Internal migration as a strategy for dealing with economic shocks: the case of Indonesia

L. Alan Winters

I was recently asked ‘what existing notion or belief do you think we need to shake off in order to foster a more sensible discussion of migration?’ The question was about international migration, but my suggestion was equally applicable to internal migration – migration from one part of a country to another. It was to inculcate the idea that migration is an entirely normal piece of human behaviour. It is not special or abnormal, but since the beginning of time people have moved to make the most of their circumstances. This is not to say that migration is comfortable or even desirable, but that it is – or at least is expected to be – better than the alternative. For all the discomfort of leaving your home and family and often living in very cramped and uncertain circumstances, migrants continue to make the move because it seems to offer better prospects for them and/or their families than staying put. A very rough estimate is that at least 700 million people (over ten percent of the world population) are internal migrants. One can hardly pretend that this is not a major phenomenon or that it should not be the subject of a good deal of research and analysis.

Having recently become Chief Executive Officer of the Migrating Out of Poverty Research Programme Consortium, based in the University of Sussex, I have been thinking quite a lot about internal migration recently. Internal migration is a great deal cheaper than international migration and so is much more accessible to poor people in developing countries. As a result, it tends more immediately to boost the incomes of poor people directly and of their families back home if they remit money (as most do) or return home with new capital, connections and capacities for economic activity.

 

An interesting example of internal migration was the way in which people in Indonesia migrated in response to the Asian financial crisis which devastated the Indonesian economy in 1998. Duncan Thomas (of Duke University) and his colleagues documented the considerable extent of migration during the crisis, much of it from urban to rural areas as the latter suffered less badly than the former. Matteo Sandi, one of my research students in Sussex, and I have been looking at whether this was a successful response to the crisis, especially over the long run (up to 2007). Using data from the Indonesian Family Life Survey, which allows us to observe the same people in 1997, 2000 and 2007 (and often before as well), we asked how the consumption of people who were affected by migration over the crisis grew compared with those who were not. We were able not only to look at the migrants and the households they left, but also at the households they joined as well as cases where the whole household moved.

 

Our preliminary results – which will be written up quite soon – suggest that in the long run, households that lost people experienced faster per capita consumption growth than non-migration affected households in both the short and long runs (from 1997 to 2000 and 2007 respectively). That is, migration seems to have alleviated pressure on family consumption. Households that were net receivers of people suffered reductions in per capita consumption relative to non-migrant ones in the short run but no significant adverse effect in the long run. This suggests that receiving households had to share their resources over more individuals at first, but that over the subsequent seven years they were able to adjust to get back to the ‘average’ trajectory. There is at least some evidence that such receiving households set off in 1997 with higher consumption than the people-exporting households, so this pattern was not necessarily inequitable as some people have suggested. We also look at households which experienced in and out migration of equal magnitude – i.e. which kept the same composition but with (some) different individuals: they showed significant gains in the long run and no significant difference from non-migrant households in the short run. (Although these households experienced both inflows and outflows of migrants over the crisis, we should not necessarily expect their experience to equal the sum of those who were net receivers and net losers of people, so there is no conflict with the results above.)

 

We are yet to unpick the details of how these results came about, but the overall story seems to be that migration helped families cope with a very major income shock in the short run and that in the long run that response had only positive or zero effects on consumption growth. That is, migration helped.

Professor L. Alan Winters is interim CEO of the Migrating out of Poverty Research Programme Consortium.