The implications of rural-urban migration for children left behind

Back on New Years Day, Tyler Cowen at Marginal Revolution posted this "China fact of the day":

More than 61 million children — about one-fifth of the kids in China — live in villages without their parents. Most are the offspring of peasants who have flocked to cities in one of the largest migrations in human history. For three decades, the migrants’ cheap labor has fueled China’s rise as an economic juggernaut. But the city workers are so squeezed by high costs and long hours that many send their children to live with elderly relatives in the countryside.

The Washington Post article which inspired Tyler's post makes for sobering reading. But was no surprise to me. One of the first things I noticed during my PhD fieldwork in Northeast Thailand in 2003, was the number of children who were being raised by grandparents. Just like Beibei in the article.

Even worse, many of the grandparents that I spoke to related a similar story: The grandchild (or often grandchildren, since unlike China, Thailand didn't have a one child policy) are left with them while the parents go to work in Bangkok. The parents start off by visiting and bringing money with them, then after a couple of years this becomes sending money without visiting, and before long the remittances start to dry up. Eventually, the grandparents are left caring for the grandchildren without assistance from the parents at all. Now, of course this does not happen in all cases. But it was frequent enough of a story for me to refine my data collection on the fly to allow me to also partially investigate the question of whether these children were being made worse off.

The result was this working paper, co-authored with Steven Lim. In the paper, we look at how migration changes Lux's traditional domestic cycle, and how children are faring (in terms of anthropometric measures) in households of different types. We found that, relative to nuclear families (where both of the child's parents are present), children in other family types have significantly lower weight-for-height (with insignificant differences in weight-for-age or height-for-age). The implication is that children living without their parents have worse outcomes. Surprisingly though, the worst outcomes of all were for those children in extended families with both parents present. We put that result down to nutritional resources being spread among more dependents in those households.

Of course, this study was purely cross-sectional so we are limited in what we can say about causality. We couldn't even control for how long each child had been in their 'current' household type, which was a big limitation. I've been meaning to go back to similar research questions using some of the excellent panel data series that are available for developing countries, which might help to resolve the question of whether the type of household really matters.