Thursday, November 20, 2014

Does Chinese aid mainly go to African presidents' villages?

A new paper released by the folks at AidData argues that Chinese aid goes disproportionately to the birthplace regions of African leaders. Media comment is building up on this paper, and an article in The Guardian seems to be leading the way. I am actually somewhat sympathetic to the argument of the paper. My own anecdotal observations suggest that at least one of the 3 primary schools financed in many African countries by the Chinese (following a 2006 FOCAC pledge to build 100 primary schools across Africa) has often been located in an African president's hometown. However, the vast majority of Chinese official aid projects are financed in capital cities -- something the researchers neglected to discuss.

Unfortunately, I think the Guardian article got a few things wrong.

1.  The Guardian article links to accurate official statement that the Chinese spend more than half of their foreign aid in Africa but then disregards the official published Chinese statistics on aggregate official aid to Africa (which would be a total of about $26 billion up to 2012*) in favor of the overblown figure circulated by AidData ($80 billion between 2000 and 2012). The article strongly implies that the AidData figure is an official figure. It is not. It purports to include all finance from China (real and "under discussion" including commercial loans, export credits, and other trade-related finance). Further, despite some improvements, AidData still gets a lot of these figures wrong, as our SAIS-CARI spot check of their hydropower project data revealed. They over count, by a big margin.

2. This misunderstanding continues in this sentence: "Researchers took data that China published on its foreign assistance," and geo-coded it. Whoa, Nelly! China did not publish that data, it was AidData that collected it, mainly from media reports.

3. The Guardian article also said that African leaders are three times more likely to spend Chinese aid in areas where they have ethnic ties. The paper in question actually said that they found no evidence for this. Their purported finding was that all things being equal, Chinese finance is more likely to go to the region of birthplace of an African leader than to another given region.

Readers of this blog know that I have been critical of AidData in the past. This summer our team looked again at their data and found that it had been improved -- although they still have a large number of "false positives" (projects that AidData says have been promised finance, but where our investigations show that these projects were wishful thinking on the part of an African government or a Chinese contractor: many never secured finance). But -- perhaps because of their name -- they cannot stop describing their numbers as "Chinese aid" even when their own discussions suggest that much of what they are tracking is trade and investment finance, for generating business.

------
* In the 2011 White Paper, the Chinese total for all aid provided globally from the start of the aid program up to 2009 was $37.7 billion with the 2014 White Paper announcing another $14.4 billion between 2009 and 2012. Assuming the percentage going to Africa was about half, this would have been $26 billion.

16 comments:

Carlos Oya said...

Deborah,

I think you should write to the Guardian and alert them about these gross mistakes. Many people read this newspaper and the paper ought to correct and I'm sure they will do if you write to them and tell them what they got wrong. they might even invite to write a brief comment on the article and the news. I really suggest you do this.

Deborah Brautigam said...

OK Carlos, I've just sent my "letter to the editor" of the Guardian. Let's see what they do with it!

Anonymous said...

Hi Deborah,

I'm the Guardian journalist who wrote this article. We have revised the article with the following correction: This article was amended on 21 November 2014 to clarify that the $80bn figure for aid to Africa between 2000 and 2012 was an estimate by AidData, not an official Chinese government figure, and that the estimate includes “pledged, initiated, and completed projects”.

As to your other criticisms of my report I'd say that I feel Roland Hodler's comments about AidData's research indicates that there is a strong bias among African leaders to spend Chinese development aid on areas where they have ethnic ties: “As soon as [a region] becomes the birthplace of an African president this region gets 270% more development assistance (from China) than it would get if it were not the birth region of the president."

Please feel free to get in touch with any further questions or concerns.

Best,

Mark Anderson
mark [dot] anderson [at] theguardian [dot] com

Deborah Brautigam said...

Hi Mark -- Thanks for correcting the figures that implied that the $80 billion was an official Chinese figure. On the other point, the authors make this pretty clear in the paper. Here's a quote from the abstract: "We do not find evidence that leaders shift aid to regions populated by groups who share their ethnicity." I'm not sure why this would be the case, as the theory behind this would suggest that political survival strategies would dictate a broader allocation of patronage. Perhaps there is something strange about their coding of projects.

Anonymous said...

Hi Deborah,

Do you refer this particular primary school?

http://tz.china-embassy.org/eng/ztgx/t787932.htm

Unknown said...

Another example of getting warped by correlations being interpreted as something causally significant. Daniel Posner published a good piece on why public works (by extension aid) patronage type of inferences can be highly misleading.

Deborah Brautigam said...

Hi Dwayne -- do you know which piece that was? I'd like to read it. Sounds quite useful.

Deborah Brautigam said...

Absolutely. I talk about this school and its location in the hometown of Tanzania's president in a working paper published by IFPRI in 2012: http://www.ifpri.org/sites/default/files/publications/ifpridp01214.pdf

Carlos Oya said...

Dear Mark (Guardian journalist),

Why don't you try to find a study that locates schools built by OECD aid in the past and whether they can be linked to 'African politicians'. There is nothing new in this story about 'aid'. Anyone remember Mobutu and the aid channelled to his tiny village? and so many other stories, some told and retold by the likes of Calderisi and Easterly?

Basically, much of this 'new' empirical evidence has little 'new' and is based on suspect raw data easy to misinterpret. Guardian readers should always be reminded the old computer adage: Garbage In = Garbage Out.

Yours

Kai Xue said...

From 1964-1976 China provided North Vietnam and then Vietnam with about 20 billion USD of military and economic assistance. If that figure is included in the total aid given up until 2009 then aid sent to Africa was less than half of the total for the period from beginning to 2009.

Deborah Brautigam said...

Hi Kai Xue,
This is an interesting question -- although China's annual budget for external assistance includes military aid (as does the foreign assistance budget of the US), I am not sure that the historic aid totals reported in the two White Papers include military aid. The first White Paper discussion suggests not, as it talks specifically about things like "civil airplanes", but I will see if my contacts in Beijing know the answer to that.

Unknown said...

We were misunderstood - DIVING INTO THE DETAILS: WHAT “AID ON DEMAND” SAYS AND DOESN’T SAY ABOUT CHINA IN AFRICA

Unknown said...

We were misunderstood: DIVING INTO THE DETAILS: WHAT “AID ON DEMAND” SAYS AND DOESN’T SAY ABOUT CHINA IN AFRICA

Deborah Brautigam said...

Here's the link to AidData's response: http://aiddata.org/blog/diving-into-the-details-what-aid-on-demand-says-and-doesnt-say-about-china-in-africa

Anonymous said...

Deborah,

That "response" from AidData contains misleading facts and most important of all, it chooses to not answer what Carlos Oya has been asking --- about the proportionality of Western aids which flowed into the birth region of the African leaders

AidData's contention that Chinese aid pattern is flawed, but in the same breath, their silence on the similar pattern emitting from the Western aid is tremendously deafening indeed

Anonymous said...

Dear Kai Xue,

The most valuable "foreign aid" China has given to Vietnam (and perhaps the whole world) is a drug called "ARTEMISININ"

It is an anti-malaria drug, developed by China, to aid their North Vietnamese comrades during the Vietnam War

Many North Vietnam fighters were fighting in malaria infested dense jungles and wetlands. Many came down with the disease and that had severely hampered the North Vietnam's effort into launching attacks on the puppet regime on the South, hence the North Vietnamese requested the help from China ... and the rest, is history

The drug "ARTEMISININ" is still in used today. Many millions of people (not only the North Vietnamese) are still alive thanks to it, and to China