Alex, Tyler and I rented a boat and our captain took us to a Salvadoran village, San Francisco, maybe 500 people.
Speaking Spanish, our captain told us, "The majority of the people from this town are in the United States." They sent back remittances, which explained the remodeled homes, the satellite dishes, the schoolchildren buying ice cream.
I thought about Richard Rogerson's paper, "Indivisible Labor, Lotteries, and Equilibrium."
Rogerson said that there's a high fixed cost of going to work (commuting time, putting on your game face). So people's work lives will tend to be all-or-nothing: 40 hours or zero hours.
In that kind of world, it's more efficient for half the population to work full-time rather for the whole population to work half-time; why should everyone have to get up early?
Rogerson shows such people would rather have a work lottery: The people who draw the short straws have to go to work and have to share their cash with everyone who didn't work.
The people of San Francisco are living in Rogerson's world: One family member draws the short straw and has to go work in the U.S. She sends back most of her money, some of which is used to build a vacation/retirement home for the unlucky worker (A Philippine example here).
There is good work on the micro-level causes of remittances: Stark and Lucas for example. But the social causes are underresearched. One study surveyed here found that even controlling for the usual individual factors, what mattered most was what town you were from.
“[W]hatever factors governed migrants’ decisions, they operated at the community level"
"[F]urther research is required on the social determinants of remittances"
Rogerson's model is often right, we just don't know why.