Secret Data Sunday – International Rescue Committee Edition

I haven’t posted for a while on this subject so here’s some background.

The International Rescue Committee (IRC) did a series of surveys in the Democratic Republic of Congo (DRC).  The final installment summed up the IRC findings as follows:

Based on the results of the five IRC studies, we now estimate that 5.4 million excess deaths have occurred between August 1998 and April 2007. An estimated 2.1 million of those deaths have occurred since the formal end of war in 2002.

The IRC’s estimate of 5.4 million excess deaths received massive publicity, some of it critical, but journalists and scholars have mostly taken the IRC claim at face value.  The IRC work had substantial methodological flaws that were exposed in detail in the Human Security Report and you should definitely have a look if you haven’t seen this critique. But I won’t rehash all these issues in the present blog post.  Instead, I will just discuss data.

One of the main clouds hanging over the IRC work is the fact that three other surveys find child mortality rates to be steadily falling during the period when the IRC claims there was a massive spike in these rates.  (See this post and this post for more information.)  In particular, there are two DHS surveys and a MICS survey that strongly contradict the IRC claims.

And guess what?

The DHS and MICS data are publicly available but the IRC hides its data.

As always, I don’t draw the conclusion of data hiding lightly but, rather, I’ve tried pretty hard to persuade the relevant actors to come clean.

Frankly, I don’t think I’m under any obligation to make all these efforts.  I haven’t sent any emails to the DHS or MICS people because there’s no need to bother, given that their data are free for the taking.  But the IRC hasn’t posted their data so I resorted to emails.

I wrote multiple times over many months with no success to Ben Coghlan of the Burnet Institute in Australia.  He led the last two rounds of the IRC research, including an academic publication in the Lancet, so he was a sensible starting point.

In the end, it would have been better if Coghlan had just done a Taleb and told me to “fuck off” straight away rather than stringing me along.  First he asked what I wanted to do with the data.  I feel that this is not an appropriate questions since data access shouldn’t really depend plans.  But I told him that I wanted to get to the bottom of why the IRC data were so inconsistent with the other data.  After prompting, he said he needed to delay because he was just finishing his PhD.  I made the obvious reply, pointing out that even while completing a PhD he should still be able to spare ten minutes to send a dataset.  On my next prompt he replied by asking me, rather disingenuously I thought,  how my project was getting on.  I replied that I hadn’t been able to get out of the starting block because he hadn’t sent me any data.  I gave up after two more prompts.

Next I tried Jeannie Annan, the Senior Director of Research and Evaluation at the IRC.  She replied that she didn’t have the data and that I should try …..Ben Coghlan and Les Roberts who led the early rounds of the surveys.

I knew that Les Roberts would never cough up the data (too long a story for this blog post) but wrote him anyway.  He didn’t reply.

I wrote back to Jeannie Annan saying that both Coghlan and Roberts were uncooperative but that, ultimately, this is IRC work and that the IRC needs to take responsibility for it. In my view:

  1. The IRC should have the data if they stand behind their work
  2. If the IRC doesn’t have the data then they should insist that Roberts and Coghlan hand it over.
  3. If Roberts and Coghlan refuse to provide them with the data then the IRC should retract the work.

She didn’t reply.

Here’s where this unfortunate situation stands.

The IRC estimate of 5.4 million excess deaths in the DRC exerts a big influence on the conflict field and on the perceptions of the general public.  It is widely, but erroneously, believed that this DRC conflict has been the deadliest since World War 2.  The IRC estimate survives largely as conventional wisdom, despite the critique of the Human Security Report.

The IRC and the academics involved keep their data well hidden,  choking off further discussion.

PS – Note that this is not only a tale of an NGO that doesn’t uphold scientific standards – there are also academics involved.  I say this because last week at least one person commented that, although Taleb’s behavior is appalling, he’s not really an academic.



One thought on “Secret Data Sunday – International Rescue Committee Edition

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s