Following my conversation with geographer-turned-payday-loan-opponent Dr. Stephen Graves, I did my due diligence by investigating his other work in the area of payday loans. Since I don’t want to be accused of being an ideologue, I felt I should give his work a fair shake to see if his position could be supported.

I read a paper he co-authored with Prof. Chris Peterson at the S.J. Quinney College of Law at the University of Utah. Regrettably, the paper is filled with false and misleading statements, faulty premises, empty hypotheses, weak methodology, and completely unjustified conclusions. I was flabbergasted that the paper was published in Catholic University’s Law Review.

In response, I provide the following critique of the paper, examining a few portions to demonstrate its flaws. The critique is not intended to be exhaustive, but to highlight the weakest portions of the work.

A primary criticism relates to the authors’ obvious inherent bias against payday loans, as evidenced by the frequent use of pejorative adjectives and hyperbole to describe the industry. The use of this language immediately undermines the work’s credibility, suggesting the authors’ were more interested in seeking data that supported their biased position than preparing a truly objective report. For example, the authors refer to defenders of the industry as “apologists” [p.2,7]; that payday lenders “poured into American neighborhoods like water over a breached dam” [p.25]; in describing the number of PDL storefronts, the authors’ render a biased judgment even before stating their conclusion: “For those who are concerned about the social, moral, and even spiritual well being of the lower and moderate income Americans, this is a profound, unprecedented, and troubling change in the American nation. “ [p.26]

False and Misleading Statements

1) “Payday lenders generally supplement revenue from…hefty late payment fees, insufficient funds fees and attorney fees” [P.6]

False. Late payment fees are only permitted in three states, hardly enough to qualify it as being something that is done “generally”. Attorney’s fees do not supplement income. They reimburse the attorney retained to collect on defaulted loans in court, as is customary.

NSF Fees do supplement income. However, it is standard practice in all business to charge a fee for a bounced check.

2) “Critics…point to a seemingly endless supply of consumer horror stories…” [p.2]

Any journalist will admit that a story about a bad consumer experience is more inflammatory and of interest to editors and readers than stories about consumers being helped. The absence of positive stories in the media only serves to fuel the ideological opposition to PDLs, but does not provide any scientific evidence regarding the product’s alleged harm.

3) “Unfortunately, payday loan borrowers frequently are unable to pay off their loans after the initial loan term”

False. 94% of all payday loans are paid off on time, per SEC filings of the six public companies in the PDL space.

4) “The best available nationwide estimate suggests that the average payday loan borrower repays $793 for a $325 loan”.

False. The authors rely on a biased source – the Center for Responsible Lending – for this data. Instead, they should have examined the Annual Report of Advance America, the largest PDL in the country. For the year ended 2008, the “Aggregate principal amount of cash advances originated was $4.296,493,000

If the CRL’s “best nationwide estimate” were true, then total revenue from customer advances would be $793 divided by $325, or 2.44 times the amount advanced. Revenue should therefore be $10,483,442,920.

In truth, fees and interest charged to customers for year ended 2008 was $676,346,000. This is equivalent to borrowers repaying $376 for a $325 loan, far less than the CRL claims. Approximately the same amounts can be found in all public company filings.

The authors’ statement, at best, demonstrates ignorance as to where proper data can be culled. At worst, they could be accused of deliberately omitting this information in favor of false data.

The CRL is a public relations front for the Self-Help Foundation, itself a credit union which uses strong-arm tactics on delinquent borrowers, and was founded by Herb and Marion Sandler. The Sandler’s, through World Savings, were major purveyors of option-ARM loans that cratered our economy. As a credit union, Self-Help is in direct competition with payday lenders, and have engaged in aggressive and misleading campaigns to have them banned in several states.

Furthermore, the CRL’s “studies” have been repeatedly disproven as being biased and methodologically unsound. They claim to be non-partisan yet they take funding from Soros and have strong ties with ACORN.

Sources:
http://lwvmilwaukee.org/critique_of_race_matters.pdf
http://www.nytimes.com/2008/12/25/business/25sandler.html
http://www.kivitv.com/Global/story.asp?S=10262229
http://www.reuters.com/article/pressRelease/idUS104573+10-Apr-2009+PRN20090410

Interested readers should ask the National Credit Union Association for the amount of overdraft and NSF fees that Self-Help generates. They will find this entity is more harmful to consumers on an economic basis than any payday loan company.

5) “Payday loans of this sort have made the industry extremely profitable”

The authors do not define “extremely profitable”. Profit does not exist in a vacuum. They make no comparison of net profit margins or absolute dollar figures to other industries, nor provide any assessment of revenue collected vs. the risk of default.

The authors’ statement, at best, demonstrates ignorance as to how to read basic financial statements and analyze basic business profitibility metrics.

6) “Consumers that..make little headway on their loan principals generate the vast majority of the payday lending industry’s profit”

False. The authors are cherry-picking phrases of other works to further their own assertions. The very report they cite [Flannery; p.6] says that evidence to support the quoted assertion is, in fact, “quite limited”. Furthermore, Flannery dismisses other studies supporting this claim as lacking proper methodology and conclusions. Flannery also states that 46.9% of store revenues come from frequent borrowers (not rollovers). 46.9% does not qualify as a “majority”.

I further challenge the authors’ claim that the location of payday loan stores proves that lenders “target” low-income and minority neighborhoods.

Flannery disagrees completely with this conclusion in his study [p.13]. Furthermore, Lehman’s critique of the CRL’s “Race Matters” study also applies to the authors’ false conclusion. This quote from Lehman is directly applicable to the authors’ false conclusion, by simply substituting the words “low-income” and “Christian” for “African-American”:

“Based upon their model of correlation and regression analysis, the authors
assert that payday loan firms locate stores in geographic locations with a
high density of African-Americans, arguing that firms “prey” on African-
American customers. However, the authors do not possess data on the
actual customer base of the specific stores used in the analysis, so there is
no way to know whether, for example, these stores have a proportionately
higher or lower number of African-American customers, or whether those
customers even reside in the census tract in which the store is located.
Theoretically, the stores may have a proportionately high or low number
of African-American customers relative to the census tract in which they
are located. And, theoretically, the stores may have a mix of customers
who reside both inside and outside of the census tract in which the store
is located. This is particularly a concern if the stores serve customers who
work in close proximity, or who shop at retail outlets close to the payday
loan store, but who have a residence in another census tract. The data
gathered by the authors of the study leave these questions unanswered.

Although the link between a store’s customer base and its geographic loca-
tion may be plausible, the data used in this study fail to show any direct
connection between location in a census tract and the actual customer
base of the stores themselves. The authors are simply asking readers to
make a “leap of faith” that the makeup of the census tract is necessarily
the same as the makeup of the store’s customer base, and then to infer
from this loose connection that the payday lending firm made a calculat-
ed decision to locate the storefront in this census tract to achieve this
result. In reality, the data used by the authors show no such connection.
As Saltes (2005) accurately points out, “[T]he racial composition of peo-
ple who live nearby a payday loan store does not necessarily imply that the
store’s customers have the same characteristics.”

The authors’ study also contains the same multicollienarity and model specification problems that Lehman discusses in the same paper.

In short, then, the authors’ choice to simply map the locations of payday loan stores do not yield the data necessary for the them to draw the conclusion they have.

Usury and Christianity

The authors’ discussion surrounding usury is also problematic.

They correctly note that the Bible prohibits taking advantage of the poor, in either intent or effect. “Taking advantage” is the qualifier which defines whether or not a loan is usurious or not. Simply put, God asks whether our actions help someone or harm them. If you take advantage of the poor, you harm them.

The flaw in the authors’ reasoning stems from their inherent and demonstrated bias that payday loans harm the consumer. They present this judgment as a postulation – despite little evidence that this is the case. In point of fact, consumer behavior and true rates of on-time repayment demonstrates the product’s value. Although few studies have been conducted with the millions of PDL customers regarding their assessment of the value of the product, the fact that 154 million transactions occurred in 2008 is indicative that, on the basis of cost alone, consumers prefer this choice over other short-term credit options.

Indeed, Dr. Graves has told me he wants payday lenders to be put out of business. Yet the Bible condones commerce. It is not holy, Godly, or biblical to shut businesses down. No Christian theologian or pastor would say it’s improper to loan money unless there is a demonstrable harm. Given the overwhelming support consumers have shown for the product, shutting down PDLs would truly be harmful to them – forcing them to less desirable alternatives.

In addition, the authors’ assertions fail to even provide a modern definition of usury. As this article demonstrates, payday loans do not fit this modern definition. It also challenges the authors’ interpretation of Nehemiah, further rendering suspicion concerning their definition of usury.

We can go one step further. We are reminded that the Scriptures are from the Mosaic Covenant, a specific Covenant God made with a specific people in a specific land, for a specific period — which ended with the New Covenant. Indeed, the First Council of Nicaea forbade clergy from engaging in usury, which was defined as interest of any kind. Anyone living in modern society recognizes that interest must be charged for loans for us to even have commerce, period.

This brings us to the very question the authors’ indicate is the driving force behind their study. “If the Bible so clearly and forcefully condemns usury, one would hypothesize that political jurisdictions with a traditional, conservative Christian perspective would adopt law reflecting this Biblical value.”

The hypothesis that Christian political jurisdictions would adopt laws reflecting Biblical value is false if lawmakers do not consider payday lending to be usury.

Methodology Critique: The “CPI Score”

The authors create a “Christian Power Rankings” using three criteria, of which two are derived from incomplete data and sampling bias.

“The per capita density of Evangelical Christians and Mormons involved trying to rank states based on the simple percentage of people whom we believe are prone to use their Christian Faith to guide them as they vote for public officials.”

“Whom we believe”? How can this be called anything but sampling bias? How can scientific conclusions be obtained when the experimenters randomly and subjectively apply their opinion to something? They even admit that, “obtaining good data based on religious affiliation is not easy”. It renders the criteria valueless, and therefore renders the CPI score, which is derived from this criteria, to be valueless, as well.

In addition, this approach ignores the fact that the composition of state legislatures at the time that payday loan enabling legislation was passed was entirely different than when the authors conducted their study!

For the 2nd criteria for CPI, in which they use the opinions of conservative Christian Groups actively engaged in the political process by using the voter scorecards of three Christian PACs: “Our purposes would have been better suited if they…rated every member of every state legislature rather than just the legislators in D.C. because the state legislators are largely responsible for crafting much of the law that regulates payday lending. As it is, we have chosen to use the scores assigned to federal legislators as a proxy for their counterparts in state government”.

Again, they use an unrepresentative substitute for their criteria. One cannot assign a proxy, such as this, when it dismisses the differences that exist between state and federal party affiliation. Also, legislative decisions made at the state level are subject to infinite political variables that differ from those at the federal level. As it is, with current attempts at federal legislation, the April 2nd hearing of the House Finance Subcommittee, showed broad differences between federal legislators opinions on payday loans and those at the state level.

With two of the three criteria for determining CPI score rendered valueless, it completely discredits the CPI score, and all conclusions derived from it.

Yet the authors claim, “Our hope is that this index….will accurately reflect the political climate of the respective states”.

Hope is all they have to go on at this point, because scientific accuracy is out the window.

Other faulty conclusions:

1) “PDLs have exploded into an industry with more than McDonald’s, BK, Sears, JCPenny, and Target combined. For those concerned about the social moral and spiritual well-being of the lower and moderate income Americans, this is a profound, unprecedented and troubling change”.

Even if one were to stipulate that payday loans are “evil”, the authors’ examples dismiss the vastly more significant social and economic costs that the cited fast food chains contribute to obesity and heart disease. The economic and social harm these products perpetuate (rising health care costs, burden on the health care delivery system), as well as to the individual health of those that use those products irresponsibly, are far more serious than being called by collection agents for a defaulted payday loan.

2) The mapping correlations they present again “fails to fully develop a broad eco-
nomic theory of the many possible determinants of payday loan storefront
locations… This is a major methodological weakness of the study, for it fails to con-
sider a host of potential determinants of storefront location that go well
beyond the demographic composition of the census tract in which the
store is located. Thus, there are likely a number of omitted variables that
may explain storefront location decisions but which were excluded from
the model developed by the authors of the study. Omitted variables lead
to severe weaknesses in regression analysis because they may lead
researchers into drawing spurious relationships that do not exist. “Finding
an apparent relationship in a regression that actually doesn’t exist can be
a consequence of omitted variable bias” (Saltes, 2005). “” [Lehman]

3) “Our data report a simple – but nonetheless important—geographic fact: there tend to be more PDLs in areas where conservative Christians live and control others. We leave it others to explain why this relationship exists”

The data leads to false conclusions, as I’ve already indicated. There is not a shred of proof that this alleged relationship exists.

Second, yet another example of bias occurs because, by the authors’ own claim, 75% of America considers itself Christian. It therefore stands to reason that three out of four of any type of retail store will exist in an area “where Christians live and control others”.

The simple truth is this: Payday loans are used by those in need of short-term credit and earning $25,000 – $50,000. Naturally, one would expect payday loan stores to cluster in neighborhoods with a demographic that is most likely to have need of the product. This is no different than gas stations and fast food chains clustering at freeway exits, and retail businesses clustering in shopping malls. Conversely, one would not expect a BMW dealer in a low-income neighborhood. The cited occurrences have nothing to do with race, ethnicity, or religious affiliation.

The idea that a business would deliberately establish an operational policy driven by any of these considerations is ludicrous in the extreme. Where a payday loan store, or a store of any kind, is located depends on dozens of variables. The idea that practical operational considerations would somehow be trumped by the neighborhood’s ethnicity or religious affiliation shows the inherent flaw in academic studies that do not examine how businesses actually operate. It also shows why grassroots, in-person scientific sampling at the storefronts are necessary, and why geographers and lawyers lack the proper education to engage in studies like this.

4) “Our findings should serve as conclusive proof that conservative Christian Americans are a prime target of PDLs”.

For all the reasons stated above, this is anything but “conclusive proof”. The authors proceed with a clear bias against payday loans. They select data from biased sources in financial competition with payday lenders, cherry-pick quotes from studies that reach divergent conclusions, selectively interpret Scriptural passages to support a flawed definition of usury, base their entire methodology on inaccurate data and sampling bias, and present data subject to omitted variable bias.

Therefore, the conclusions reached A) are grossly lacking in confidence to remotely prove correlation of any kind, much less “conclusive proof”, and B) provide no proof of causation. Instead, the author’s bias against payday loans renders the paper into a screed against payday lenders, with a distasteful inferred swipe at Christian legislators for being “hypocritical”.

With the hypothesis itself of no value, and the methodology and conclusions also demonstrated to be false, the entire paper is rendered meaningless. It renders unfair judgment on payday lenders and Christian politicians. Worst of all, it implies favoritism towards paternalistic government policy.

This is precisely the type of poor research methodology, tainted by bias, that payday loan opponents have utilized for years.

Be Sociable, Share!