01 April 2014

Is economic growth helpful for reducing malnutrition? New paper says no. I think it is wrong.

So, I thought it would be good to start off my IFPRI blogging days with a piece about a new paper in Lancet Global Health on the link between economic growth and under 5 malnutrition by Sebastian Vollmer and colleagues.  They use 121 DHS surveys from 36 countries over the time period 1990 to present to conclude:

 "the quantitatively very small to null association seen in our study suggests that the contribution of economic growth to the reduction in early childhood undernutrition in developing countries is very small, if it exists at all."

This is so counter to all the other research out there (some involving me so I do have a vested interest), that it merits a closer look and some colleagues and I have written a letter to the Lancet Global Health to say why we think the study is flawed and draws the wrong conclusions.  We hope they publish it.  After they have made their decision I will share it. 

Many of the arguments we make in the letter are rather dry and technical, but they boil down to which of the graphs below (from the paper) you are trying to fit a line to.  

The first graph shows the association between the levels of stunting and GDP per capita--clearly a strong and downward sloping relationship.   

The second graph shows the relationship between changes in stunting and changes in GDP per capita over 3-5 year periods.   Here there is a much weaker negative relationship (although still significant at 7.3% level). 

The first graph shows the longer term relationship between GDP and stunting, the second one shows the shorter term relationship.  

Why are they so different?  Well, it takes time for GDP per capita to work its way into improved household income and improved health and watsan services.  The regression work focuses solely on the short term effects but does not clearly portray the results as such.  

Economic growth is not sufficient for malnutrition reduction, it may not even be necessary, but it is certainly almost always potentially useful.  Country case studies tend to show the important role economic growth plays in malnutrition reduction (e.g. see this one for Vietnam). 

We know that nutrition specific interventions can only take us so far (see Bhutta Lancet paper 2) and that nutrition sensitive interventions are full of potential, but as yet somewhat unrealised (see Ruel and Alderman Lancet paper 3).  So while for sure we cannot rely solely on economic growth to reduce malnutrition we need economic growth to pick up some of the slack.  We know it won't do it well in the short run.  We know it won't work in every country.  But on average, longer term, it  has to be useful.  That's what the vast majority of the evidence says and I don't see anything in this new paper to convincingly counter that.  

No comments: