Dean Karlan on Commitment Savings Research: Video Transcript
Hi. We all have weaknesses. Sometimes it’s exercise. Sometimes it’s eating. Sometimes it’s spending money on frivolous things. This is true whether we are rich or poor, fat or slim. The challenge we all face in life is how do we make these trade-offs. How do we trade off the things we say we want to do and the things we actually do. How do we, for instance, set a plan and then fulfill that plan?
When we think about this in the savings space, it’s a very important issue, particularly for the poor, because there are a lot of things that tempt us on a daily basis. If we want to be able to accumulate savings so that we can have a nest egg for a rainy day or if we live in an agricultural setting and we want to have a nest egg so that when it’s planting season we have enough money to buy fertilizer to invest in a farm – how do we manage to get money away from ourselves and safe, in a place and committed, so that it doesn’t get spent, so that it can be used at the time of need?
This is the challenge that the poor face around the world because they don’t have the luxury, for instance, of a certificate of deposit or a retirement savings account, which are very natural commitment savings accounts that we have access to in the developed world. So, one of the things we tested in the Philippines was whether giving people a way of tying up their money could actually lead to real changes in behavior, real changes in outcomes for the household.
So, in the Philippines we gave people an account, we called it a “Seed Account” which did nothing other than lock their money up – it locked their money up until they reached their goal. People were asked, “Do you have a goal?” If their goal was time-based then they were told fine we are going to give you a separate account. You can keep your money in there until you reach that date. If their goal was an amount of money – I want a $150 because that’s what I need for cement in order to finish the floor of my house – then they would be given a goal that was an amount. They couldn’t withdraw the money until they reach that amount. We set this up as an RCT. We found that in the first year those who were offered this account saved 80% more than they would have otherwise, which is determined by understanding the changes that took place in the control group. The most important result was that the women who received this account actually ended up, at the end of the year, with more power in the household. We saw more household durable goods purchased when women got these accounts. This was very striking result to understand how not just access to financial services, but access to this particular type of financial service, could increase power for women in the household.
That was the first start and there are lots of questions one could ask as to how exactly to make these work and you can easily imagine in different cultures, different outside market environments that there are different answers on ways these products need to be rolled out. The two most important things that I think we need to understand more about – there is new research that is forging ahead in this area – is how do people become self-aware. How do they become self-aware enough to recognize that locking money away from their future selves can help them reach their goals?
The second big struggle that I think we have around how to design these accounts is deciding how strict to make the commitment. The basic concept behind any sort of commitment device is about increasing the price of vice. The question is how high do you increase that price of vice. If the price of vice is too high people won’t engage in the activity at all in the first place. They won’t open the account. If it is a commitment contract to exercise more they won’t ever sign the contract. The price of vice has to be higher than it would be normally but not high enough to scare people away from engaging in that activity. But it can’t be so low that it doesn’t have any bite. So, how do you strike that balance.
In Uganda, we did a study where we gave people commitment savings accounts. It was a school setting with children and we used two different options. In one, money was locked in for educational expenses and, in the other, the money was not locked in for educational expenses but it was going to be paid out in cash. But paid out in cash at the time when it was convenient for them to buy educational supplies because there was an educational fair going on in the area where they could use money and buy their workbooks and their pencils. People to a large extent did not use the vouchers for educational expenses instead they used the other one. We actually saw test scores go up as a result of getting more workbooks and things of this nature. It was a striking result showing how a soft commitment works better than a hard commitment but it also worked better than nothing. It was a big positive impact. The minute the commitment is pushed too high then those benefits were not generated. This is not to say that soft commitment is the answer but it does suggest that there is a trade-off that we face in designing these types of products and we have to see how to strike that balance right. So one of the things that is obvious about this space is that as with any evaluation and product there is no one size fits all formula for designing such products. We can’t do one study at one place and then have the answer for the entire world.
We have seen some extremely interesting pilots that show the power of these accounts. Besides the one in the Philippines, the other in Kenya – that was really striking in showing that a commitment device to farmers at harvest season to invest in fertilizers at planting season was hugely successful. There is a commitment saving account in Malawi that was tested that found, similarly, giving people commitment savings account led also to higher investment in fertilizer. What’s striking there is that the money flowed through the liquid account which really suggests that there is something about soft commitment – about teaching people about their own self- control and how to overcome their cash management problems on their own using a bank account can lead to positive benefits.
We have also tried the weakest thing we could think of to do in Ghana, where we put no rules on the account, all we did was give it a label. Now you’re helping him. So, it’s still a commitment savings account but the only cost to deviating is psychological – it’s internal, its dissatisfaction with your own behavior and this too led to a 30% increase. So, we are trying to figure out how low we can go in terms of weakening what the commitment is in order to give people as much flexibility as they need but still guide them in the direction you want. From there we are actually going in the field with a new study that’s even lower where we are just giving people two things – rather than one account they get two accounts and there is not even a label on them. We want
to see how this partitioning of money alone can lead people towards achieving their goals in savings and actually changes the way they allocate their money and allocate their expenditures in the households. So, it’s a lot of exciting work. But a lot of it is at the pilot phase where we are still seeing a lot of micro-tests, which are really necessary to understand the household and individual decision making process and how these things interact with different dynamics when men and women have different levels of power in the household.
Whenever we think of public policy we should always ask, what’s the market failure? Why are we talking about the issue in the first place? Why aren’t we just letting the markets work, firms offer things and people buy them and everyone is happy?
What are the markets failures here? There are two. One could be a behavioral failure. Which is really another way of saying there is a principal agent problem between me and my future self or the person I want to be and the person I am. So I need things that help me be the person I want to be. So, if it’s about temptation, what it means is that I’m trying to squirrel money away from my future self. I don’t trust my future self. I’d like to be able to write a contract with my future self and commit that future self to doing better things for me. So a commitment device is just that. It’s a way of solving that internal principal agent problem.
The other set of market failures come from the household. Maybe I’m squirrelling money away to keep it from my husband. It is another way of saying that I have a principal agent problem within the household. I want to commit to a certain thing. I‘d like the household to engage in a certain behavior. I don’t have total control over the resources nor my husband and if I let the money stay in the household then I lose control over it.
Both of these are market failures, effectively. One within myself and one within the household. And in both settings a commitment saving device is a vehicle for helping me solve that problem and attain the goal that I have in life in terms of how I am going to use financial services to save up money for a rainy day, to save up money for investments for my children, whatever the case may be. Hopefully through a sequence of these pilots we are really starting to get clarity on what macro-level policies can be for large banks and large financial institutions so they can roll out products that are both profitable for the banks and at the same time really helping people achieve their goals and really using financial services to help achieve those goals. Thank you.
Dean Karlan on Commitment Savings Research: Video Transcript