Everybody seems to hate payday loans, but millions of people choose them voluntarily each year. So do we know as much about payday loans as we think?

A recent "Liberty Street Economics" blog post by myself and three other authors summarizes three sets of peer-reviewed research findings on payday loans, with links to all the relevant studies. Despite all the opinions about payday loans, commentators are not always armed with the facts. So this type of research is crucial.

What does the research tell us? First, while payday loans are indeed expensive, that does not necessarily mean big returns for lenders. The typical brick-and-mortar payday lender charges $15 per each $100 borrowed every two weeks, implying an annual percentage interest rate of 391%. But on the flip side, research shows that payday lenders earn no more than competitive profits.

At a 391% APR, how can payday lenders just be breaking even? First, these loans default frequently, so the stratospheric APRs are only expected rates, not actual rates. And the loan amounts are very small compared to loans made by banks, so in some cases the high APR is just enough to recover overhead.

Payday lenders could theoretically charge even higher rates to improve their returns. But with there being more payday loan stores in the U.S. than Starbucks coffee shops, competition is intense and actually holds down prices, resulting in risk-adjusted earnings at payday lenders that are comparable to those at other financial firms.

Second, despite the concerns about consumer protection issues with payday loans, the evidence is mixed on whether those concerns are valid.

A handful of peer-reviewed journal articles test whether access to payday loans helps or harms consumers. On the harm side, studies find that access to payday loans leads to more difficulty paying bills, more involuntary bank account closures and reduced military preparedness by "airmen" who had payday lending troubles. On the help side, studies find that payday loan access is associated with less difficulty paying bills, fewer bounced checks and reduced foreclosure rates after natural disasters. Two studies find neutral results.

Why might consumers be drawn to payday lenders if the product was hurting them? One has to consider the alternatives. If multiple checking account overdrafts are more expensive than taking out a single payday loan — and this can easily be the case — then a payday loan is a rational choice.

The third main area addressed in the body of research is the important issue of payday loan "rollovers," which can be very costly. Typically, if a $100 payday loan were rolled over, the lender would charge an additional $15 for every rollover. About half of initial payday loans are repaid within a month so most of these borrowers pay the fee just once or twice. But about 20% of new payday loans are rolled over six times or more. These consumers end up paying more in fees than the amount originally borrowed. Are these borrowers overly optimistic about their ability to quickly pay back a loan? Again, the evidence is mixed.

One study finds that counseling prospective payday loan borrowers about the cost of rollovers reduced their demand for the product by 11%. A second study finds that 61% of payday borrowers were able to predict within two weeks how long it would take them to pay back their loans, with the rest divided equally between those who over-predicted and those who under-predicted. A third finding by an expert reviewing the available evidence concluded that the link between over-optimism and rollovers "is tenuous at best."

Despite the evidence being mixed, the Consumer Financial Protection Bureau is proposing new far-reaching rules for payday lenders. Lenders would be required to engage in costly underwriting to assess borrowers’ ability to pay. Borrowers would be limited to at most two rollovers for each payday loan, after which the loan would be converted to a term loan at a lower or zero interest rate.

These regulations may simply drive payday lenders out of business, mirroring the experience in states that capped payday loan APRs at 36%. Low-rollover borrowers would be worse off. High-rollover borrowers may or may not be better off, depending on whether they can find alternative forms of credit and how much that credit costs.

My colleagues and I believe that more research should precede wholesale reforms. One area to focus future studies is to determine how many loan rollovers translate into the product being used irresponsibly. If a payday loan is being overused, converting a borrower to a longer-term loan seems prudent and responsible. But how many rollovers is too much?

Existing research suggests that two rollovers are likely too few to identify the truly overly optimistic borrowers. Additional studies are warranted, in part since some states cap the number of allowed payday loan rollovers while they are unlimited in other states. Careful analysis of how borrowers fared in these two sets of states would help inform the regulators.

Whenever possible, financial regulation should be grounded in the results of objective peer-reviewed research, not based on the "analysis" provided by industry or activist groups.

Robert DeYoung is Capitol Federal Distinguished Professor of Finance at the University of Kansas. He has no affiliations with the payday lending industry.