Video: Does poverty lower productivity?

Suanna Oh: Thank you. I'm excited to discuss my joint project with Supreet Kaur, Sendhil Mullainathan, and Frank Schilbach. So we've already seen a good overview of recent works that investigated the links between poverty and cognition and decision making. We were motivated by sort of a natural follow-up question. If poor people have scarcity and are more, you know, are suffering from a draining attention, then could this distract them at work, and would this reduce their productivity at work? We took this question to the setting in rural India where a lot of population are financially vulnerable. So we asked this question, how worried are you about your finances, to our sample during – before our experiment started. And we found that 70 percent of the people told us that they're very worried about their finances, and another 18 percent told us that they were quite worried about their finances.

To provide some context, a large population here are casually employed. So during agricultural harvest seasons they can find agricultural jobs, but during the lean season they have difficulty finding jobs. And if they can find jobs, it's usually temporary contract work in manufacturing or construction. So financial struggle is part of their daily life, and yet this is now sort of a default, you know, environment for them. They are always worried about their finances. So our question recruited some of these people into a manufacturing worksite so that we can actually measure their productivity at work. And then what we do is we directly test the relationship between financial constraints and productivity bygiving them a large amount of cash and seeing how relieving their financial constraints affect their productivity. And we – and we try to provide some tests for whether this effect is driven through the channel of attention and cognitive function.

So before I discuss the exact design, let me first discuss our setting. So in the manufacturing worksite, the workers are asked to produce disposable plates. So these are plates made of leaves, as in this picture. And what they do is they stitch the dried leaves together using wooden sticks. And we partner with local contractors so that the training and the quality standards for these products match the market quality, and then we pay piece rates on the leaf plates that they produce. So – and I want to emphasize that we employed them for two weeks, and this sort of seasonal contract work is very common in this area. So this work environment was very real to them. It was sort of a real work setting that translated into real earnings for them.

So now our – let me tell you about the main test. So our main question is do financial constraints affect productivity. And to do this, what we did was we told – we recruited them for two weeks but gave them different payment schedules. So some of the workers would be paid earlier, and the – while the rest of the workers would wait – have to wait until the end of the experiment to realize all of their earnings. So here is a diagram that displays sort of example schedules. On the first day, everyone received training and then received flat payments. This sort of helped them understand that we're going to deliver our payments, and then – and then seal the – seal the contract. And on day five we announced to them what their individual payment schedule will be.

So some of them, as on the top, will be paid on day eight. So they'll receive about a week's earning as one cash drop on day eight. And then they'll receive the remaining money at the end of the experiment, on day 12. For other people, as in the last line, they would know that they would only get their payments at the end of the experiment. So what we are curious about is during the treatment period, so during the period after day eight, do the people who received cash early behave more productively compared to the people who have not received any cash.

And we thought this cash payment should be large enough to relieve their financial constraints. This is because it is so difficult for them to find work during this season that the amount of cash we were giving out were close to almost a month worths of wages for them. And when we found that, at the baseline, 80 percent of our sample at outstanding lines at the time, and a lot of the people who received cash early paid of some of these loans right away, within the two days after they received cash. And of course the people who didn't receive cash were not able to pay off their loans to the same extent.

And the result we find, the main result we find is that the hourly output of the early cash group does increase by five percent more after payment compared to the people who have not received cash. And we found that this effect is statistically significant at five percent level. And in particular, we found that this positive effect of cash payment is concentrated in those who are particularly financially vulnerable.

So we – during the baseline, we asked a number of questions to see – or to measure any variation in their wealth. So although all workers are casual labour workers and are generally poor, there were some variations. So one question we asked was do you own any land, and the other is do you receive any food credit from the government. The third question is can you get 1000 rupees as a loan if there was an emergency. And another question is what kind of – what type of housing do you live in. And then we also combined these answers into an index variable. And as you can see in this graph, for those who are more – relatively poorer, the effect of receiving cash is particularly high. It's close to ten percent effect on hourly output, and – and they’re statistically significant. But for those who are relatively richer, the – the effect of cash is not statistically distinguishable from the control group, who did not receive any cash.

So the question is are these effects coming through from the channel of cognition. To do this, what we focused on is attentional mistakes. So while the task is something that anyone can learn to do, it does require you to pay attention. Otherwise, you may make three types of mistakes. You may make double holes, meaning that you're putting a stick in and then taking out because it's a mistake. Or you may use too many leaves or too many stitches, meaning that you're doing unnecessarily work, which will slow you down. And we find that early cash group do make fewer mistakes during – after they receive cash, compared to the control group. And again, these effects are driven by those who are particularly poorer in our sample. So this effect of mistakes between the two groups is also statistically significant, at five percent level.

So the last way that we tried to see whether this effect is coming through cognition is by looking at how they behave when we induce focus on finances through salience exercises. What I mean is we tried to bring their attention to their financial troubles and tried to see what kind of effect this has. We we did this by telling them a story about a poor farmer, much like themselves, and then also asking them personal questions about their own loans and financial troubles.

And conceptually, this could have two competing effects. It could – because they're thinking about their financial issues, it may motivate them to work extra hard because, when they work harder, they can take more money home. But if the – the scarcity effect exists, it could also distract them and then make them more prone to mistakes or lower their productivity. Our idea was that if we ease their financial constraints by giving them cash, the negative cognition effect may be reduced, meaning that the people who have received cash may – may have a more positive effect compared to those who haven't received cash when we induced this – when – induced concerns about finances through this exercise.

And so what we did here, after we implemented these financial salience discussions, we told them suppose you needed to come up with a large amount of money for an emergency. How would you obtain the funds? Please think about this while you're working, and then at the end of day we'll ask you about, you know, how you are going to cope with this emergency. So this was to actively engage them about thinking about emergency situations.

And what we found is that after – or, you know, during the period where people are actively thinking about their emergency, there is an over a positive effect on output with increases in mistakes. So it seems that the motivation effect may be dominating in this setting. However, this positive effect was concentrated only in the post-cash period. So among those who have already received cash, talking to them about their financial issues actually helped them work better. However, if they have not received any cash and were already very concerned about their finances, talking to them about their finances didn't help them produce more. And while the attentional outcomes regarding the mistakes were noisy and suggestive, they did go in the same direction.

to conclude, we found some effect that relieving financial constraints have a positive impact on people's productivity. And then we show – we hopefully show some compelling evidence that this could be operating through cognition. Thank you.

(Applause)