17 years in a refugee camp: on the trail of a dodgy statistic

Photos of refugee camps always look like this: the subject of a future post

Seventeen years: the average length of stay in a refugee camp. The figure came up in a conversation last week with a French journalist who’s making a documentary about camps, but I’d heard it many times before—perhaps you have, too. Here are a selection of international sources quoting this figure, which I turned up with a bit of cursory googling in English and an even briefer bit in French:

  • A 2013 article on TakePart.com (“a digital news & lifestyle magazine and social action platform for the conscious consumer”), quoting Liesl Spitz of FilmAid.
  • From 2012, the Khalid Hosseini Foundation supporting a UNHCR campaign.
  • A 2008 Australian Broadcasting Corporation (Hobart) news article about an Oxfam educational resource.
  • From the same year, an edited volume in one of the best academic series on refugee studies, Berghahn’s Studies in Forced Migration: Hajdukowski-Ahmed et al., Not born a refugee woman (see p16).
  • The info page for the Ideas Box, a Libraries Without Borders initiative (and toolkit) for providing education in refugee camps.
  • A Radio France news article about the above, including an interview with the designer, Philippe Starck (yup, that Philippe Starck).

You find the same figure being quoted all over the world, by all sorts of different sources, and not just in English and French: in this Turkish news article from just two weeks ago, ‘Angelina Jolie stated that the average period for which refugees live in refugee camps was 17 years’ [Angelina Jolie, mülteci kampında mültecilerin yaşadıkları sürenin ortalama 17 yıl olduğunu belirtti]. Any statistic that’s so freely quoted by so many different people deserves a bit of critical attention, especially when no-one ever gives the source. And as it happens, my friend Eleanor Davey and I did a bit of digging around about this one last year.

Where does the figure come from? It appears in the 2006 edition of the UNHCR publication, The State of the World’s Refugees, which is probably how it began circulating more widely: lots of journalists, think-tankers, academics, and NGO media teams read this report and quoted from it. If you look on the UNHCR website you can read it too: you’ll find the reference to ‘seventeen years’ in chapter 5, ‘Protracted refugee situations: the search for practical solutions’, p109:

It is estimated that the average duration of major refugee situations, protracted or not, has increased: from 9 years in 1993 to 17 years in 2003.

But the authors are quoting directly from another source: an internal UNHCR document from 2004. If you read that document—and please do, it’s only twelve pages long—you’ll see that it is actually quite cautious about the figures it gives.

Two striking things about the quote itself:

  • It doesn’t refer to camps. At all.
  • It states the situation in 2003.

Those two things already raise big problems for our commonplace figure of 17 years being the ‘average length of stay in a refugee camp’.

First, in most refugee situations, protracted or not, many refugees don’t end up in camps—indeed, they often actively avoid them. Of the 200,000 or so Guatemalan refugees who entered Mexico in the early 1980s, just under 50,000 were settled in camps and given refugee status: ‘the others—an estimated 150,000—were defined as undocumented economic migrants subject to deportation’.* If you go to Jordan or Turkey today, you won’t just find Syrian refugees in the enormous camps near the border: you’ll find them in Amman, Istanbul, and other cities. Even if the average duration of a protracted refugee situation is seventeen years (and it isn’t, as we’ll see in a moment), that doesn’t say anything about the average length of stay in a refugee camp. Some people who should know better blur this distinction without quite ignoring it: here’s one example, and here’s another.

Second, things have changed a lot since 2003. The quote itself estimates that in the ten years since 1993 the average length of a protracted refugee situation had increased from nine years to seventeen: significant and rapid change. What has happened since then? Tina Rosenberg uses the same information in this New York Times blog from 2011, but she doesn’t conflate ‘protracted refugee situation’ and ‘refugee camp’; she also takes care to give the dates to which the information actually applies. I wish Angelina Jolie and Philippe Starck had been as careful.

It gets worse. Read the 2004 document and you’ll see that the figure is very clearly presented as an estimate. (Some of those recent reports state it as ‘fact’.) It’s also based on what UNHCR itself called a ‘crude measure’: ‘25,000 persons or more who have been in exile for five or more years in developing countries’. Crude is the word. I can see the point of having a working definition by length of time displaced: you need some idea of what ‘protracted’ means, with a necessarily arbitrary threshold. I’m not sure why there’s also a limit by size of displaced population, though, nor why this definition is shaped by an even more arbitrary decision which ensures that ‘protracted refugee situations’ can’t happen in rich countries. The document doesn’t say how countries are classified as ‘developing’ or not, either. (This definition now appears to be obsolete.)

The document also notes that Palestinian refugees aren’t included. In institutional UN terms, that makes sense: this is a UNHCR document, and Palestinian refugees have never been UNHCR’s responsibility. As the document says, they’re covered by the United Nations Relief and Works Agency for Palestine refugees (UNRWA), set up in 1948—before the 1951 Refugee Convention, and before UNHCR was created. But any figure for ‘protracted refugee situations’ (or for that matter prolonged stays in refugee camps) that doesn’t include them is clearly meaningless: even other UNHCR documents from around the same time pointed out that they represented the ‘oldest and largest’ protracted refugee situation.

The point of this post isn’t to downplay the seriousness of protracted refugee situations, whether the people caught up in them are encamped or not. Nor is it driven by a desire for accuracy. (Well, okay, a little. “Factual Inaccuracy Found On Internet!”) The problem with this untrustworthy figure—bandied around endlessly despite being deeply problematic, over a decade out of date, and not actually about refugee camps—is what it does to refugees themselves. Like the image of ‘the refugee’, like the identikit photos of refugee camps (subject of a future post, as mentioned in the caption to the one at the top of the page), this figure reduces the enormous complexity of protracted refugee situations, and the much greater complexity, richness, and difficulty of the lives of the people living through them, to a mere cipher. It gives those of us who aren’t refugees a sorry little fact to wring our hands over—one of those things we can feel good about feeling bad about—then set aside. When people use it, they’re not really talking about refugees at all.

Screen Shot 2015-07-03 at 22.58.17
Wikipedia is more careful than almost all of the sources listed above

*Kristi Ann Stølen, ‘Contradictory notions of the state: returned refugees in Guatemala’, in Christian Krohn-Hansen and Knut G. Nustad, State formation: anthropological perspectives (London: Pluto Press, 2005), pp142–167, this quote at p151 and the figure for encamped refugees—46,000—at p153.

Click images for source.