In the debate about global trade imbalances, we often hear it said that because Americans produce nothing that China exports to the US, any move to restrict Chinese imports to the US would have no employment effect on Americans.  A forced contraction in Chinese exports would simply result in an equivalent increase in the exports of some other country, let’s call it Mexico.  Mexico would benefit from US trade action, but the US wouldn’t.

There are many reason for opposing trade war between the two countries – the two most important being, in my opinion, that trade war will result in slower global growth than a negotiated settlement, and that the importance of the US-China relationship involves a lot more than economic issues.  A troubled relationship between the two spells potentially bad outcomes for issues such as the environment, global terrorism, and nuclear proliferation.  But the argument that China trade has no impact on US employment is I think a very weak one.

In the first place, China doesn’t simply produce slippers, lighters and toys to sell to the US.  Chinese growth is heavily capital intensive, and China produces many things that Americans produce or used to produce until quite recently – including automobiles, steel, chemicals, advanced metal products and, soon enough, aircraft.  Remember also that China’s import substitution will have as big an impact on trade as export support.

Anyway, it is hard for me to imagine that the explosion in US imports from China in the past decade could consist of nothing that Americans produced themselves but were already importing from other countries.  If that were truly the case, wouldn’t China’s rising exports and trade surplus in the past decade be balanced wholly by declining exports and trade surpluses in other countries?  US imports and the American trade deficit should have held fairly steady.

But even if I am wrong, and China and the US produce absolutely nothing in common, it is irrelevant.  The claim that there can be no employment impact in the US of a contraction in Chinese trade could only be true if all trade settled only on a bilateral basis.

Globalization is not bilateral

There may have been a time, two or three hundred years ago, when trade settled mostly on a bilateral basis, but I would submit that by the late 19th Century, and certainly in the past few decades, trade never settles bilaterally.  It must settle multilaterally.  In that case it is pointless to talk about the overlap between US and Chinese production.  Let me try to explain why.

Let us assume there are four countries in the world: China, the US, Mexico and Brazil.  Let us also assume that China runs a large trade surplus with the US and that the two countries produce completely different sets of goods in which there is no overlap.

If the US takes actions to force up the value of the renminbi and so reduce Chinese exports to the US by $100, Americans would still need to buy those goods from someone else, albeit at higher prices, since according to our admittedly unrealistic assumptions they do not and cannot produce them on their own.

Rather than buy these products from China, Americans will buy them from Mexico.  In that case nearly the full quantity of lost Chinese production would shift to Mexico (higher prices would reduce US demand somewhat).  Chinese workers would get fired and Mexican workers get jobs.

So not a single unemployed American would benefit, and consumers would have to pay higher prices, right?  Not quite.  This isn’t the end of the story.  If Mexico does not intervene in its currency, two things are going to happen in Mexico.  First, the peso will strengthen as demand for pesos rises among American importers.  I am assuming that the surge in exports causes no capital inflow into Mexico, but if it does, that would put even more upward pressure on the peso.

Of course a rising peso shifts income from Mexican exporters to Mexican households.  It also reduces the profitability and demand for other Mexican exports and so causes Mexican exporters to reduce other production by some amount.  Total Mexican production in that case rises by less than $100, and either American or Brazilian production increases to fill the gap.

Mexican imports rise

More importantly, unemployed Mexican workers will now get jobs and will earn income that they weren’t earning before.  Real household income in Mexico will have risen by quite a bit.

Since it is a pretty safe bet that Mexican workers don’t save 100% of their additional income, especially if they were formerly unemployed, Mexican consumption must also rise.  Notice, then, that total Mexican production rose by less than $100 and total Mexican consumption rose by some large amount, perhaps very close to $100.  I am making the unreasonable assumption that there is no increase in Mexican investment associated with the increase in Mexican exports, but if there were any increase in investment, it would increase Mexico’s domestic demand even further without increasing Mexican production.

So Mexico’s trade surplus will not rise by anywhere near $100 dollars.  It will rise by a lot less than that.  Since the rising renminbi shifted Mexican exports upwards by $100, that must mean that Mexican imports also increased, and if that is the case they must have imported from somewhere – perhaps from the US or Brazil.

If the increase in Mexican imports was satisfied by increased US exports, then of course there is likely to be a positive impact on US employment.  If they imported part of it from Brazil, then we have to go through the whole exercise for Brazil and come back to at least some additional US exports to Brazil.  Mutatis mutandi a reduction in China’s trade surplus must result in a reduction in the US trade deficit and an increase in US employment as long as Brazil and Mexico don’t intervene in trade.

One obvious flaw in my argument is that I have left China out of the picture once I assumed a drop in its exports.  But why couldn’t the full increase in Mexico’s imports come from a surge in Chinese exports to Mexico?

In fact it could, if China intervened to counter the impact of declining exports to the US, say by reducing interest rates or by increasing subsidies to manufactures in other ways.  The rebalancing impact of an increase in the renminbi, or of an increase in US tariffs, would be offset by higher subsidies, paid for of course by the Chinese household sector.  In that case China’s trade surplus could even rise.

This would create a worse problem both for China and the world, especially for poor Mexico, who will be dragged even deeper into the unsustainable US-China imbalances.  It would mean that China still exported deficient demand into a world struggling with anemic demand growth while, at the same time, worsening its domestic imbalances, perhaps by increasing investment in a country where investment levels are already dangerously high.  And it would mean that US unemployment remained where it was.

Trade is more complex than we assume

So what would I conclude from this little exercise?  Three things.

First, the no-overlap argument – that the US cannot benefit from a reduction in the Chinese trade surplus because the US produces nothing that China sells – is silly.  Not only does the US produce (or could produce) many things that China sells, but more importantly it doesn’t matter if it does.  Trade does not have to settle bilaterally.  In fact it almost never does.

Second, what matters is the totality of Chinese intervention.  If a rising renminbi, or trade tariffs in the US, are met by countermeasures within China, there might very well be no net trade rebalancing and even more dangerous distortions within the Chinese economy.

That is why it is probably better to target trade surpluses rather than just the currency, or just interest rates, or just wages, or just taxes, or just direct subsidies, or just any of a dozen factors.  It is encouraging, then, that US Treasury Secretary Tim Geithner seems to be more intent on arriving at a deal limiting current account surpluses rather than forcing a revaluation of the renminbi – although according to reports Germany is very much opposed to the idea.

Of course even targeting trade surpluses has its problems.  If China reduces its trade surplus not by boosting domestic demand but rather by boosting investment in commodities (stockpiling commodities), it simply maintains its domestic imbalances while increasing its exposure to commodity prices and pulling commodity-exporting countries even deeper into the global imbalance.  It also does little to rebalance globally because for the major commodity exporters the propensity to consume out of additional revenues is quite low.  In that case the imbalance has simply been shifted elsewhere.  I discuss this in a letter today to the Financial Times responding to Martin Wolf’s column on the subject.

The third conclusion is the all these things matter in the US too.  Measures targeted just at China might or might not work, depending on the Chinese response, and the wrong Chinese response can make both countries worse off (much worse off in the case of China).  If the US really wants to see its trade deficit decline, it should move aggressively to alter the balance between domestic production and consumption in a more permanent way – perhaps by raising consumption taxes, although this will work mainly by increasing US unemployment if China increases its intervention in the currency or in interest rates and credit.

Clearly the US needs to move aggressively to shift global demand in its favor because most countries are doing just that – that is what beggar-they-neighbor means, and in a beggar-thy-neighbor world whoever does not respond will bear most of the brunt of the pain.  There is no point pretending that we are not in a world in which nearly every country is cheating on trade, and will continue cheating as long as it is able.

But the wrong aggressive moves can easily make things worse.  I guess this means that the whole trade problem really will best be resolved by an intelligent multilateral agreement, with no cheating and no free-riding, that involves a set of determined and coordinated policies that bring the imbalances down gradually over the next several years.

But that is asking for a lot.  My guess?  Trade disputes will be resolved through more tariffs and currency interventions.  No one out there, it seems, is willing to do more than wish the imbalances away.  Actually to take the painful rebalancing steps is still not on anyone’s agenda.