    See Douglas (2014, p.2).

  4. 4.

    Due to limited availability of data, the paper focuses on actual storefronts to the exclusion of online payday lenders. However, William H. Sorrell (2014, p.1), Attorney General of Vermont, recently stated that “Online lenders nationwide (currently numbered at over 200) earned over $18 billion dollars in income from high-interest, small-dollar loans made in 2012.” Yet, according to the Consumer Financial Protection Bureau (2013), these payday loans still make up a minority of the total loan volume, and the loans are offered with fees equal to or higher than storefront loans.

  5. 5.

    It should be note that in the late 1990s some payday lenders began partnering with nationally chartered banks and payday loans became “bank loans” because such banks were not subject to state-imposed fee caps or usury laws. However, the Federal Deposit Insurance Corporation took actions in 2003 and 2005 that, according to Stegman (2007, p. 179) “… rendered the rent-a-bank model obsolete.”

  6. 6.

    Changes in credit supply are proxied by two dummy variables, with 0 before a state banned payday lending and also a 0 before a state passed enabling legislation for payday lending, and a 1 in both cases after the banning and enabling changes. They rely on annual store counts obtained from Stephen Inc., which is an investment bank that tracks the payday lending industry.

  7. 7.

    Colorado has a tiered structure of rates and fees based upon the loan amount.

  8. 8.

    As a result of the Talent-Nelson Amendment to the John Warner National Defense Authorization Act of 2007, a 36 percent annual percentage rate cap took effect on October 1, 2007, for all payday loans made to military borrowers on active duty.

  9. 9.

    The codes are 522291 (consumer lending) and 522390 (other activities related to credit intermediation).

  10. 10.

    It should be noted that when we refer to the number of payday lenders, we are referring to the number of stores since each store must have a separate license.

  11. 11.

    Our study is related to that of Prager (2009) and several of the papers he discusses, but relies on more recent data, a somewhat different set of variables to explain the concentration of payday lending stores, and a different estimation technique to deal with multicollinearity.

  12. 12.

    Rank order correlations were also calculated for the same variables as in Table 1. The results are quite similar to those already reported, with one notable exception. The correlations between the percentage of the population that is Asian and the income and education variables are now significantly positive, and significantly negative for the poverty rate and the percentage of the population that is aged 65 and over. These correlations are not unexpected.

  13. 13.

    A check on the stability of the estimated coefficients in the ridge regression was conducted and the results indicate that the coefficients are quite stable.

Notes

Acknowledgment

The authors are extremely grateful to Richard Cebula for inviting us to write and present this paper as well as helpful comments. Thanks are also due to Kang Lee for assistance with the ridge estimation application.

References

  1. Bhutta, N. (2014). ”Payday loans and consumer financial health”. Journal of Banking and Finance, 47, 230–242.CrossRefGoogle Scholar
  2. Burke, K. Lanning, J. Leary, J. and Wang J. (2014): ”CFPB Data point: Payday lending.” Consumer Financial Protection Bureau.Google Scholar
  3. Carrell, S., & Zinman, J. (2014). In Harm's Way? payday loan access and military personnel performance. Review of Financial Studies, 27, 2805–2840.CrossRefGoogle Scholar
  4. Consumer Financial Protection Bureau. (2013) “Payday loans and deposit advance products.”Google Scholar
  5. Douglas, D. (2014, March 26) “There are almost as many payday lenders as McDonald’s and Starbucks. No, really.” Washington Post, p. 2.Google Scholar
  6. Gallmeyer, A., & Roberts, W. T. (2009). Payday lenders and economically distressed communities: a spatial analysis of financial predation. The Social Science Journal, 46, 521–538.CrossRefGoogle Scholar
  7. Manage, N. (1983). Further evidence on estimating regulated personal loan market relationships. Quarterly Review of Economics and Business, 23, 63–80.Google Scholar
  8. Melzer, B. T. (2011). The real costs of credit access: evidence from the payday lending market. Quarterly Journal of Economics, 126, 517–55.CrossRefGoogle Scholar
  9. Morgan D. P. and Strain M. R. . (2008) “Payday holiday: How households Fare after Payday Credit Bans.” Federal Reserve Bank of New York Staff Report No. 309.Google Scholar
  10. Morgan, D. P., Strain, M. R., & Ihab, S. (2012). “How payday credit access *affects overdrafts and other outcomes”. Journal of Money, Credit, and Banking, 44(2-3), 519–531.CrossRefGoogle Scholar
  11. Morse, A. (2011). Payday lenders: heroes or villains?”. Journal of Financial Economics, 102, 28–44.CrossRefGoogle Scholar
  12. Prager, R. A. (2009) “Determinants of the Locations of Payday Lenders, Pawnshops and Check Cashing Outlets.” Federal Reserve Board Finance and Economics Discussion Series # 2009-33.Google Scholar
  13. Sorrell, W. H., Attorney General-State of Vermont. (2014) Letter to DISH Network regarding advertisements for payday lenders in Vermont.Google Scholar
  14. Stegman, M. A. (2007). Payday lending. The Journal of Economic Perspectives, 21, 169–190.CrossRefGoogle Scholar
  15. Stegman, M. A., & Faris, R. (2003). Payday lending: a business model that encourages chronic borrowing. Economic Development Quarterly, 17, 8–32.CrossRefGoogle Scholar
  16. Zinman, J. (2010). Restricting consumer access: household survey evidence on effects around the Oregon rate cap. Journal of Banking and Finance, 34, 546–556.CrossRefGoogle Scholar

Copyright information

© International Atlantic Economic Society 2015

Authors and Affiliations

  • James R. Barth
    • 1
    • 2
  • Jitka Hilliard
    • 1
  • John S. JaheraJr.
    • 1
  1. 1.Auburn UniversityAuburnUSA
  2. 2.Milken InstituteSanta MonicaUSA

Personalised recommendations