Most women end up borrowing from a money lender at very high interest rate. And they even end up borrowing from the money lender when they are about to deliver. When they do this they get the money at a very high interest rates because they need the money right then and there.
The reason we think this is puzzling is this: people borrow from money lenders for health shocks, because after all it’s a shock. But delivery is not a shock, you have known about it for at least five or six months. Yet in those five months, people don’t save up. And you can’t say that they don’t have the money because here is what happens: you deliver, you borrow and then in the next five months you repay.
So if you had the money post delivery to repay, the woman isn’t working then, the woman wasn’t working in the previous five months, so obviously the money is there. It’s just a planning problem. We find many examples of that where people can prepay and save a lot on interest or post-pay and pay a lot of interest. Several health shocks are predictable yet people don’t save up for that. We have kind of worked out a financial product to facilitate this and we are now trying to launch it with the IFMR Trust, which is based in Chennai.
One interesting study that you carried out was on discrimination in job interviews on the basis of race and name. Can you tell us something about that?
Aha! In the US, this is a very big topic. There African-Americans are in a minority and many people feel that they are discriminated against. If you look at wages, you find that African-Americans earn less. But you never know what’s discrimination and what’s just difference in skill. So maybe they earn less because the employer looks at the resume and says well you just haven’t done as good as a job or you don’t have the same education or you don’t have the same skills. So is it the employer who is really discriminating? Or is it that employers are picking the best candidates and it just so happens that we have a system that the best candidates are not African-American?
That’s a hard theory to test. But we found a kind of cute way to test it. We sent out resumes that were made up in response to ‘help wanted’ ads. We just changed the name on the resume be it an African-American name or a white name. Then we set up phone lines, and we see who got called back. So we sent out five thousand fictional applicants, and we found that the African-Americans were much less likely to be called back. And what’s funny there is that the same resume sometimes gets sent out with an African-American name and sometimes with a white name. So we know that the skills were literally the same. But the gap was huge, the same resume got 50% more calls if it was white than if it was black. And that was just astonishing to us. We even redid a version of the study in India with Muslims and Hindus. We are awaiting the results.
You did a study partnering one of the big banks in South Africa and found that psychological cues go a long way influencing borrower behaviour…
We were looking at a bank giving out loans at very high rates to customers who had already borrowed from them. We sent out letters and we randomised what interest rates people got. So some people got a higher rate, some people got a lower rate and from that we could understand what is the demand curve; when interest rates change, how much does demand for the loan change. But at the same time we varied little features of the letter. So there was a photo in the corner that could be a man or a woman. What we were interested in was, let’s see how much do these psychological cues matter.
What were the results?
It turns out that the interest rate does matter. But quantitatively, it is not nearly as important as psychologically. So a photo of a woman on a letter sent to men raises a take-up of a loan by the equivalent of 5% of the interest rate. If I put a female photo in a letter sent to men and raise the interest rate by 5% I’d have have the same take-up as I would have if I were to offer a loan at 5% lower, without the female photo. There are a lot of subtle cues that drive demand a lot more than you might imagine.
You carried out a very interesting study on the entire process of getting a driving licence in Delhi...
What we did was we took people who were about to go and get the driver’s licence. We put them into different groups. One group we just followed. To another, we gave money if they got the licence fast. Between the temporary and the permanent licence, you have to wait 30 days. So we said if you do it in 31 days, we will give you a cash bonus.
What we found is kind of stark. The people in the bonus group —- they had an incredibly higher rate of getting a licence, without ever knowing how to drive. This kind of illustrates a deeper problem that people always forget about corruption. People focus on the bribe. That’s fine. That’s money. But that is not really the problem; what really is the problem is it that regulation gets distorted.
What the study told me was bribe-paying wasn’t the problem. It meant that we were no longer regulating who could drive and who couldn’t. So we were giving licences to people who had no ability to drive. So why even have the regulation to begin with. In our data, literally, the people who can drive had no difference in getting the licence vis a vis people who couldn’t. It turned out that a lot of people used agents. So we then hired a bunch of actors to go to the agents. So some of the actors said to the agent: “Oh I would like a licence, how much does it cost?”
Comments