Thanks to PGL for alerting me to this post by Steve Kyle, now at Angry Bear. Steve refers us to comments by Dean Baker, in which Baker argues that a yuan revaluation against the dollar is essential to any improvement in the US current account balance - higher savings in the US will not be enough to do it.
PGL points out that one can really think of this is a question of cause and effect (otherwise known as a chicken-or-the-egg debate). Would a cheaper dollar induce higher savings, thus bringing about the needed change to the US's savings/investment balance that would reduce the CA deficit? Or would an improvement in US savings cause the dollar to lose value, thus inducing US exports to rise and imports to fall and thereby improving the CA deficit?
Obviously, exchange rates and domestic consumption/savings behavior are both jointly and simultaneously determined. But if I had to pick one to be more exogenous, I would vote for savings behavior. I think that the US's underlying savings/investment balance has a lot more influence over the dollar exchange rate than the other way around. Dean suggests the opposite, and specifically argues that because the weak yuan has entailed the Chinese central bank buying lots of dollars, the Chinese exchange rate policy has effectively kept interest rates low in the US, which in turn has depressed savings.
But I'm not convinced. While I agree that interest rates in the US are lower today than they would be without Chinese exchange rate intervention (not to mention intervention by lots of other central banks around the world, particularly among certain OPEC countries), I think that there's pretty weak evidence that savings behavior is affected much by interest rates.
As a quick look at what I'm talking about, it's worth simply plotting interest rates against the savings rate in the US, as I've done in the following picture. Both series show 6-month moving averages.
I think it's hard to see a clear relationship between interest rates and the savings rate in the US in this picture. Personally, I believe that personal saving as a fraction of disposable income (which is what the red line shows) depends a lot more on changes in household balance sheets than it does on interest rates. While we probably shouldn't take that notion too literally, it is an interesting coincidence that the two big drops in savings rates that we've seen over the past decade both coincided with the final, frenzied stages of dramatic asset price appreciations - in the stock market in 1999, and in the housing market in 2005.
But I don't want to pretend that I really understand household savings behavior. What I do want to argue is that I'm skeptical of the notion that a different yuan/dollar exchange rate will somehow make the US savings rate rise. At the same time, it's fairly easy for me to tell a story about how an improvement in the US savings rate would cause the US CA balance to rise. That's why, if I had to choose a necessary first step for an improvement in the CA deficit, I'd vote for a change in US savings behavior.