Today, David Roodman at the Center for Global Development responded to our guest blogger Arvind Subramanian’s post (and forthcoming paper) on the effects of aid on manufacturing exports. Here, Arvind replies:
I cannot think of a more thoughtful follower of, and contributor to, the aid effectiveness literature than David Roodman (Aart Kraay is another). So, I am really very pleased with his bottom line assessment of my paper that he trusts this paper “more than most” in the aid growth literature.
That said, there is one point about his blog that merits a response. David gently chides Bill Easterly for his tweet where Bill interprets and presents our paper as showing that aid is bad for manufacturing exports. David’s point is that our paper strictly speaking only establishes a relative effect—that exportable sectors grow slower than non-exportable sectors —and not a total or overall effect: that aid leads to slower growth in exports as a whole.
But two points are worth noting. In a longer version of the paper, that I will post on my web-page, we do find evidence that aid leads to slower growth of the manufacturing sector as a whole. For methodological reasons, this result is less strong than our core result about relative effects. But, we certainly don’t get the empirical result that aid raises growth in all sectors that David claims (rightly) is theoretically possible.
Perhaps more important, Bill’s tweet does capture the spirit of our paper. Whether and how manufacturing exports can be an engine of overall growth is still debated. But the historical experience is strongly suggestive that if export sectors grow slowly or grow slower than other sectors, overall growth is affected. So, our paper could be interpreted not as a lament about the effects of aid on export sectors but as a celebration of its effects on non-export sectors. But, in my view and also in the historical record, between export and non-export sectors as an engine of growth, there is no contest.
4 Comments
Arvind, thank you for the generous response. But one thing leaves me quizzical: it seems as if this repsonse is based on a claim of having done a credible cross-country regression of manufacturing sector growth on aid. This is only a hair different from the sort of work that made Bill sigh about the:
Arvind, Bill, David -
I’m confused about something that I’ve seen cursorily referenced many times, but never explained particularly well. I appreciate the insight about how aid can crowd out sectors that might be the dynamic drivers of economic growth, but I often come across references to the impacts of aid on the exchange rate, and I fail to see the transmission mechanism. Granted my experience is limited, but most large aid projects, the kind that spend enough money to move exchange rates significantly, aren’t implemented by local firms. For example, when USAID or MCC contracts to build a road in Tanzania, no one is taking those hundreds of millions of dollars, converting them to Tanzanian shillings, and hiring a local firm to do this work, but this seems to me to be what is implied when people talk about aid moving exchange rates. Rather, what happens on the projects I’ve worked on is that US dollars are given to a construction firm from a developed nation, which then may convert a small portion of those funds to local currency to pay local workers, etc., but are these expenses big enough to move exchange rates so significantly that exports begin to become less competitive? This seems like a stretch to me. Do you know of any research out there that systematically looks at this effect, rather than just assuming it happens?
Arvind replied to Eric’s question at cgdev.org.
there is a second disease connected with aid, apart from the forex distortions that is called dutch disease:
the aid sector tends to suck away the best people in your economy, paying crazy salaries, resulting in an internal brain drain.
For a business trying to survive, it is damn hard to compete with the aid sector. Any competent economist/accountant/trader you employ prefers to earn more money as a driver for unicef, then to earn less as an economist/accountant/trader.
I don’t need a paper to understand that, just like the effects of dutch disease on exports.
One Trackback
[...] Arvind Subramanian has replied on Aid Watch to this [...]