Scott Sumner  

Is China trade inflationary?

Banning the Benjamin Would Red... Kleiman on Trade and Taxes...

I like contrarian arguments, but the hypothesis in the title is a bit too much for even me to accept. Instead I'll argue that the fans of the recent paper by Autor, Dorn and Hanson (ADH) are implicitly making this claim---probably.

I'm going to work with the AS/AD model, so let me start off by reminding you that the AS/AD model (i.e. the simple diagram) doesn't really rule out any hypothesized transmission mechanism. There are four possible co-movements in P and Y, and any are possible depending on the type of shock that hits the economy.

As I pointed out in a recent post, it's not entirely clear what ADH were arguing, but most people (including me) seem to read them as questioning the standard view (among economists) that free trade with China benefits the US. ADH point out that lots of jobs were lost in specific areas, and there's not much evidence of workers being reallocated to other sectors. I criticized this argument, partly on the grounds that they were drawing macro implications without an adequate macro model. For instance, there was no discussion of factors such as monetary offset.

I also pointed out that the period they considered (1990-2007) saw excellent macro outcomes for the US, and we were not at the zero bound. Thus it was reasonable to assume that the Fed had the path of NGDP pretty much where they wanted it. Furthermore, any attempt to boost AD through actions such as trade barriers on imports, almost certainly would have triggered offsetting tightening by the Fed, leaving total AD no higher than before. Indeed if you view the Fed as an inflation targeter, not an NGDP targeter, we would have been worse off with trade barriers.

Then I argued that the "reallocation of labor problem" was the only plausible transmission mechanism that could make their hypothesis true, and also pointed out that ADH frequently discussed the reallocation problem.

So if the hypothesis that China trade reduced AD makes no sense on either empirical or theoretical grounds, then let's assume that fans of the ADH paper are thinking in terms of the reallocation problem. What are the implications of this hypothesis, in terms of the AS/AD model?

The implication might surprise you. If we assume that the Fed keeps AD stable, and offsets shocks that would move it left or right, then China trade can only have hurt the US economy by reducing AS. Now that's certainly possible; if reallocation didn't go well, it would result in higher unemployment, would depress RGDP. But it would also raise prices. In other words, the only plausible mechanism by which China trade could have hurt the US as a whole (rather than just specific groups impacted by imports) is also a mechanism that implies China trade had an inflationary effect on the US economy.

Maybe it did.

I would guess that if you polled Americans, the majority might believe that imports from China hurt the US economy. I think a majority would also acknowledge that China trade somewhat lowered the cost of living, relative to no trade with China. That means we shouldn't take seriously what average people believe, because their beliefs are not even internally consistent. They don't understand economics. Asking average people about the net benefits of trade is like asking them whether they favor the Copenhagen interpretation or the "Many Worlds" interpretation of quantum mechanics.

I don't agree with those who claim ADH is an important paper calling into question the claim that free trade with China benefits the US economy. That doesn't mean I am not willing to consider their views---I am willing. But only if they combine their support for ADH with a loud affirmation that:

Trade with China has an inflationary effect on the US!
Will they proudly affirm that view? Or will they shy away, in fear of being ridiculed? Let's keep a close watch on this issue, and see how it plays out. Please send me links to any posts you see where ADH fans claim China trade is inflationary for the US. If we don't see any, then we need to draw the obvious conclusion.

PS. If you are wondering whether the same argument applies to fiscal stimulus during the 1990-2007 period, the answer is yes. During that period fiscal stimulus was expansionary if and only if it was deflationary (assuming monetary offset). That's actually an implication of the New Keynesian model; I'm not making it up.

PPS: I have a post on trade deficits over at TheMoneyIllusion.

Comments and Sharing

COMMENTS (10 to date)
Lorenzo from Oz writes:

China trade as inflationary: a view that implies that is, indeed, odd.

Some years ago, I read a commentary piece by a prominent Australian economic journalist arguing that central banks were heroically keeping up the money supply in the face of the tsunami of Chinese exports. There is a touch of that in this Bloomberg piece on the deflation in durable goods.

fralupo writes:

Assigning to the trade deficit the blame for the counter-cyclical monetary policy response to that deficit is interesting.

Wouldn't a better retelling be that trade deficits (in the AD-AS model) are recessionary and monetary responses to recessions are inflationary?

ThomasH writes:

I do not see how you can think that a study of the costs of opening of trade with China for a particular group of people can be construed as finding that is was on balance a negative supply shock or that given suboptimal monetary policy translates into a negative monetary shock.

Market Fiscalist writes:

Lets assume that monetary policy is perfect and it allows RGDP to always be at its optimal level.

Lets say that a lot of American are engaged in production of goodA. Then the Chinese find a way of producing goodA more efficiently. Its price falls leading to an overall fall in the price level. While all other Americans are better off because of cheaper goodA, the producers of GoodA are worse off, even if physical output remains the same. They will remain worse off until they learn to produce GoodA as efficiently as the Chinese, or switch to another good that they can produce cheaply enough to allow them to maintain their previous consumption levels.

William writes:


I am always puzzled when you talk about supply shocks : How can you sympathize with cost-push theories if inflation is all about money? Aren't there then two types of "inflation" (demand (money) and supply-side which just has to do with relative prices) and we constantly mix both up?

Market Fiscalist writes:

As the fed was (successfully) targeting inflation during the period in question then any underlying effect of trade with China on prices would have been masked , wouldn't it ?

Philo writes:

"Asking average people about the net benefits of trade is like asking them whether they favor the Copenhagen interpretation or the 'Many Worlds' interpretation of quantum mechanics." Except that if you ask them about quantum mechanics they are likely to respond "No opinion," while if you ask them about economic issues a good many will, with passion, offer you an opinion. (True, the passion is suspect, since very few would be willing to bet their own money on the accuracy of their opinions.)

Scott Sumner writes:

Lorenzo, Yes, I've seen that sort of story as well.

fralupo, When you contemplate adopting a new trade policy, you take the Fed's dual mandate as a given. Any effect of that trade policy, assuming the Fed continues targeting the variables it has always targeted, are best treated as being caused by the trade policy.

If I'm wrong then other examples of offset, such as "crowding out", would make no sense even if "true."

Thomas, You said:

"I do not see how you can think that a study of the costs of opening of trade with China for a particular group of people can be construed as finding that is was on balance a negative supply shock "

I never did. I only commented on studies that claimed China trade reduced overall employment, a very different argument. Indeed I specifically acknowledged that China trade reduced employment in specific industries, something all economists have understood since the days of David Hume.

William, Under the Fed's current monetary regime, supply shocks will cause higher prices. In addition, under the optimal monetary regime (NGDPLT) supply shocks would cause higher prices.

Market Fiscalist, The Fed had a dual mandate, hence if China trade reduced aggregate employment, then it should have been inflationary.

Philo, Actually, if you explain the QM to people, most will oppose the many world's view. Recall that this view implies that trillions of near perfect copies of you are created each second.

Market Fiscalist writes:

"if China trade reduced aggregate employment, then it should have been inflationary.".

True, but China trade can hurt some Americans without reducing employment. For example: If China started producing goods cheaper than Americans could, and some of the Americans previously producing those goods switched to producing other goods that paid lower wages (or just got lower wages for introducing the same good).

In this case the underlying pressure on prices would be downwards (cheaper imports and lower local wages) and the fed would not have to exceed its inflation target to maintain whatever its target employment level might be.

Peter writes:

The reduction in AS from the labor reallocation problem would raise prices, but wouldn't the increase in cheap imports from other countries be a positive supply shock making the net impact on inflation undetermined? Or are you already accounting for the cheaper imports? Or is that not even relevant to the point you're making?

Also, would we have to consider changes on industry composition from trade? Say an industry is opened up to trade from china and so that industry's output in the US shrinks so labor demand in the industry falls and people are laid off. Those people have trouble "reallocating" so there is the decrease labor supply. Since we are assuming AD is kept growing normally, output in another industry must be increasing. But, that increased output might come from a capital intensive industry (Is it Hecksher Ohlin that predicts these kinds of reallocations due to trade?). So labor demand does not increase in this industry enough to offset the fall in labor demand from the other industry. So we have a fall in labor supply and an offsetting fall in labor demand - leaving the change in wages undetermined and the impact on inflation undetermined. Am I hopelessly confused?

Comments for this entry have been closed
Return to top