Fernald – Mexico cash transfer program

I have spent the last 10-15 years involved in the evolution of Mexico’s large scale cash transfer program, and was involved in the follow up, the three to five year follow up and the seven to ten year follow up of child development outcomes from a large randomized control trial that began in 1997. 

So the most amazing thing about this study is that it was randomly assigned at the community level in 1997 which is rare in a kind of a large social study like this, and so we were able to follow children who had been enrolled in the program, they’re now probably 15 years old and we followed them at two different time points. So there are two key findings that came out. The first is there were effects on reducing behaviour problems in the children who had been enrolled longer in the program than those who hadn’t been. So the main effect of reducing behaviour problems was clear at three to five years and also at seven to ten years, that was a very clear and significant finding for both of those time periods. Those were looking at the program effect. The second thing we really wanted to do was look at a dose response relationship to the cash the households were receiving. So for example a household that received, this was now looking at children who had been enrolled in their program their entire lives, this is looking then at how much cash the household is receiving, and then looking at how the children develop and what we found was that they households that received more cash as part of the program, controlling for all of the baseline variables, and all of the housing, and all of the household and parental variable etcetera, the household that had received more cash, those children were the ones who performed better in terms of cognition, in terms of language, in terms of having fewer behaviour problems. 

So it was a finding showing really that among children who had been enrolled in a program for their entire lives, there’s really a clear dose response relationship, and improvement in cognitive and language outcomes for those children who had received more cash, who had been in households that had received more cash.  

There’s a very high take up rate so I should mention that the Mexican cash transfer program increases income by 20 per cent, so it’s one of the biggest increases. For example I just wrote a paper about the Ecuador conditional cash transfer program and that one only increases income by seven per cent. So there’s a very big range in terms of how much income households receive and the Mexican one is at the higher end of the range, so the compliance rate is quite high, in Mexico it’s reported to be 97 per cent compliance in all groups, so this allowed us to be able to do this dose response analysis without worrying about a significant part of the population dropping out.