Taylor, J Edward. 2004. "Remittances, Savings, and Development in Migrant-Sending Areas." Pp. 157-73 in International Migration: Prospects and Policies in a Global Market, edited by Douglas S Massey and J. Edward Taylor.
Taylor begins by stating that other scholars are essentially incorrect in referring to remittances as a negative for local development. He says this is because they underestimate the amount of remittances (in part because it's hard to estimate the value of in-kind remittances) and don't consider how remittances boost the economy in the area that receives them, or the multiplier effect they have in the local economy in the receiving area. Taylor believes that remittances are helpful because they have an "equalizing effect" in receiving areas, since they increase the incomes of households at the bottom to middle of the economy, and can even help those households achieve economic mobility (p. 157).
Taylor then begins to attempt to quantify remittances. He makes a few significant points. First, remittances are not equally distributed around the world; some countries and areas within countries receive more of them. In some areas, remittances are greater than a quarter of the value of export revenues. Globally, they are greater than the value of official development assistance. Studies also suggest that the value of remittances is a significant percent of household income in receiving families. In short, the value of remittances is not at all trivial.
In the next section, Taylor refers to the "new economics of labor migration" (NELM). According to NELM, "migration is hypothesized to be partly an effort by households to overcome market failures that constrain local production" (p. 160). For example, if an area does not have a good insurance market, remittances provide a form of insurance against a loss of farm income because the remittance income will be there even if the crops are lost or prices crash so that the crops don't bring in the needed income.
Taylor says this is a different way to view the connection between migration and development compared to neo-classical economics and dependency theory. A major difference is that this approach connects the reasons for migration with the effects of migration to sending areas. It leads to new types of research because researchers now don't just ask about remittance income and migration, they also ask about all aspects of farm income and production. This allows for new forms of analysis compared to before.
I find it interesting that so much of the literature Taylor reviews is framed as a simple matter of markets (i.e. migration is due to "imperfections in capital markets" on p. 161), and there is no mention of social and political factors. I also find it difficult to understand all of the economics in the lit review, and suspect much of it to be bullshit. (That goes along with my general suspicion that a lot of what economists say is bullshit.)
One point that makes sense is that remittances might increase agricultural productivity by allowing receiving families to invest in their farms in ways that increase productivity. However, remittances may also be a way to reduce risk by spreading it out over diverse activities. Taylor says it could be a combination of the two, and I believe it is. Even simply being able to purchase inputs outright instead of on credit reduces the need to pay interest; or being able to purchase land instead of leasing it increases profits. That is, in the former case, to the extent that the purchased inputs actually help instead of hurting the farm, which is something I have my doubts about. (For example, see Glenn Stone's work on farming and deskilling in India.)
In citing Adams (1991), Taylor notes that policy can play a role, because policies that depress prices for agricultural output discourages investment. That is, if you aren't even going to make much money for growing something, why would you invest money into growing it (or growing more of it)?
Taylor finally sheds light on why some scholars think remittances are bad. In the first year of migration in one study, "a $1 change in remittances produces a less-than-$1 change in total incomes of remittance-receiving households" (p. 161). I can only imagine this is because the labor of the family member who migrates was worth more at home on the farm than the amount they can send in remittances. If that's not the case, then I totally don't understand how the heck this can be possible. In any case, measured 6 years later, every $! sent in remittances produces more than $1 in increased household income. Taylor later makes clear that this is because the remittance income allows the family to invest in a way that increases their ability to make money (for example, purchasing land or agricultural inputs, purchasing livestock, or buying inputs outright instead of on credit).
His point is that there is variation "across time and settings" in the effects of remittance income and "the positive effects clearly depend on the magnitude of migrant remittances and the profitability of investing in new production activities or techniques" (p. 162). He notes that this can be limited in three ways: environmental and market constraints, or policies that disadvantage agriculture. In other words, if your land is infertile in the first place, you won't get much bang for your buck by investing in a tractor or hybrid seeds, etc. Likewise, if the prices you'd get for what you produce are low either (perhaps because of policies like free trade agreements), you won't get much economic benefit by investing in agriculture. In any case, Taylor thinks this variation could account for why some studies found remittances are more harmful than beneficial.
In the next section (p. 162), Taylor states that treating each household as its own isolated economic unit is wrong because it ignores the interaction between households. Honestly, this is flipping obvious and any economist who misses it should have his or her degree revoked. When one household in a village receive money from abroad, they spend it locally, and that puts money into the hands of other families in the area. Or foreign corporations as the case may be if they are buying seeds from Monsanto or DuPont. But some of the money goes to the local agro-input dealer at least. This is freaking economics 101. The money multiplies in the local community. The agro-input dealer takes the extra income and uses it to buy whatever he or she needs too - shoes, school fees, food, medical care, etc. And whoever sells them the shoes or food, etc, turns around and spends the money to meet their own needs, and so on. Presumably the entire value of the money is not re-spent each time, because some of it may be saved. Additionally, some of it is going to distant corporations and not to local families. (For example, if a family purchases DuPont seeds at the agro-input dealer, part of the value of their purchase goes back to DuPont. But the part retained as profit stays in the community, as do the part that pays wages of the workers in the store.) So thank you Taylor for explaining elementary economics to a lot of people who should have known better.
Jumping off of this point, Taylor says that it's crucial to examine household expenditures in remittance-receiving households because that tells you how the remittances provide indirect benefits to the local economy through the multiplier effect.
Later, Taylor provides a clearer explanation based on a study in Michoacan, Mexico: "Remittances from migrants stimulated non-remittance income in the Michoacan-survey households in three ways. First, they enabled migrant households to purchase inputs (i.eg., fertilizer) that increased income in the short run. Second, they provided migrant-sending households with funds to invest in income-producing assets - particularly livestock - which created new sources of local income in the long run. Third, they created expenditure linkages in the local economy that transmitted the positive effects of remittances to other households - including those that did not have migrants in the United States" (p. 166). As a result, every $1 received resulted in $1.85 additional income in the receiving household. Within the village as a whole, every $1 in remittances increased village income by $1.60 (p. 167).
I have to admit, this article is doing a lot to reinforce my general notion that economists are idiots. Taylor's essentially explaining all of this, all of which makes perfect logical sense, because it flies in the face of what economists had accepted as truth about remittances.