Hello everyone. As usual I’ve come to ask for feedback on my latest research. I can’t emphasize enough how much it helps to blog it out, if only because it forces me to sit down and try to summarize things in a few hundred words.
My current research is looking at the effect emigration has, if any, on institutions. Institutions come in various forms. The state is an institution, but family, religion, and even organized crime are too. Broadly speaking institutions are those rules that govern society, both formal and informal. Institutions have increasingly been acknowledged as being one of the key (if not the key) determinants of a nation’s wealth.
Despite the importance of institutions, we know relatively little about them. By no means is this due to a lack of trying, and in there have been some earnest attempts to tackle the issue. Acemoglu’s Why Nations Fail is one such attempt.* For the time being the goal in institutional studies is to properly explain how and why institutions form.
My goal is to neither explain the origin of institutions or to measure their impact on economic well being. I take it for granted that good (and bad) institutions populate the world. Instead I am interested in how different institutions interact with one another.
My former boss at Cato has looked at how immigration has influenced a destination country’s (the USA) institutions. He finds little effect. In my project I try to look at the problem from the opposite end – how does emigration influence an origin country’s institutions. To measure the impact of emigration I use remittance data.
Remittances come in two form. There are monetary remittances, which are cash transfers from emigrants to their family members and friends back home. There is a broad economic literature on the former and its affect on development outcomes. There is however little (if any- I haven’t found any at least) economic work on social remittances. Social remittances is the transfer of ideas from emigrants to their family members and friends. In general the economic remittance literature has not yet attempted to connect itself with the institution literature despite both being part of the larger development literature.
Most work on social remittances has been done by sociologists. Thus far though most of the work has been qualitative and/or focused on how social remittances tie migrant communities with their origin countries. There has been little work on how this communication translates to changes in institutions.
Political scientists are currently taking the lead on the question. Earlier this year Abel et al. published a paper looking at how remittances affect democratic transition. They find that increased monetary remittances decreases voter turn out and thus weakens the political base of populist-based autocracies. Another recent paper by Miller et al. find that emigration increase the possibility of civil war by giving opposition parties an external funding source.
I think Abel and Miller’s work the best thus far in seeing how emigration affects institutions. My biggest concern with Abel’s paper is that he looks at democratic transition events, but there is no reason why democracy must lead to better institutions. Hong Kong and Singapore alternate as the most economically free states in the world, but neither is a bastion of democracy. India is the world’s largest democracy and by most metrics has awful institutions.
Miller’s work on the other hand looks at how the probability of civil war increases, but civil war in itself is not always bad. On occasion war is necessary for the improvement of institutions**.
To remedy my concerns I look at how remittances marginally influence institutions. I use the Fraser Institute’s Economic Freedom in the World summary index as my measure of a country’s institutions. My regression tables are found below. All observations are for north American (including central America but excluding the Caribbean) countries from 1994-2012.
Column 1 is a simply regression between a country’s EFW score and remittances as a percent of GDP. Initially we find a negative correlation between the two – a 1 percentage point increase in remittances is associated with a 0.01 point decrease in its EFW score. Is this a sign that brain drain, the emigration of high skilled migrants, is reducing the institutional qualify of origin countries? Not quite – it’s simply caused by the lack of control variables. At this point remittances is a proxy for a country being undeveloped.
Columns 2-4 are me playing around with various control variables. The interaction of phone subscriptions with remittances is my attempt to proxy for social remittances. Presumably emigrants are more likely to call back home, and exchange ideas, if their family members and friends have a phone to be contacted at. The 1 year lagged EFW index symbolizes the ‘stickiness’ of institutions: in the short run institutions do not drastically change.
Column 5 is simply column 4 re-run using clustered errors and country fixed effects. Country fixed effects, for those of you who have been spared endless hours of statistical classes, is a technique that allows us to account for unobserved characteristics of a country that do not change across the observed time span. This is usually done to account for such things as culture or geography.
In this final iteration we find that a one percentage point increase in remittances increases a country’s EFW index score by 0.05 points. This is a marginal effect, but its not irrelevant. See the Cato Institute’s interactive map of economic freedom. The difference between the United States and Russia is about one point despite the former presumably being a bastion of freedom.
*Why Nations Fail has been discussed on NOL several times before, see here and here.
** But let me emphasize that this is rarely the case and war should be the last option. We really do need to make our own NOL foreign policy quiz.
|VARIABLES||EFW Index||EFW Index||EFW Index||EFW Index||EFW Index|
|Remittances as a percent of GDP – Fraser EFW||-0.01*||0.04***||0.00||0.01**||0.05***|
|Fixed telephone subscriptions (per 100 people)||0.03***||0.01*||0.01|
|Remittances * Phone||-0.00||-0.00||-0.00**|
|EFW Index 1-year lag||0.81***||0.78***||0.54***|
|Income Per Capita in 000s, Constant 2005 dollars.||0.00*||-0.00||-0.01|
|Country Fixed Effects||No||No||No||No||Yes|
Standard errors in parentheses in columns 1-4. Robust errors in column 5.
*** p<0.01, ** p<0.05, * p<0.1