The recent surge in remittances from migrant workers to their native countries may be overestimated, according to research by Michael Clemens and David McKenzie, published in the July 2018 issue of the Economic Journal. Their study suggests that the true increase in remittances may be too small for it to have detectable growth effects back home. What's more, any growth effect of remittances may be offset by the ‘opportunity cost’ – the fact that to get that money, workers have to leave.
These findings do not undermine the strong evidence that migration and remittances have notable effects on reducing poverty for migrant workers and their families. But they do help to explain why countries with big shares of the apparent worldwide surge in remittances have not achieved substantially higher growth. The researchers conclude:
‘Migration and remittances appear to expand greatly the economic possibilities of families without harming – but also without greatly helping – the economic possibilities of the broader economies from which they depart.’
Migrants sent about US$450 billion to developing countries in 2017, ten times more than they did just 20 years earlier. These ‘workers’ remittances’ are more than four times the value of all foreign aid for development. The United Nation is developing a new Global Compact on Migration, and many hope that such an agreement will help economic growth around the world by encouraging remittances to poor regions.
But economists searching the data have largely failed to detect such an effect. They find that countries receiving a larger share of the worldwide surge in remittances wave have not achieved substantially higher growth. The new study explores how that could possibly be the case, and offers three surprising answers.
First, the surge in remittances in recent decades may be largely an illusion. Much of it may reflect more and more financial flows being counted as remittances but which are not.
For example, the study compares aggregate measures of remittances into Mexico, collected by the central bank, with nationwide surveys of how much Mexican households report receiving from abroad. Those two numbers are quite similar until the year 2000. But shortly thereafter, the central bank’s measure of total remittances tripled, while the household-reported measure barely changed.
This may arise in part from illicit financial flows – such as clandestine cash movements related to international organised crime – being increasingly counted in official statistics under new reporting standards enacted after the terrorist attacks of 2001. It is no surprise that Mexican growth was unaffected, because this represents mostly a change in how money flows were counted, not changes in the flow of money.
Second, the true increase in remittances may simply be too small for its growth effects to be distinguished amid the statistical noise. Suppose the true rise in remittances to developing countries over the past decade is, say, half of the standard estimates (US$75 billion more over the past ten years, instead of US$150 billion).
The combined GDP of all developing countries is around US$45 trillion per year. That means that the per year rise in true remittances to developing countries would be 0.017% of GDP per year – just one sixtieth of one percentage point per year. That is likely to be smaller than the margin of measurement error for most economic data from many poor countries, so it would be very hard to detect in the average country.
Third, the single largest reason that some countries get more remittances than others is that some countries have more workers abroad than others. When workers are abroad, they do not produce for the home country economy. So any growth effect of the remittances they send is offset by the ‘opportunity cost’ of the fact that in order to get that money, workers have to leave.
Some analyses have found positive effects on growth from the same number of workers sending more money (such as when exchange rates shift, so that the same number of dollars abroad is worth more pesos or CFA francs at home). But that is not the principal reason why remittances rise in normal settings. They mostly rise because more workers are leaving, and their absence has an offsetting effect on growth.
None of these results question the effects of remittances on poverty in poor countries. The study concludes that there is strong evidence that migration and remittances have first-order effects on reducing poverty for the workers who migrate and for their immediate and extended families.
Yet while remittances are a vital economic lifeline for many of the poor, economists lack strong evidence that they are an engine of economic growth for the broader economies from which migrants come. Migration and remittances appear to expand greatly the economic possibilities of families without harming – but also without greatly helping – the economic possibilities of the broader economies from which they depart.
‘Why Don’t Remittances Appear to Affect Growth?’ by Michael Clemens and David McKenzie is published in the July 2018 issue of the Economic Journal.