Lit in Review: Child Undernutrition and Labor

Harttgen, Klasen, and Vollmer, "Economic Growth and Child Undernutrition in Africa," GlobalFood Discussion Paper.

They add to the short but slowly growing literature on the impact of economic growth on child undernutrition that I based my dissertation work on. They add Klasen (08), Friedman and Schady (09), and Subramanyam et al (11) to the ones I knew about.

They identify a micro-macro paradox: increasing household income reduces undernutrition significantly but increasing average incomes doesn't make nearly the same difference. It is rather disturbing to me to see for how many SSA countries there has been significant improvement in GDP/capita, but a worsening of child hunger. Among the things they find that matter more than GDP/capita, household asset rank matters, suggesting that inequality may be an important part of the story.

Bhagowalia, Menon, Quisumbing, and Soundararajan, "What Dimensions of Women's Empowerment  Matter Most for Child Nutrition?' IFPRI Discussion Paper 01192

Women's empowerment can be a fairly vague (their preferred word: complex) term, so it is nice to see the authors breaking various dimensions of it down to see which are correlated with better outcomes. From the abstract: "Results from logit models indicate that both a greater degree of women’s empowerment and greater maternal endowments [read: height and education] are associated with better long-term nutritional status of children [in Bangladesh]. Attitudes toward domestic violence have an effect on child stunting and mobility; participation in decisionmaking is an important influence on dietary diversity. Consistent with previous studies, maternal height and maternal schooling decrease the probability of stunting, and maternal schooling is positively associated with dietary diversity. While these are not immediate measures of empowerment, they are positively associated with child nutritional outcomes and reflect prior investments in women and girls." However, they have almost no control variables in the regressions, so it is uncertain how much is due to these factors directly, and how much is the fact they signal correlation with socio-economic status or other factors.

De Paoli and Mendola, "Does International Migration Increase Child Labor?" Centro Studi Luca d'Agliano Development Studies Working Paper 339.

An interesting data combination allows them to look out how out-migration of workers from different skills impact local labor markets, and how those impacts affect some 200,000 children studied in 38 different countries. They find that allowing low-skilled workers to emigrate increases local wages for the parents, reducing child labor. It might have also been the case that it raised the wage for child labor, which would increase child labor, but this does not seem to be a significant factor. They also see that female emigration reduces child labor by more than male emigration. "According to our estimates, a 10% increase in the migration rate decreases the probability of child labor by 1.2 percentage points (p.p.) and the total time of weekly work by 7.3 hours." I count 209 reported regressions and, at first glance at least, the vast majority tell the same overall story.

Analogy of the day: Immigration and cookies

Wronging Rights has a thought-provoking analogy on immigration. The basic question is whether immigration is a zero-or-negative sum game or a positive sum game. Taub's analogy is that some people think of immigration like love (the more you share, the more you have - a positive sum game) while others think of immigration like a cookie (the more you share, the less everyone else has - a zero sum game). To put in an equation, which side do you think would be bigger:

      Gains to immigrants     OR    losses to recipient country + losses to the families left behind (brain drain)

So the natural answer from an economist is to try to measure the gains and losses. Barro, Card, Clemens, and many others have a slew of literature showing that the gains to immigrants are really big, the losses to host countries are small, and there may even be a brain gain instead of a brain drain. Then we even have Taub and Yglesias arguing the recipient country gains too, so that EVERY term is a gain. Conclusion: immigration is really positive sum. If we're feeling particularly wonkish, we'll even write down a social welfare function to show that if every human deserves equal attention, the gains to poor immigrants are even bigger compared to any conceivable losses to the already-wealthy recipient countries.

Unfortunately, what most people don't tend to do is actually sit down to measure gains and losses and find out empirically whether immigration is like love or a cookie. They use a substitute question, such as how they feel about immigrants, or our burgeoning welfare state, or people getting out of poverty or something else, and from there develop a line of reasoning to justify their emotions*.

This is part of the reason it is so difficult for economists to convince people of the gains from immigration. If people have at heart an assumption that the world is zero-sum because of how they feel about the question, new numbers won't resolve the issue. Put another way, if people assume there is some accounting identity out there that says

      Gains to immigrants = losses to recipient countries + brain drain

then new research showing the gains to immigrants are even larger won't convince anyone. They'll just assume that the losses must be larger too. New research on brain gain won't convince them either because it just means losses to the recipient country are even larger than they thought. Combine that with 1) the fact that Americans think there are twice as many foreigners already and 6-7 times as many illegal immigrants in the coutnry as there are and 2) new research showing that correcting that misperception doesn't change attitudes much, and we have a problem. This makes it all the more amazing that Clemens succeeded in convincing the US government to remove restrictions on Haitians coming to the US.

If the real issue is about how people feel, then what needs to change are hearts rather than minds. The research will provide evidence for them to hang their hats on once they've switched over to the Love camp, but by itself won't bring about the change we need.