Differences in Consumer Credit Choices
By Banked and Unbanked Mississippians
DO NOT CITE WITHOUT WRITTEN CONSENT OF THE AUTHOR
November 6, 2014
Thomas W. Miller, Jr. Professor of Finance and Jack R. Lee Chair in
Financial Institutions and Consumer Finance
Department of Finance and Economics
P.O. Box 9580
Mississippi State University
Mississippi State, MS 39762
E: [email protected]
Ph: (662) 325-1997
Special Acknowledgement
I first heard about “Consumer Credit and the American Economy” when Mr. Franc Lee, CEO of First
Tower, Inc., took me to visit some folks at the American Financial Services Association headquarters in
Washington, D.C. Two of the people present were Dr. Tom Durkin and Dr. Greg Elliehausen. During the
conversation, I learned that they were working on a manuscript on consumer credit. How convenient, I
thought. I was in the process of putting together a course in consumer credit at Mississippi State
University, so such a manuscript would be very useful for my students and me. Little did I know the true
importance of this book for new researchers in the area of consumer finance.
Through the kindness of the authors, I was able to use manuscript in class for two years before it
appeared in print. When Oxford University Press asked me to review the book, I told them that I would
not be an unbiased reviewer. I told them not only did I think they should publish the manuscript, but
that they should put their best editors on the project. (Despite my disclosure, Oxford University Press
insisted that I write a review for them.)
This paper builds on material found in Chapter 8 of the book. Each time I read the chapter, I am
astounded at the depth and breadth of knowledge displayed by the authors. I thank them greatly for
their accumulated knowledge and willingness to share. It is my intention to expand and build upon this
book by conducting additional original and rigorous research in the area of consumer credit—especially
in the small dollar loan space. My aim is to help anyone who reads my work to understand the costs,
benefits, and workings of the small dollar loan market.
Page | 1
Differences in Consumer Credit Choices
By Banked and Unbanked Mississippians 1
Abstract. This paper presents the results of a 2011 survey of a random
sample of 400 Mississippians. The survey questions centered on the use
of non-bank supplied consumer credit products.
Two previously unanswered questions were asked: 1) Do consumers
know where to go to get a loan that suits their needs? and 2) Do
consumers understand the terms of their loans? Concerning question 1,
whether the Respondent has a bank account or not matters and so does
whether their education level stops at high school. Concerning question
2, what matters is whether the educational level of the Respondent
stops at high school or not. Whether they have a bank account or not
does not matter, statistically speaking.
By usage, evidence from the survey suggests that consumers avoid
credit products that might trap them in the so-called “debt-trap”. The
survey respondents used payday loans least often of all surveyed
products. About three times as many Respondents used title loans and
pawn loans—both of which are secured by collateral and are nonrecourse. Many more Respondents used installment loans than any
non-installment loan product.
1
I thank Todd Zywicki and Tom Durkin for encouraging me to pursue research in consumer finance. I thank
Mercatus Center staff for their creativity and feedback, especially Rob Raffety and Hester Peirce. I thank Carolyn
Moore Miller for research and editing assistance with this manuscript. I also thank Frank Adams for conferring with
me on various aspects of reporting the survey results.
Page | 2
Differences in Consumer Credit Choices
By Banked and Unbanked Mississippians
“If the test of a subject’s historical importance is the amount of
controversy it generated, then consumer credit is one of the most
significant subjects in the history of the American twentieth century.”
---Lendol Calder (1999)
I. Introduction and Overview
The topic of consumer credit, especially small dollar loans, continues to generate passionate debate.
Sometimes, rigorous research results that buttress the forceful opinions expressed might be missing.
Over time, the regulations and laws concerning small dollar loans that have emerged from the debate
form a remarkable patchwork of inconsistencies. For example, some states prohibit access to some
forms of consumer credit, while neighboring states prohibit access to other forms of consumer credit.
To arrive at a well-formed policy concerning issues in consumer credit, logic dictates that all participants
in the debate use the same set of facts. It is incumbent, therefore, on researchers to prepare rigorous,
data-driven results to present to lawmakers, policy advocates, and other interested parties.2
The continued existence of consumer credit products, despite the ever more restrictive web of
regulation, confirms there is a demand for consumer credit. In some instances, the supply of certain
types of consumer credit has been legislated out of existence.3 Even if lawmakers banned all types of
consumer credit, consumers would simply seek out “illegal” sources of credit.4
2
In the spirit of full disclosure, the analysis that follows necessarily carries two biases through which the author
views personal finance. First, the author asserts that the opportunity to seek consumer credit is a fundamental
liberty. Second, the author asserts that nearly all of-age citizens are in the best position to make their own
decisions concerning their finances.
3
Consider the availability of some consumer credit products in the state of Arkansas. On March 18, 2008, Arkansas
Attorney General Dustin McDaniel sent letters to 156 payday lenders, ordering them to stop issuing new loans and
void any current and past due loans or face legal action. The last payday lending business in Arkansas closed on
August 11, 2009. In addition, the Arkansas Constitution imposes a 17% interest rate cap on any loan—thereby
rendering small dollar loans as uneconomic to make by installment lenders.
4
Illegal lenders charge more for their loans because they bear the additional risk of prosecution and conviction for
their illegal activity. Loan sharking is an important source of income for organized crime. According to Sifakis
(2005), the going interest rate on loan shark loans has traditionally been twenty percent per week, or an annual
percentage rate (APR) of 1,040%.
Page | 3
Few areas spawn so many emotionally charged opinions and controversy as does the confluence of nonbank supplied credit products and the financial condition of low-income citizens. Unfortunately, for
many important consumer credit policy questions, there is a lack of rigorous empirical studies in this
particular area.
To add information to the debate, the focus of this paper is on consumer use of non-bank consumer
credit products. Specifically, the aim of this research is to document what consumers themselves have
to say about which credit products they use. The main research motivation is to seek some evidence on
two previously unexplored questions: 1) Do consumers know how to obtain loans that suit their needs,
and 2) Do consumers understand the terms of loans made to them?
For many people, an important source of financing is personal savings. Consequently, consumers were
asked to disclose their intended, and actual, savings habits. Other important information gathered by
the survey was whether the consumer had a bank account, as well as the income and education level of
the consumer. As a result, one can study whether the answers to the main questions asked in the study
relate on whether the Respondent has a bank account, as well as their income and education level.
The paper proceeds as follow. Section II is a narrative of the demand and supply conditions which
existed for non-bank supplied consumer credit in the state of Mississippi at the end of 2011—the time
of the Survey. Section III contains a description of the survey method used in the study. Section IV
contains the empirical results. Section V is a summary.
II. Demand and Supply Conditions for Non-Bank Supplied Consumer Credit in
Mississippi.
II.A. A Summary of the Economic Conditions in Mississippi
The economic conditions of the State of Mississippi make the state a natural laboratory to research
many policy questions concerning un-banked consumers, non-bank supplied credit products, and the
financial condition of lower-income citizens.
According to various U.S. census documents5, Mississippi ranks significantly below the median in the
United States in terms of many wealth, income, and education measures. For example, in 2009:
5
www.census.gov; quickfacts.census.gov.
Page | 4

Housing and Home Ownership. The median home value in Mississippi was about half
that of the US median home value. For Mississippi, the median value was $100,200,
while for the US, the median value was $181,400. Mississippi, however, ranked 5th in the
US in terms of household home ownership percentage. For the US, 67.4 percent of
households were homeowners. For Mississippi, this percentage was 75.5.

Income. Per capita and median household income levels in Mississippi were about
three-fourths the national levels. For Mississippi, per capita income6 was $30,399 (rank
50th) while for the US per capita income was $40,208. The median household income in
Mississippi was $36,646 (rank 50th) while the median household income was $50,221 for
the US. The percent of households in Mississippi that had an annual income under
$25,000 was 36 percent (rank 50th). By comparison, 24.7 percent of households in the
US had annual income less than $25,000.

Education. Mississippians hold Bachelor’s degrees or higher at about two-thirds the
average rate in the USA. In Mississippi, 19.6 percent of the population over the age of 25
held a four-year college degree or higher. For the US, 27.9 percent of the population
over the age of 25 held a four-year college degree or higher.
II.B. Sources of Non-bank Consumer Credit in Mississippi
All five of the major non-bank consumer credit products exist in the State of Mississippi: 1) Lease-toOwn Transactions, 2) Pawn Loans, 3) Vehicle Title Loans, 4) Payday Loans, and 5) Traditional Installment
Loans from finance companies.7
Table 1 summarizes some aspects of these loan products in the State of Mississippi.8 For example, in
2011, there were 1,053 licensed check cashers (a category that includes payday lenders). Given the
estimated 2011 state population, there was a licensed check casher or payday lender for every 2,100
Mississippians over the age of 18.9 There were 525 licensed small loan companies, whose main lending
6
Per capita personal income is total personal income divided by total midyear population.
A description of these loan products and their terms for Mississippians appears in the Appendix.
8
To see the full report of the Deputy Commissioner of the Mississippi Department of Banking and Consumer Credit
report, navigate to http://www.dbcf.state.ms.us/documents/annual-report11.pdf.
9
In 2011, the Census estimates the total population of Mississippi to be 2,967,299, with 24.7% under the age of 18.
7
Page | 5
product is a personal installment loan. The outstanding loan balance for the Small Loan category
suggests an average loan balance of about $2,450.10 Additionally, there were 219 licensed pawnbrokers,
and 402 licensed title pledge lenders.
It is interesting to note is that the number of transactions in the Check Casher / Payday Loan category is
the smallest of the four categories of loan products. However, interest in Payday loans by federal
regulators is quite high—as is the press coverage of this loan space.
Dividing the dollar amount of the loans outstanding by the number of licenses outstanding yields a
snapshot estimate of the average loan capital required per each licensee. For Check Casher / Payday
Loans, this amount is $43,409—the smallest capitalization of the four types of consumer loan products.
Pawnbroker and Title Pledge Lender capital amounts are $46,891 and $76,730, respectively. In stark
contrast to these levels, the Small Loan category has a loan capital estimate of $1,530,045per licensee.
This amount is about 20 times the level of Title Pledge and about 35 times the level of the average Check
Casher / Payday Lender. These contrasting levels suggest that the fundamental business model of
making small dollar Installment Loans differs significantly from the business models of the other three
categories of consumer loan choices.
III. Survey Method
In December 2011, The Survey Research Laboratory at Mississippi State University conducted a
statewide telephone survey at the request of Economy Watch, a now defunct publication of the College
of Business at Mississippi State University. In addition to the questions asked for the Economy Watch
article, the survey included a set of questions concerning the use of consumer credit products.11 Using
random digit dialing, The Survey Research Laboratory formed a random sample consisting of 75%
landline phone numbers and 25% cell phone numbers. Interviewers called 9,500 numbers to get 400
people to answer the 32-question survey (hereinafter “Respondents”).
10
This loan category represents a mature industry that contains loans that exceed twelve months in duration. One
can assume, therefore, an approximate steady state as consumers pay off old loans and take out new ones. In this
case, the dollar amount outstanding will be roughly stable, as will be the number of transactions.
11
No results presented in the current study appear in Economy Watch. The answers to the survey questions that
appear in Economy Watch were about questions like: the general economic conditions in the State of Mississippi,
whether the Respondents felt their families were better off now than one year ago, and whether it is a good time
to purchase consumer durables.
Page | 6
Consistent with survey research practice, the responses were weighted by gender, race, and age so that
the responses mirror prior knowledge about gender, race, and age distribution in the state. Table 2
contains the actual survey responses as well as the weighted observations.12 As shown in Table 2, the
weighted observations in Panels A, C, and D are consistent with the data presented from the 2010
Census. As a result, the sampling error (for dichotomous response options with a 50/50 split) is no
larger than ±5% at a 95% confidence level.
IV. Survey Results and Statistical Tests.
IV.A. Savings Habits and Plans
One way for consumers to finance unexpected expenses is to have a “rainy day fund.” Consumers
undoubtedly know that to the extent that they have saved some money, they can use this money to pay
for unexpected financial shocks. Despite the wish by most observes that all households have emergency
savings, many consumers do not have such savings.13
Lusardi, Schneider, and Tufano (2011) examine the ability of American households to come up with
$2,000 within 30 days to help weather a financial shock. They document that approximate one-half of
American household certainly could not, or probably could not, do so. Consequently, when faced with a
financial shock that requires a payment, these household likely need access to some form of credit.
One goal of this part of the paper is to document savings habits and plans by Respondents in the survey.
A further goal is to document whether these habits and plans vary by whether anyone in the
Respondent’s household (including the Respondent) has a checking or savings account at a bank.
Table 3 and Figure 1a present survey results concerning savings habits and planned saving. About half of
all Respondents reported that they were able to save money last year, while eighty percent say that
they plan to save money in the coming year. This finding almost certainly represents a “triumph of hope
12
Each survey response receives a weight based on gender, race, and age. The weighting scheme works like this.
Suppose the Respondent is a black female in the 18-24 age group. To calculate the weight for this Respondent, one
divides the percent of the population in Mississippi that black females age 18-24 represent by the corresponding
percent in the sample. The weight will be greater than one if too few Respondents in the sample are from this
category, and will be less than one if there are too many. Similar weights are calculated for black males, white
males, and white females for this age group as well as three others: 25-44, 45-64, and 65+.
13
For example, Grinstein-Weiss, Russell, Tucker, and Comer (2014) recommend that policy makers should explore
opportunities to facilitate and fund “innovative collaborations between academic, research, business, and
government partners [to] generate new approaches to long-standing problems (e.g., lack of emergency savings)
and provide effective interventions that can significantly improve the financial well-being of American
households.”
Page | 7
over experience.” That is, there is to believe, ex ante, that 2011 was significantly different from 2010. If
so, it is likely that in 2010 half the Respondents would have reported that they saved in 2010, and eighty
percent likely planned to save in 2011. Only half of them, however, did save in 2011.
Conditional on saving in 2011, about 95% of the Respondents stated that they intended to save in 2012.
This percentage is most likely a reasonable estimate because this group of consumers had displayed
savings behavior. Note, however, that 65% of the Respondents who did not save in 2011 stated that
they intend to save in 2012.14
The last two panels in Table 3 and Figures 1b and 1c separate the Respondents into those who affirmed
that someone in their household (including themselves) have a checking or savings account at a bank
(hereinafter “Banked Respondents”) from those who do not. In the Survey, 15.8% of the Respondents
said that no one in their family had a banking account (hereinafter “Unbanked Respondents”). This level
is consistent with a recent study by the FDIC that reports 15.1% of citizens in Mississippi are unbanked.15
About 55% of the Banked Respondents reported that they saved money in the previous year, compared
to about 33% of the Unbanked Respondents. These two proportions are statistically significantly
different, with a p-value of 0.0021. That is, percentage of Banked Respondents who reported that they
saved, was a statistically higher percentage of Unbanked Respondents who reported that they saved.
Nearly 82% of Banked Respondents reported that they planned to save the next year. About 70% of
Unbanked Respondents also reported that they planned to save the next year. (The survey did not
attempt to discover which financial service, if any, the Unbanked group will use in their planned saving.)
These two percentages are not statistically significantly different. That is, a Respondent’s answer to the
question of whether they planned to save next year is likely to be the same whether the Respondent
was Banked or Unbanked.
IV.B. Credit Products used by Respondents
14
Note that the weighting scheme employed introduces rounding error. One can see an example of this rounding
error in Table 3. The number of people planning to save is 318, but in the panel directly below, 191 + 126 = 317.
15
2011 FDIC National Survey of Unbanked and Underbanked Households, pg. 126. Susan Burnhouse and Yazmin
Osaki, lead authors.
Page | 8
IV.B.1 Results by Bank Account. Table 4 and Figure 2 present Survey results concerning the use of loan
products by Respondents. Respondents could state that they use one or more of these products. The
results are presented for two groups: the Banked or Unbanked.
Panel A of Table 4 contains results for financing sources whose main product can be classified as a noninstallment loan. These three are 1) Pawn Shop, 2) Payday Lender, and 3) Title Lender. Panel B of Table 4
contains results for loan sources whose main loan product can be classified as an installment loan. These
three installment loan sources are 1) Finance Company, 2) Credit Union, and 3) Bank.
For each non-installment product listed in Panel A, the “Percent Yes on Product” from Unbanked
Respondents exceeds the “Percent Yes on Product” from Banked Respondents. Further, a Chi-Square
test for equal percentages shows that these percentages are not equal.
By way of contrast, for each installment product listed in Panel B, the “Percent Yes on Product” from
Banked Respondents exceeds the “Percent Yes on Product” from Unbanked Respondents. Further, a ChiSquare test for equal percentages shows that these percentages are not equal at the 0.05 level for
Credit Unions and Banks. For Finance Companies, the Chi-Square test for equal percentages shows that
the null hypothesis of equal percentages is rejected at the 0.10 level (the p-value is actually 0.0700).
Based on the statistical results shown in Table 4, Unbanked Respondents used non-installment loan
products at a higher rate than did Banked Respondents, while Unbanked Respondents used installment
loan products at a lower rate than did Banked Respondents. Figure 2 displays the contrasting responses
for “Percent Yes on Product” by loan source.
IV.B.2. Evidence Concerning Debt-Trap Avoidance. A current concern of many regulators, lawmakers, and
consumer advocates is the so-called “debt-trap.” One common definition of a consumer debt-trap, or
cycle of debt, when the consumer takes out a short-term lump sum loan and is unable to pay back the
interest and principal when they are due. Often, the consumer pays the interest expense, and “rolls
over” the principal. If this cycle continues, the borrower can wind up paying a considerable amount of
additional interest.
As shown in Table 4, however, Respondents used Payday Loans the least of any product. Although
Payday Loans are perhaps the easiest type of loan to obtain, only ten Respondents of 400 reported that
they had used the service of a Payday lender in the past two years.
Page | 9
Moreover, data in Table 4 shows that consumers use products that will not trap them in the so-called
“debt-trap” more often than they use products that have the potential to do so. Compared to Payday
Loan usage, about three times as many Respondents used Title Loans or Pawn Loans (both of which are
secured by collateral and are non-recourse). That is, the terms of these agreements do not obligate the
consumer to pay any money to the Title or Pawn Lender. The consumer can simply leave their collateral
with the lender and leave with the proceeds. Consumers also chose to finance the acquisition of goods
by using a Rent-to-Own agreement more frequently than they chose to use Payday loans. In a Rent-toOwn agreement, the consumer can return the item and can stop paying without further obligation..16
Another way to avoid a cycle of debt is to borrow money using an installment loan. In an installment
loan transaction, the borrower and lender review the financial condition of the borrower before the
lender extends the loan. After agreeing on the terms, the consumer has a known, and fixed, monthly
payment that is comprised of interest and principal. As shown in Panel B of Table 4, many more
Respondents used the services of Finance Companies, whose product is an installment loan, than any
non-installment lender. Respondents used the services of Credit Unions and Banks even more often
than they used Finance Companies.
Overall, as noted above, the Respondents used Payday loans least often. As detailed in the Appendix, for
the products listed in Table 4, Payday loans have the highest allowable interest rate in Mississippi. Pawn
Shops and Title Lenders have a lower allowable interest rate, and consumers used them more often
than Payday loans. Finance Companies have an even lower allowable rate, and consumers used them
more often than Pawn Shop and Title Loans. In addition, consumers, by their choices in evidence, seem
to try to avoid using a product that could entangle them in a debt trap. Instead, consumers most often
chose credit products that were either non-recourse or gave them a clear pathway out of debt.
IV.B.3. Results by Income. Figure 3 displays the percent of Respondents answering “Yes” to whether they
used a particular financial product within the past two years. The results are shown by income level. One
thing to note, however, is that 91 of the 400 (22.8 percent) Respondents refused to disclose their
income. Another thing to note is that in Figure 3, a “flat spot” indicates a response of zero. There are
several interesting aspects to Figure 3. First, note that all income levels reported using the services of a
16
The Appendix contains more detail about each of the products that appear in Panel A of Table 4, along with a
description of the workings of a small-dollar installment loan.
Page | 10
bank. And all but one, those with income exceeding $150,000, reported using the services of a finance
company.
Recall that the median household income in Mississippi in 2011 was about $35,000. Figure 3 shows
graphically that the loan product usage of those Respondents who reported their income as under
$35,000 differed from that of Respondents who reported their income as $35,000 and higher.
Respondents with less than $35,000 income were the only ones using Payday lenders, and were also the
majority users of Rent-to-Own, Title lenders, and Pawnshops.
IV.C. Do Respondents Know Where to get Loans that Suit Their Needs, and Do They Understand the
Terms of Loans They Have Taken Out?
In this section of the paper, I present and discuss the responses to two important survey questions. First,
did Respondents know where to get loans that suit their needs? The wording of the question on the
survey was: “Would you say you know how to obtain a loan that suits your needs?” Second, did
Respondents understand the terms of loans they have taken out? The wording of the question on the
survey was: “Would you say you understand the terms of loans you have taken out?”
The first blocks in Table 5 show the Likert scale responses for each of the questions highlighted in bold
italics. It is common when reporting differences in Likert scale responses to add together the number of
responses from the highest possible category and the number of responses in the second highest
possible category. Similarly, it is common to add the number in the lowest possible response category to
the number of responses in the second lowest possible category. In the subsequent discussion, I refer to
these newly created categories as “Agree” and “Disagree.”
When looking at all responses, overwhelmingly, Respondents answered that they know where to obtain
a loan that suits their needs: more than 65% agreed, while 17.8% disagreed. Roughly 3.7 times as many
Respondents agreed than disagreed. When asked whether they understood the terms of the loans that
had taken out, more than 70% of the Respondents agreed, while 12.8% disagreed. Roughly 5.5 times as
many Respondents agreed than disagreed when asked whether they understand their loan terms.
IV.C.1. Responses by Bank Account
Page | 11
When we separate the Respondents by whether they have a bank account, the results differ
substantially. The results for Banked Respondents largely agree with the overall results; 69.8% of Banked
Respondents agreed with the statement that they knew where to get a loan, while 18.1% disagreed.
That is, Banked Respondents were statistically significantly more likely to agree than disagree with the
statement that they knew where to get a loan that suits their needs.
For the Unbanked Respondents, 48.4% of them agreed that they knew where to get a loan, while 37.1%
disagreed. These percentages for Unbanked Respondents are statistically insignificant. That is, people
without a bank account were statistically equally likely to agree or disagree with the statement that they
knew where to get a loan that suits their needs.
For the question of whether they understand the terms of the loan, however, the Banked and Unbanked
had similar responses. For Banked Respondents, 74.8% agreed with the statement that they understood
the terms, while 8.7% disagreed. For Unbanked Respondents, 62.9% agreed that they knew where to get
a loan, while 27.4% disagreed. In both cases, these percentages are statistically significantly different.
That is, both Banked and Unbanked Respondents were statistically significantly more likely to agree than
disagree with the statement that they understood the terms of the loans that they had taken out.
IV.C.2. Responses by Income
When we separate the Respondents by whether they report income at a higher or lower level than the
median for the state of Mississippi, the results are similar to the bank account results. There is a
confounding factor, however, when dealing with these income results. About 25% of the Respondents
refused to answer the question about their income level.
For Respondents who reported income over $35,000 per year, 81.8% of them agreed with the
statement that they knew where to get a loan, while 5.3% disagreed. For the Respondents with a
reported income less than $35,000 per year, 60.9% of them agreed that they knew where to get a loan,
while 25.0% disagreed. These percentages are statistically significantly different.
For the question of whether they understand the terms of the loan, however, the two groups also had
similar responses. For Respondents who reported income over $35,000 per year, 86.4% agreed with the
statement that they understood the terms while 5.3% disagreed. For the Respondents who reported
income of less than $35,000 per year, 69.9% of them agreed that they knew where to get a loan, while
17.3% disagreed. In both cases, these percentages are statistically significantly different.
Page | 12
IV.C.3. Responses by Education Level
For purposes of the survey, there are three categories of education. Category One includes people
whose maximum formal education is a high school diploma. In the survey, 171 Respondents fit into this
category. Category Two includes Respondents whose maximum education lies beyond a high school
diploma, but not beyond an undergraduate degree. There were 179 Respondents in this category.
Respondents whose maximum formal education ended with a degree higher than an undergraduate
degree comprise Category Three. This category includes master’s degrees, law degrees, medical
degrees, and doctorates of various kinds. There were 46 Respondents in this category. In stark contrast
to their refusal to disclose their income, only four of 400 Respondents refused to answer the question
about the level of their education.
When one separates the Respondents by level of education, one sees a similar result for both questions.
As shown in Table 5, the level of disagreement strictly fell and the level of agreement strictly rose with
the level of formal education. For example, those who disagreed with the statement that they knew
where to get a loan were 26.3% for Category One, 12.8% for Category Two, and 0.0% for Category
Three. Similarly, those who disagreed with the statement that they understood the terms of the loan
were 18.7% for Category One, 7.8% for Category Two, and 2.2% for Category Three.
Between the categories of High School and Undergraduate educations, the difference in disagreement
and agreement levels differed in a statistically significant way for both questions. The P-value for the
Chi-Square test is 0.0008 concerning the question of knowing how to get a loan that best suits one’s
needs, and less than 0.0001 concerning the question of understanding the terms of the loans that one
has taken out.
Within each education category, however, there is no statistical difference between the percentages of
Respondents who disagree (or agree) with the two questions. For example, for those with a maximum
formal education of an undergraduate degree, the P-value for the Chi-Square statistic is 0.1182 for the
test of the difference between those who disagree with both statements, and 0.1495 for the test of the
difference between those who agree with both statements.
IV.D. Ordinal Regression Results
Page | 13
Two questions that are a major focus of this paper are whether consumers know where to get a loan
that suits their needs, and whether consumers understand the terms of loans that they have taken. For
the purposes of the results presented in Table 6, Respondents were divided between those whose
maximum education level is a high school degree, versus those who studied beyond high school. Also,
Respondents were divided at roughly the median income level in the state of Mississippi at the time of
the survey.
Table 6 contains Chi-square test results about how Respondents voiced their agreement or
disagreement to these questions. One can see, via the various P-values being less than 0.0500, that the
null hypothesis of no relationship is rejected in each case at traditional statistical levels. That is,
Unbanked Respondents disagree with both questions at a statistically higher rate than Banked
Respondents. The same statistical conclusion is reached concerning educational level and income.
Table 6 contains bivariate results. Studying the combined effects of the existence of a bank account,
education level, and income, however, on how Respondents agree or disagree can lead to a richer
understanding of these relationships.
It is important to note that while nearly one fourth of Respondents refused to disclose their income
level, only one percent refused to disclose their educational level. Because income and education are
well-known to be highly correlated, one feels comfortable that excluding income in multivariate analysis
will not unduly influence the inferences.
Nonetheless, a multivariate technique exists that allows multiple effects to be analyzed with statistical
confidence. This technique is ordinal regression. In ordinal regression, one can study how cumulative
distribution percentages of responses vary.
IV.D.1. Cumulative Distributions
Traditionally, multivariate ordinary least squares would be the choice by researchers to investigate the
effects of more than one explanatory variable. The variable that one is trying to explain is the level of
agreement by Respondents. Because responses to this variable are categorical, ordinary least squares is
a poor choice. Although the categories can be recast ordinal values (5 = strongly agree down to 1 =
strongly disagree), a statistical problem persists.
Page | 14
If one were investigating, for example, the relationship between miles per gallon and driving speed,
both variables are continuous, not ordinal. More importantly, the difference between 15 miles per
gallon and 16 miles per gallon is exactly equal to the difference between 21 miles per gallon and 22
miles per gallon. It is exactly equal to one, and one can measure this difference with precision. There are
no subjective feelings involved. In the analysis here, even though one can recast the categorical
responses as ordinal, one cannot say that the difference between Strongly Agree (=5) and Agree (=4) is
the same as Strongly Disagree (=1) and Disagree (=2). The ordinal rankings are artificially set, and can
differ substantially among Respondents.
Table 7 contains the cumulative percentages between Survey responses, by whether Respondents have
a bank account and by their education level. One can glance at these results and surmise that there is a
difference. Figures 7 and 8 show these results graphically. Again, one can look at these graphs and see
that the lines are separated. The questions, however, are whether this separation is statistically
significant, and what happens when both bank account and education level are included to explain the
cumulative distribution of Survey responses.
IV.D.2. Knowing Where to Get a Loan
Table 8 contains the results of four separate ordinal regressions concerning the responses to the
question about knowing where to obtain a loan. There is a consistent set of statistically significant
results.
Regressions (A) and (B) contain one independent variable: whether the Respondent has a bank account
and whether the maximum education level is high school. The regression coefficients for both variables
are statistically significantly negative. The inference for these two coefficients is that for Unbanked
Respondents, the cumulative percentage of how they agree is lower than for Banked Respondents.
Another way to state this inference is that the Unbanked Respondents are more likely to disagree with
the statement that they know where to go to obtain a loan that suits their needs. One can make the
same inferences for the group of Respondents whose maximum education level is a high school degree.
Regression (C) includes both explanatory variables. Both explanatory variables are statistically
significantly negative. The inference in this regression is that the cumulative percentage agreement for
the group of Unbanked Respondents is lower than it is for Banked Respondents for the question of
whether they know where to go to obtain a loan that suits their needs. This effect is still significant
Page | 15
when the maximum education level of high school is included in the regression. Similarly, the cumulative
percentage agreement for the group of Respondents with a maximum education level of high school is
lower than it is for those whose maximum education level lies beyond a high school degree.
Consequently, one can conclude that both variables are of interest concerning the question.
Regression (D) contains three explanatory variables, each of which measure a joint effect. For example,
the first independent variable is for the group of Unbanked Respondents whose maximum education
level is high school. The second variable is for the group of Unbanked Respondents whose maximum
education level is beyond high school. The third variable is for the group of Banked Respondents whose
maximum education level is high school. Consequently, the reference group for each of these three
groups is the set of Banked Respondents whose education level is beyond high school.
Two of the three variables are statistically significantly negative at the 1 percent confidence level, while
the third has a p-value of 0.063. One can infer that the cumulative percentage agreement for each of
these groups lies below the cumulative percentage agreement for the reference group, but the effect is
weakest between the groups who differ only by whether they have a bank account.
Overall, a bank account and an education level of more than a high school diploma results in the highest
cumulative agreement with the question of whether a Respondent knows how to obtain a loan that
suits their needs. Moreover, both effects matter when trying to explain the difference in cumulative
percentage agreement between groups.
The level of the Nagelkerke Pseudo R-Square suggests that these models are likely to be a poor predictor
of the outcome for any particular Respondent. This conjecture, however, does not rebut the fact that
there is a statistically significant difference between groups of Respondents.
IV.D.3. Understanding the Terms of the Loan
Table 9 contains the results of four separate ordinal regressions concerning the responses to the
question about understanding the terms of loans. The results for this question are not as consistent as
they are for the question about knowing where to go to get a loan to suit the needs of the borrower.
But, these results appear to help distinguish between the effects of the two independent variables.
In regression (A), the independent variable for whether the Respondent does not have a bank account is
insignificant. The interpretation of this result is that the cumulative percentage of how Respondents
Page | 16
agree about understanding the terms of the loan is statistically the same between the groups of Banked
and Unbanked Respondents. Note that this result is at variance with the Chi-Square test presented in
Table 6.
The independent variable in regression (B) is whether the maximum education level of the Respondent
is a high school diploma. The regression coefficient is statistically significantly negative. The inference is
that for the set of Respondents who did not study beyond high school, the cumulative percentage of
agreement is lower than for the group of Respondents who studied beyond high school. Another way to
state this inference is that the group of Respondents who did not study beyond high school is more
likely to disagree with the statement that they understand the terms of their loans.
Regression (C) includes both explanatory variables. Only the explanatory variable for education level is
statistically significantly negative. The inference in this regression is that the cumulative percentage
agreement for the group of Respondents who did not study beyond high school is lower than it is for
those who did. Moreover, this difference between the cumulative percentages does not change when
one accounts for whether the Respondent has a bank account. Consequently, one can conclude that
only the independent variable for education level is of interest concerning the question of whether
Respondents say that they understand the terms of their loans.17
Regression (D) contains three explanatory variables, each of which measure a joint effect. For example,
the first independent variable is for the group of Unbanked Respondents whose maximum education
level is high school. The second variable is for the group of Banked Respondents whose maximum
education level is high school. The third variable is for the group of Banked Respondents whose
maximum education level is beyond high school. Consequently, the reference group for each of these
three groups is the set of Banked Respondents whose education level is beyond high school.
Two of the three variables are statistically significantly negative at the 1 percent confidence level, while
the third is insignificant. One can infer that the cumulative percentage agreement for each of the groups
who have a maximum education of high school lies below the cumulative percentage agreement for the
reference group, regardless of whether the Respondent has a bank account. Moreover, the cumulative
17
Sometimes, when two independent variables are highly correlated, a researcher can incorrectly judge a variable
to be insignificant when, in fact, it is significant. There are two countervailing pieces of evidence in this case
however. First, the correlation coefficient between the two independent variables is the same as it is for the
results presented it Table 8: that is, 0.252. Second, when entering the regression by itself, the independent
variable for whether the Respondent does not have a bank account is insignificant at the 0.0500 level.
Page | 17
percentage agreement for the group with a maximum education beyond high school is not statistically
significantly different from the reference group.
Overall, an education level of more than a high school diploma results in the highest cumulative
agreement with the question of whether a Respondent understands the terms of their loans. Moreover,
this result is seemingly not influenced by whether the Respondent has a bank account.
As with the results in Section IV.D.2, the level of the Nagelkerke Pseudo R-Square suggests that these
models are also likely to be poor predictors of the outcome for any particular Respondent in the Survey.
Again, this inference, however, does not rebut the fact that there is a statistically significant difference
between groups of Respondents.
Unlike the results of the models presented in Section IV.D.2, an important ordinal regression assumption
is upheld for three of the models. An assumption made in the ordinal regression technique is that the
odds ratio is equal at each threshold (i.e., at each break point between agreement levels). This
assumption stems from the fact that the ordinal model constrains these odds ratios to be equal through
the proportional odds assumption. One examines the appropriateness of this assumption by using the
test of parallel lines. This test compares the ordinal model that has one set of coefficients for all
thresholds to an unconstrained model with a separate set of coefficients. If the unconstrained model
yields a significantly better result than the constrained model, then one rejects the assumption of
proportional odds. In Regressions (B), (C), and (D), one does not reject the null that the odds ratio is
equal at each threshold.
V. Conclusions
Few areas spawn so many emotionally charged opinions and controversy as does the confluence of nonbank supplied credit products and the financial condition of low-income citizens. To add information to
the debate, the focus of this paper is on consumer use of non-bank supplied consumer credit products.
Specifically, the aim of this research is to document what consumers themselves have to say about
which credit products they use. The main research motivation is to seek some evidence on two
previously unexplored questions: 1) How do consumers respond to the questions of whether they know
how to obtain loans that suit their needs, and 2) Do consumers understand the terms of the loans made
to them? The research method used in this study is a telephone survey that resulted in a random sample
Page | 18
of 400 Mississippians. Further evidence concerning the answers to these two questions is provided
based on whether Respondents have a bank account, as well as their income and education level.
In the sample of consumers surveyed, Respondents overwhelmingly say that they know where to go to
get loans that suit their needs and they overwhelmingly say that they understand the terms of their
loans.
Perhaps not surprisingly, Banked Respondents saved at a higher rate than the Unbanked. Evidence
presented in this research shows that Unbanked Respondents tend to be the users of payday loans, title
loans, pawn loans, and the services of rent-to-own companies. Additional evidence presented shows
that income level and the likelihood of having a bank account are related, as are income level and the
type of credit product employed.
Concerning the question of whether people know where to go to get a loan that suits their needs, two
factors studied matter when looking deeper at the question. That is, whether the Respondent has a
bank account or not matters, and so does whether their education level stops at high school.
Statistically, however, one cannot be sure that the results hold across all levels of agreement: from
“strongly disagree” to “strongly agree.”
Concerning the question of whether people understand the terms of their loans, only one of the factors
studied matters. What matters is whether the educational level of the Respondent stops at high school
or not. Whether they have a bank account or not does not matter, statistically speaking. Statistical tests
suggest that one cannot reject the assumption that these results hold across all levels of agreement:
from “strongly disagree” to “strongly agree.”
There are many ways for consumers to obtain non-bank supplied credit. These products have starkly
different features and clientele. It is incumbent upon all discuss to non-bank supplied credit to recognize
these differences. That is, it is imperative to avoid lumping all these products together under the
umbrella of “high cost loans” or “small-dollar loans.” These credit products differ substantially from one
another, just as do other consumer services.
Viewed under the finding by Lusardi, Schneider, and Tufano (2011) that at least one fourth of American
households could not come up with $2,000 in thirty days, the non-bank supplied loan space takes on
even more importance. Because of the economics of their business model, banks generally do not lend
sums of $2,000 and less. Recent research documents the importance of keeping non-bank supplied
credit products available to consumers. For example:
Page | 19

Zywicki and Okolski (2009) conclude: “The bottom line is that restrictions on auto title lending
will eliminate an important funding option for many consumers, especially those of lower
income, and will incentivize the use of more risky or dangerous credit channels.”

Fritzdixon, Hawkins, and Skiba (2014) conclude: “Instead of banning title lending, policymakers
should foster a market with information that will help customers understand the true cost of
title loans.”

Morse (2011) studies the payday loan market and finds that “communities with payday lenders
show greater resiliency to natural disasters” and her estimates of welfare measures “suggest
that payday lending enhances the welfare of communities.”
Evidence in the current study shows that consumers prefer products that will not trap them in the socalled debt-trap. The Survey Respondents used payday loans least often. About three times as many
Respondents used Title Loan and Pawn Loans—both of which are secured by collateral and are nonrecourse. Another way a consumer can `1avoid being in a cycle of debt, is to borrow money using an
installment loan. Evidence from the Survey shows that many more Respondents used the services of
Finance Companies, whose product is an installment loan, than any non-installment loan product.
Consumers used the services of Credit Unions and Banks even more than they used Finance Companies.
Page | 20
Appendix. Five Non-Bank Provided Personal Loan Choices.
This appendix contains a brief overview of five widely-available personal loan and financing products
available to Mississippi residents at the time of the 2011 Economy Watch Survey. (There are some other
sources of consumer loans that did not appear in the Survey.18) Some of the products that are available
in Mississippi have been legislated out of existence in certain other states. For example, the
constitutional interest rate cap of 17% in Arkansas means that traditional installment loans are
unprofitable for lenders. As a result, Arkansas residents who seek an installment loan must travel to a
state that borders Arkansas to obtain a traditional installment loan.19 In addition, payday lending has
recently been legislated out of existence in Arkansas.20
A1. Lease to Own Transactions
Lease-to-Own, or commonly Rent-to-Own, agreements involve the transfer of possession of a consumer
good by the lessor to the lessee, in exchange for a weekly or monthly payment by the lessee to the
lessor. Generally, these assets fall into the categories of furniture, electronics, or appliances. Although
these transactions are not loans, they do represent a method of financing the acquisition of an asset.
The lease generally has a monthly term. At the end of each term, the consumer (lessee) has the option
to renew the lease for another term or terminate the agreement. If the consumer chooses to terminate
the agreement, the consumer has no further obligation after the consumer returns the leased asset to
the lessor. If the consumer makes the agreed upon number of payments, the consumer receives title to
the asset.
As an example, a recent rent-to-own circular21 offered a 60” Smart LED 1080p TV (and stand) at a “Buy it
Now” price of $1,648.99, plus tax. A consumer can lease-to-own this television with 24 monthly
payments of $109.99. If the consumer purchases a service plan that includes a lifetime reinstatement
18
A non-exhaustive list of other loan products includes: Bank Deposit Advance Loans, Income Tax Refund
Anticipation Loans, Online Payday Loans, Credit Cards (including Subprime Credit Cards), Peer-to-Peer Loans, and
Installment Loans from Sovereign Nations within the US.
19
See Durkin, Elliehausen, and Hwang (2014) and Lukongo and Miller (2014).
20
See Wann and Wann (2012) for a history of payday lending in Arkansas.
21
http://www.shoplocal.com/aarons/weekly-ads-sales/-98053/
Page | 21
option, the total monthly payment is $120.98.22 Including the service plan, the total lease-to-own cost of
the TV is $2,903.52. The circular states the cost of the lease service as $1,254.53. A 24-month lease with
monthly payments of $120.98 and a “Buy it Now” price of $1,648.99 implies an annual percentage rate
(APR) of 61.50%.23
According to a 2000 Federal Trade Commission survey on the rent-to-own industry,24 Respondents
chose a rent-to-own transaction for a variety of reasons. Notably, they include: 1) “the lack of a credit
check,” 2) “the ability to obtain merchandise they otherwise could not” and, 3) “the convenience and
flexibility of the transaction.” In the survey, the most common reason Respondents cited for
dissatisfaction was dissatisfaction with the lessor’s high prices.
A2. Pawnbroker Loans
Durkin, Elliehausen, Staten, and Zywicki (2014) state “Pawnbroker loans are among the oldest forms of
credit, stretching back to antiquity.” Despite this long history, financial economists have devoted little
collective effort to studying the pawn industry and its benefits to consumers. Bos, Carter, and Skiba
(2012), discuss the handful of academic studies and describe the pawn process in detail.
In a pawn transaction, the consumer offers a tangible item to the pawnbroker, who pays cash to the
consumer and takes possession of the item. Today, the most commonly pawned items are jewelry.
Consumer electronic equipment, firearms, tools, and musical instruments are also frequently pawned.
The pawnbroker will generally ask whether the consumer wants to sell or pawn the item.25 If the
consumer wishes to pawn the item, the parties negotiate the amount that the pawnbroker will “loan”
on the item. If there is an agreement, the consumers delivers possession of the item to the pawnbroker,
and the pawnbroker gives the agreed upon cash to the consumer and issues a pawn ticket that precisely
details the terms of the transaction and redemption.
22
From the circular: “LIFETIME REINSTATEMENT: After a minimum period of time you can return a leased product
and later reactivate that least at the same store. You will receive the same or comparable condition item as
returned and pay on the remaining balance of your original lease.”
23
A similar TV could be purchased for a price between $800 and $1,050 (plus stand) at a well-known electronics
chain (http://www.bestbuy.com, navigate to LED TVs, select Vizio 60-64”.) Of course, if a consumer buys the TV
outright and finances it, the consumer loses the option to abandon the lease. Nevertheless, a lease payment
$120.98 for 24 months on a base of $1,050 implies an annual percentage rate of about 125%.
24
Lacko, McKernan, and Hastak (2000).
25
The pawnbroker will also likely ask how the consumer acquired the item. This question helps the pawnbroker
ascertain whether the consumer is the rightful owner of the item.
Page | 22
In a pawn transaction, the consumer does not need to show any proof of income or credit history. The
pawnbroker does not report the customer’s performance on the transaction to a credit-reporting
agency. A pawn transaction is not a loan in the traditional sense because the consumer has no obligation
to repay the sum obtained in the pawn transaction. The pawnbroker has no recourse if the customer
abandons the pawned item. One can view a pawn transaction, therefore, as a sale with a renewable,
month-to-month repurchase agreement.
In the typical pawn transaction, to redeem the pawned item, the consumer must pay various charges for
interest, storage, and other fees, in addition to the sum originally advanced by the pawnbroker. The
maximum allowable fees vary according to state law.26 In Mississippi, § 75-67-313 (2013) sets these
amounts: “A pawnbroker may contract for and receive a pawnshop charge in lieu of interest or other
charges for all services, expenses, cost and losses of every nature not to exceed twenty-five percent
(25%) of the principal amount, per month, advanced in the pawn transaction.”27
As an example of a pawn loan, suppose a Mississippian brings a mounted moose head to a pawnshop.
The pawn dealer assesses the pawn value of this personal treasure as $500. If the mounted moose head
has considerable sentimental value to the customer, say $1,000, the customer is likely to redeem the
pawn ticket. Assuming maximum allowable charges, at the end of the month the consumer has three
choices: 1) abandon the property, 2) extend the pawn another month by paying $125 (=0.25 times $500)
or, 3) pay $625 and reclaim the property.
A3. Payday Loans
A Payday loan is a short-term, lump sum loan. Most of the loans are a term of 30 days or less. (Payday
loans are also known as cash advance loans, delayed deposit loans, and deferred presentment loans.) In
a traditional payday loan, a borrower writes a check to a lender in exchange for a short-term cash loan.
The lender agrees not to cash the check until a date specified in the loan agreement.
To obtain a payday loan, lenders generally require borrowers to have an active checking account,
provide proof of income, show valid identification, and be at least 18 years old. Payday lenders generally
do not require a traditional credit report.
26
As a beginning reference, http://www.pawnshopsonline.info/state-regulation-interest-rates-comparisons/.
27
ARTICLE 7. MISSISSIPPI PAWN SHOP ACT. Sources: Laws, 1993, ch. 598, § 1, effective from and after
July 1, 1993. Navigate to http://www.dbcf.state.ms.us/documents/cons_finance/pawnshopact.pdf.
Page | 23
Payday lending exists throughout the US. As of March 13, 2014, according to the website for the
National Conference of State Legislatures: “Thirty-eight states have specific statutes that allow for
payday lending. Eleven jurisdictions do not have specific payday lending statutory provisions and/or
require lenders to comply with interest rate caps on consumer loans….[while]…Arizona and North
Carolina allowed pre-existing payday lending statutes to sunset. Arkansas repealed its pre-existing
statute in 2011.”28
Under the Mississippi Check Cashers Act,29 a payday loan agreement must disclose the terms of the loan,
including the loan amount and the annual percentage rate (APR). The lender will generally require the
borrower to write a personal check for the loan principal plus a loan fee, i.e., interest on the loan. The
loan agreement might allow the lender to withdraw (or attempt to withdraw) the sum owed from the
borrower’s bank account, i.e., cash the check, at the loan due date—regardless of whether the borrower
has sufficient funds in the account.30 If the borrower does not have sufficient funds, the borrower will be
subjects to Non-Sufficient Funds (NSF) fees.
Under Mississippi law, the largest check a payday loan borrower can write is for $500. The amount of
the check must include the loan principal and allowable fees. For a check written for $250 or less,
Mississippi law allows a payday lender to charge a fee of up to $20 per $100 advanced to the
borrower.31 For example, if a borrower writes a check for $240, the lender advances $200 to the
borrower and keeps the check, which includes $40 in fees. Assuming this loan is for two weeks, the
Annual Percentage Rate is: $40/$200 times 26 = 520%.
B4. Title Loans
A vehicle title loan is similar to a pawn loan, but with an important difference: in a pawn transaction, the
consumer gives possession of the item to the pawnbroker; under the terms of a title loan, the borrower
retains possession of the pledged collateral. As in a pawn loan, if the borrower defaults on a title loan,
28
http://www.ncsl.org/research/financial-services-and-commerce/payday-lending-state-statutes.aspx. For more
information on the legal status of payday loans by state, navigate to http://www.paydayloaninfo.org/stateinformation.
29
Miss. Code Ann. § 75-67-501 (2013).
30
If the lender deposits the check, but the bank returns it unpaid, the lender can charge only one $30 NSF fee—
and only if the agreement disclosed the NSF fee.
31
For checks written for an amount that exceeds $250 up to $500, Mississippi law allows a charge of no more than
$21.95 per $100 advanced to the borrower.
Page | 24
ownership of the collateral (the vehicle) is transferred to the lender. Like pawn loans, title loans are nonrecourse. If the borrower defaults on the loan, the lender can sell the vehicle. If the vehicle is sold for an
amount that is less than the amount owed, the borrower does not have to make up the difference. If,
however, the vehicle is sold for more than the outstanding amount owed, the borrower might
participate in the excess sale proceeds.32
A basic title loan is a one-month lump sum loan with the principal and interest due at the end of the
month. If the borrower cannot repay the principal, the title lender can allow an interest-only payment to
roll the loan over for another month.
The variety in state laws around the country makes for differences in the title loan transaction.33 In
Mississippi, MS Code § 75-67-413 (2013) states “A title pledge lender may contract for and receive a title
pledge service charge in lieu of interest or other charges for all services, expenses, cost and losses of
every nature not to exceed twenty-five percent (25%) of the principal amount, per month, advanced in
the title pledge transaction.” States also regulate the loan amount of in a title loan. In Mississippi, the
maximum amount is $2,500.
The application process for a title loan is straightforward. To secure a title loan, the borrower must have
a clear title to the vehicle, and the borrower must allow the title lender to place a lien on the vehicle.
According to Hawkins (2012), the borrower might, sometimes need to provide references and proof of
income. The borrower does not need to provide a credit history.
Suppose a Mississippian brings a 2003 Chevrolet Tahoe to a title lender. The title lender can inspect the
vehicle, if present, and/or look up values for similarly equipped vehicles. Suppose this vehicle has a
wholesale appraisal of about $3,900 and the lender makes a loan of $2,500. Assuming maximum
allowable charges, at the end of the month the Mississippian has three choices: 1) transfer ownership to
the title lender, 2) extend the loan for another month by paying the title lender $625+$250 (=0.25 times
$2,500, plus a required 10% reduction in the principal) or, 3) pay $3,125 and reclaim the vehicle.
B5. Finance Company Installment Loans
32
In Mississippi, 85% of the surplus from the sale goes to the borrower. See: MS Code § 75-67-411 (2013).
As a beginning resource, navigate to: http://www.responsiblelending.org/other-consumer-loans/car-titleloans/tools-resources/car-title-lending-by-state.html.
33
Page | 25
In the early 1900s a battle raged against illegal “loan sharks,” and an alternate new loan source emerged
through the collaboration of lenders who wanted to offer this new product and consumer advocates,
notably Mr. Arthur H. Ham of the Russell Sage Foundation. What emerged was The Universal Small Loan
Law written in 1916. By 1940, all but nine states had adopted some version of this proposed law.
The striking feature of this law was that it allowed for interest rates higher than allowed under existing
usury laws. Of course, illegal “loan sharks” and those who favored low interest rate ceilings lobbied long
and hard against this legislation. When collaborating on the Uniform Small Loan Law, the parties agreed:
1) Legal installment lenders must be able to earn a reasonable profit. Therefore, the interest rate was
initially set at 3 to 3.5 percent per month; 2) Small loans were defined as “up to $300” (in today’s
dollars, about $7,137), and; 3) The interest rate would be re-examined periodically to sustain the
industry.
In Mississippi, MS Code § 75-17-21 (2013) sets the maximum allowable finance charges by licensees
operating under the Small Loan Regulatory Law. The maximum interest rate is about the same as that
set forth in the Universal Small Loan Law of 1916. For an unpaid balance up to $1,000, the maximum
annual rate is 36 percent (3 percent per month). For amounts over $1,000 to $2,500 the maximum rate
is 33 percent; for amounts over $2,500 to $5,000 the maximum rate is 24 percent, and; for amounts
over $5,000 the maximum allowable annual rate is 14 percent. In addition, a licensee can contract for
and charge a closing fee of 4 percent or $25, whichever is greater, for loans of $10,000 or less.
As an example of an installment loan, suppose a consumer wants to borrow $1,000 to pay for vehicle
repairs. The terms of the loan are 12 months, an annual interest rate of 36 percent (3 percent per
month), and no closing fee (for ease of calculation). To calculate the loan payment, we use the following
two equations:
Where:
.
Page | 26
In this example, the Present Value Factor is
the resulting monthly payment is
and we can calculate $C = $100.46. The total of interest and principal payments equals the payment
times the number of payments, or $100.46 times 12, or $1,205.55. The consumer borrowed $1,000, so
the consumer pays $205.55 in interest over the life of the loan. Notice that the consumer does NOT pay
$1000 times .36, or $360 of interest. The difference between $360 and $205.55 occurs because the
amount owed each month declines, or amortizes, over the length of the loan. Therefore, even though
the interest rate of 36 percent determines the size of the installment payment, the interest income
received by the lender is $205.55, or 20.56 percent of $1,000.
Page | 27
Table A1. Summary of Products, by Loan Source.
Loan Source
Mainline Consumer Loan Product
Approximate Minimum
Loan in Mississippi
Rent-to-Own
Renewable Periodic Lease
Depends on Item
12-24 Months, Renewable Monthly
Pawn
Renewable Repurchase Agreement
Depends on Item
One Month, Renewable
Payday
Lump Sum Loan--Unsecured
$200
Two Weeks, Can be Rolled-Over
Title
Lump Sum Loan--Secured
Maximum is $2,500
Finance Company
Installment Loan
$500
6-24 Month
Bank
Installment Loan
$2,000
24 Months
Typical Term
Loan Source
Loan Process
Check Credit
Can
Help
Can
Report to
Improve
Hurt
Credit Bureau Credit Score Credit Score
Rent-to-Own
Verifiable Income; References
Minimal
Sometimes
No
Yes
Pawn
Negotiation about value of item pawned
Mimimal
No
N/A
N/A
Payday
Verifiable Income
Minimal
Noa
N/A
N/A
Title
Negotiation, vehicle value; verifiable income Minimal
No
N/A
N/A
Finance Company
Detailed Underwriting Process
Yesb
Yes
Yes
Yes
Bank
Detailed Underwriting Process
Yes, with Credit Bureaus Yes
Yes
Yes
Page | 28
REFERENCES
Bos, Marieke, Carter, Susan Payne, and Paige Marta Skiba. 2012. “The Pawn Industry and Its Customers:
The United States and Europe,” Law and Economics Working Paper Number 12—26, Vanderbilt
University Law School. SSRN: http://ssrn.com/abstract=2149575.
Calder, Lendol. 1999. “Financing the American Dream: A Cultural History of Consumer Credit,” Princeton
University Press, Princeton, New Jersey.
Carruthers, Bruce G.; Guinnane, Timothy W., and Yoonseok Lee. 2012. “Bringing ‘Honest Capital’ to Poor
Borrowers: The Passage of the Uniform Small Loan Law, 1907-1930.” Journal of Interdisciplinary History,
Vol. 42:3, 393-418.
Durkin, Thomas A., Elliehausen, Gregory, Staten, Michael E, and Todd J. Zywicki. 2014. “Consumer Credit
and the American Economy.” Oxford University Press, New York, New York.
Fritzdixon, Kathryn, Hawkins, Jim, and Paige Skiba, 2014. "Dude, Where's My Car Title? The Law,
Behavior, and Economics of Title Lending Markets." University of Illinois Law Review, August, pp. 1013-1058.
Ham, Arthur H., 1922. “Small Loan Legislation: Progress and Improvement.” Russell Sage Foundation,
Division of Remedial Loans, Pamphlets. RL 35. New York, New York.
Hawkins, Jim, 2012. “Credit on Wheels: The Law and Business of Auto Title Lending,” Washington and
Lee Law Review 69, No. 2.
Lacko, James M., McKernan, Signe-Mary, and Manoj Hastak. 2000. “Survey of Rent-to-Own Customers,”
Federal Trade Commission, Bureau of Economics Staff Report.
Lukongo, Onymbe (Ben) and Thomas W. Miller, Jr. 2014. “The Consequences of the Constitutional
Interest Rate Cap in the State of Arkansas.” Working paper, Mercatus Center, George Mason University.
Lusardi, A., Schneider, D. J., and Tufano, P. 2011. “Financially fragile households: Evidence and
Implications.” Brookings Papers on Economic Activity, 83–150.
Morse, Adair. 2011. “Payday Lenders: Heroes or Villains?” Journal of Financial Economics 102 (1), p.2844.
Olney, Martha L. 1991. “Buy Now, Pay Later: Advertising, Credit, and Consumer Durables in the 1920's.”
University of North Carolina Press, Chapel Hill, North Carolina. (out of print)
Sifakis, Carl. 2005. The Mafia Encyclopedia, 3rd Ed., Facts on File Books, New York, New York.
Wann, Garry S. and Christi R. Wann. 2012. “The History of Payday Lending in Arkansas,” Journal of the
Academy of Business and Economics, 12, Issue 1.
Zywicki, Todd and Gabriel Okoloski. 2009. “Potential Restrictions on Title Lending,” Mercatus on Policy,
Mercatus Center, George Mason University 62.
Page | 29
Table 1. Summary of Consumer Loan Outlets in Mississippi in 2011, by Category
Mississippi
Licenses
as of 12/2011 Transactions
Loans
Outstanding
Four Consumer Credit Choices
State of Mississippi Category:
Author Description of Loan Product:
Check Casher/Payday Loan
Lump Sum Short-Term Loans
Small Loan
1,053
180,884
$45,709,994
Personal Installment Loans
525
327,892
$803,273,591
Title Pledge
Lump Sum Loan Secured by Vehicle Title
402
786,377
$30,845,280
Pawnbroker
Non-Recourse Repurchase Agreement
219
540,548
$10,269,227
Source: Mississippi Department of Banking and Consumer Finance, Annual Report 2011
http://www.dbcf.state.ms.us/documents/annual-report11.pdf
Page | 30
Table 2. Weighting Demographics
Survey
Observations Percent
Weighted
Observations Percent
2010
Census
Percent
A. "Which race do you most identify yourself with?"
White
Black
All Other
Refused
245
139
13
3
61.3%
34.8%
B. "Do you consider yourself Hispanic/Latino?"
Yes
No
Don't Know
253
137
7
3
63.3%
34.3%
63.1%
34.9%
1
398
1
C. "What is the respondent's gender?"
Male
Female
Refused
147
251
2
36.8%
62.8%
185
213
2
46.3%
53.3%
47.7%
52.3%
19
72
162
130
17
4.8%
18.0%
40.5%
32.5%
52
135
125
68
17
13.7%
35.5%
32.9%
17.9%
13.6%
33.8%
33.8%
18.8%
D. "What is your age?"
18-24
25-44
45-64
65+
Refused
Source for 2010 Cens us da ta:
http://www.cens us .gov/popes t/da ta/s tate/a s rh/2013/i ndex.html
Page | 31
Table 3. Savings Habits and Plans of Respondents
Yes
No
a
Did you Save Last Year?
Percent (of 400):
Do you Plan to Save Next Year?
Percent (of 400):
202
50.5%
318
79.5%
Yes
a
Saved Last Year, plan to this year?
Percent (of 202):
b
Did Not Save Last Year, Plan to this Year?
Percent (of 195):
Bank Acct, Yes:
Yes
No
DK/NS
Refused
Chi2, Equal Proportions:
P-Value:
Bank Acct, No:
Yes
No
DK/NS
Refused
Chi2, Equal Proportions:
P-Value:
DK / NS
Refused
b
2
0
65
16
1
195
No
DK / NS
Refused
191
94.6%
7
3
0
126
64.6%
58
11
0
Did you Save Last Year?
Yes
No
DK/NS Refused
174
145
0
0
21
42
0
0
4
7
2
0
3
2
0
0
9.5
0.0021
Did you Plan to Save Next Year?
Yes
No
DK/NS Refused
261
49
10
1
44
14
4
0
8
2
2
0
4
0
0
0
2.4
0.1221
Actual Survey Questions:
In the past year, have you been able to save any money for the future?
In the coming year, do you plan to save any money for the future?
Does anyone in your household (including yourself) currently have a
checking or savings account at a bank?
Page | 32
Table 4. Respondent's Answers to Whether They Used this Credit Product
Panel A. Non-Installment Loan Products
With Bank Account
Percent Yes on Product
Without Bank Account
Percent Yes on Product
Chi2, Equal Percentages:
P-Value:
Pawn Shop
Yes
No
12
284
4.1%
9
15.3%
50
11.1
0.0009
Payday Lender
Yes
No
5
291
1.7%
5
8.5%
54
Title Lender
Yes
No
8
288
2.7%
7
11.9%
52
8.3
0.0040
10.2
0.0014
Credit Union
Yes
No
59
237
19.9%
Bank
Yes
No
166
130
56.1%
Panel B. Installment Loan Products
With Bank Account
Percent Yes on Product
Without Bank Account
Percent Yes on Product
Chi2, Equal Percentages:
P-Value:
Finance Company
Yes
No
40
256
13.5%
3
5.1%
3.3
0.0700
56
5
8.5%
4.4
0.0366
54
14
23.7%
45
20.6
0.0000
Page | 33
Table 5. Percentages Overall and Percentages by Bank Account, Income, and Education
Would you Say You Know How to
Obtain a Loan that Suits Your Needs?
Strongly Disagree
2
3
4
Strongly Agree
DK / NS
Refused
Total:
46
25
40
50
212
26
1
400
All Respondents
11.5%
17.8% Disagree
6.3%
10.0%
12.5%
65.5% Agree
53.0%
6.5%
0.3%
Would You Say You Understand the
Terms of Loans You Have Taken Out?
33
18
35
43
242
27
2
400
Bank Account:
Strongly Disagree
2
3
4
Strongly Agree
DK / NS
Refused
Total:
Strongly Disagree
2
3
4
Strongly Agree
DK / NS
Refused
Total:
Strongly Disagree
2
3
4
Strongly Agree
DK / NS
Refused
Total:
No
16
7
7
2
28
2
0
62
25.8%
11.3%
11.3%
3.2%
45.2%
3.2%
0.0%
Yes
26
16
32
48
176
22
1
321
All Respondents
8.3%
12.8% Disagree
4.5%
8.8%
10.8%
71.3% Agree
60.5%
6.8%
0.5%
Bank Account:
8.1%
5.0%
10.0%
15.0%
54.8%
6.9%
0.3%
Income:
< $35,000
> $35,000
28
17.9%
7
5.3%
11
7.1%
0
0.0%
17
10.9%
14
10.6%
16
10.3%
23
17.4%
79
50.6%
85
64.4%
6
3.8%
2
1.5%
0
0.0%
0
0.0%
157
131
Maximum Education Degree:
High School
College
Grad School
30
17.5%
15
8.4%
0
0.0%
15
8.8%
8
4.5%
0
0.0%
16
9.4%
24
13.4%
0
0.0%
14
8.2%
29
16.2%
8
17.4%
81
47.4%
97
54.2%
34
73.9%
14
8.2%
6
3.4%
4
8.7%
1
0.6%
0
0.0%
0
0.0%
171
179
46
No
12
5
2
4
35
5
0
63
19.4%
8.1%
3.2%
6.5%
56.5%
8.1%
0.0%
Yes
17
11
30
39
201
21
1
320
5.3%
3.4%
9.3%
12.1%
62.6%
6.5%
0.3%
Income:
< $35,000
> $35,000
19
12.2%
6
4.5%
8
5.1%
1
0.8%
13
8.3%
8
6.1%
13
8.3%
23
17.4%
96
61.5%
91
68.9%
7
4.5%
3
2.3%
0
0.0%
0
0.0%
156
132
Maximum Education Degree:
High School
College
Grad School
22
12.9%
10
5.6%
0
0.0%
10
5.8%
4
2.2%
1
2.2%
15
8.8%
19
10.6%
1
2.2%
17
9.9%
24
13.4%
2
4.3%
91
53.2%
114
63.7%
37
80.4%
15
8.8%
8
4.5%
4
8.7%
1
0.6%
0
0.0%
0
0.0%
171
179
45
Page | 34
Table 6. Chi-Square Tests
Would you Say You Know How to
Obtain a Loan that Suits Your Needs?
Disagree
Agree
Chi 2, Equal Percentages:
P-Value:
Disagree
Agree
Chi 2, Equal Percentages:
P-Value:
Disagree
Agree
Chi 2, Equal Percentages:
P-Value:
Bank Account
Yes
No
42
23
224
30
20.8
0.0000
Max. Education Level
High School Grad. School
48
23
94
167
22.8
0.0000
Income < $35000?
Yes
No
39
32
95
167
8.1
0.0044
Would You Say You Understand the
Terms of Loans You Have Taken Out?
Disagree
Agree
Chi 2, Equal Percentages:
P-Value:
Disagree
Agree
Chi 2, Equal Percentages:
P-Value:
Disagree
Agree
Chi 2, Equal Percentages:
P-Value:
Bank Account
Yes
No
28
17
240
39
15.4
0.0001
Max. Education Level
High School Grad. School
35
15
108
177
17.9
0.0000
Income < $35000?
Yes
No
27
23
109
176
4.4
0.0364
Page | 35
Table 7. Cumulative Percentages by Bank Account and Education
Would you Say You Know How to
Obtain a Loan that Suits Your Needs?
Strongly Disagree
Disagree
Neither
Agree
Strongly Agree
Strongly Disagree
Disagree
Neither
Agree
Strongly Agree
Cumulative Percentages:
Unbanked
Banked
25.8%
8.1%
37.1%
13.1%
51.6%
30.0%
54.8%
44.9%
100.0%
100.0%
Cumulative Percentages:
Max Ed., High School
Max Ed., Grad. School
17.5%
7.0%
26.3%
11.3%
44.4%
27.0%
52.6%
43.0%
100.0%
100.0%
Would You Say You Understand the
Terms of Loans You Have Taken Out?
Strongly Disagree
Disagree
Neither
Agree
Strongly Agree
Cumulative Percentages:
Unbanked
Banked
19.4%
5.3%
27.4%
8.7%
37.1%
24.5%
43.6%
36.7%
100.0%
100.0%
Cumulative Percentages:
Max Ed., High School Max Ed., Grad. School
Strongly Disagree
12.9%
4.8%
Disagree
18.7%
7.9%
Neither
36.8%
22.7%
Agree
46.8%
34.1%
Strongly Agree
100.0%
100.0%
Page | 36
Table 8. Ordinal Regression Results, I.
Would you Say You Know How to
Obtain a Loan that Suits Your Needs?
Intercepts
(A)
(B)
(C)
(D)
Strongly Disagree
Disagree
Neither
Agree
-2.20
-1.679
-0.756
-0.226
-2.334
-1.817
-0.894
-0.361
-2.415
-1.886
-0.948
-0.411
-2.43
-1.902
-0.962
-0.425
Unbanked
Standard Error
Wald, Significance:
-0.819
0.251
0.001
Maximum Education, High School
Standard Error
Wald, Significance:
-0.675
0.258
0.009
-0.61
0.191
0.001
-0.512
0.197
0.009
Max. Ed. High School * Unbanked
Standard Error
Wald, Significance:
-1.13
0.301
0.000
Max. Ed. Grad School * Unbanked
Standard Error
Wald, Significance:
-0.834
0.449
0.063
Max. Ed. High School * Banked
Standard Error
Wald, Significance:
-0.549
0.214
0.010
Nagelkerke Pseudo R-Square:
0.025
0.027
0.042
0.043
Chi-Square Test of Parallel Lines
Significance
11.34
0.010
9.87
0.020
17.84
0.007
58.26
0.001
Page | 37
Table 9. Ordinal Regression Results, II.
Would You Say You Understand the
Terms of Loans You Have Taken Out?
Intercepts
(A)
(B)
(C)
(D)
Strongly Disagree
Disagree
Neither
Agree
-2.50
-2.01
-0.97
-0.49
-2.72
-2.23
-1.18
-0.69
-2.75
-2.26
-1.20
-0.71
-2.74
-2.24
-1.19
-0.70
Unbanked
Standard Error
Wald, Significance:
-0.48
0.26
0.065
Maximum Education, High School
Standard Error
Wald, Significance:
-0.31
0.27
0.250
-0.62
0.20
0.002
-0.58
0.21
0.005
Max. Ed. High School * Unbanked
Standard Error
Wald, Significance:
-0.94
0.31
0.002
Max. Ed. High School * Banked
Standard Error
Wald, Significance:
-0.54
0.22
0.015
Max. Ed. Grad. School * Unbanked
Standard Error
Wald, Significance:
-0.13
0.50
0.801
Nagelkerke Pseudo R-Square:
0.008
0.027
0.030
0.030
Chi-Square Test of Parallel Lines
Significance
10.06
0.018
2.93
0.402
11.37
0.078
15.15
0.087
Page | 38
Page | 39
Page | 40
Page | 41
Page | 42
Page | 43
Page | 44
Page | 45
Page | 46