04 October 2014

Does FAO really cheat on hunger numbers?

An article by Martín Caparrós in the New York Times has been getting a lot of air time.

The article basically accuses FAO of cooking the hunger numbers from 1990--revising them upwards so that the current numbers make it look like progress is being made (see the State of Food Insecurity in the World, or SOFI, from 2014).

Caparrós stops short of accusing FAO of "conscious corruption". Instead he says the revisions and their direction, are "a symptom of an institutional culture that has to prove it is achieving important progress".

The argument is that it is hard to imagine why the 1990 numbers have increased from 786 million (estimated in 1992) to the current estimates for 1990, 995 million.

Well, as you know, I am no fan of the undernourishment numbers--they are very crude estimates of hunger, but if we want numbers for every country every year, they are the best we have at present.  

But has FAO really been cooking these numbers, even if "unconsciously"?  I say no. 

Lets go back to the regional estimates.  I could not find any SOFI's before 2005 online (something to note FAO - your links go back to 1999 but they do not work). 

Comparing the 1990-92 estimates for the numbers of people undernourished (millions) from the 2005 and 2014 SOFI reports we get:

Developing World:          823.8 (2005), 994 (2014)
East Asia:                         198.8 (2005), 295 (2014)
     >China:                        193.5 (2005), 289 (2014)
SE Asia:                              78.4 (2005), 138 (2014)
LAC:                                    59.5 (2005), 68.5 (2014)
Near East and N Africa:  24.8 (2005),      6 (2014)
Sub Saharan Africa:        170.4 (2005), 176 (2014)

As you can see, most of the increases are from East and South East Asia, and most of those are from three countries: China (up nearly 100 million), Myanmar (from 4 to 27 million) and Indonesia (from 16 to 36 million).  In Sub-Saharan Africa the estimates have barely moved.

So what is going on in  these 3 countries?  Well that would take a lot more research than this, but in 1990 these countries were not exactly transparent with their data and information.  Arguably they are more so now, and as their data become more available and more scrutinised, they probably begin to edge closer to the realities of 1990.

Governments have stronger incentives than FAO to use numbers to their advantage, and I suspect this is what may have happened back in the 1990s.

Did FAO cook the books? I would have said no before seeing these numbers and now I am even more sure.  FAO could do a better job of explaining these revisions, however.  

Is the NYT article poor journalism?   Yes--they should have checked the regional numbers,  or at least gotten a counterpoint view.  Their story corrodes the faith people have in our collective efforts to reduce hunger and in the institutions that are charged with leading the effort.  That's the real story. 


Graham Bell said...


Anonymous said...

A reply by FAO has been published by the NYT. You can read it at http://www.nytimes.com/2014/10/02/opinion/world-hunger-numbers-and-solutions.html?_r=1