This slowdown could be a blip — a few months does not a new trend make. But we doubt it, because there is a good explanation for the change: the sharp increase in value of the U.S. dollar in foreign markets.
When the value of the dollar increases, it means that a dollar is worth more in terms of foreign currencies. A year ago, it cost about $1.40 to buy one euro; today, because the dollar has gained strength, it costs $1.12.
That’s a sweet deal if you’re an American in Paris. But it’s a tough deal if you’re an American exporter, because it means our exports cost more in foreign currencies and imports from abroad are cheaper here. Although this data is also noisy, the U.S. trade deficit in manufactured goods has increased by more than $100 billion between the first quarter of 2014 and the first quarter of 2015.
Therein lies the connection to the flattening trend of manufacturing jobs, a connection that has been widely recognized in recent economic commentary. As Politico put it, “The rising dollar — lifted by stronger U.S. growth and weakness in Europe and elsewhere — makes domestic products more expensive abroad and limits the need for U.S. manufacturers to hire more workers to make more products.”
Meanwhile, as these dynamics have been unfolding in international markets, the White House is busy selling a trade agreement — the Trans-Pacific Partnership — that it says will take down trade barriers and boost American exports. Yet the TPP contains no currency provisions, meaning there are no enforceable rules in the deal that would prevent our trading partners from managing their currency so that it stays low relative to the dollar, invoking similar dynamics to those just described.
The recent strengthening of the dollar probably has more to do with relative growth rates between countries than currency manipulation — at least in 2014, the U.S. economy was speeding up as the economies in Europe and Asia were slowing. But the weakness in manufacturing employment over the past few months is a reminder of how currency valuations between countries can play out in ways that make it a lot harder for our producers to compete on a level playing field in international markets.
How big a deal is this? Previous research has found that a few of the countries in the TPP have a history of managing their currency to subsidize their exports. We estimate that currency interventions by Japan, Malaysia and Singapore have cost us 250,000 to 320,000 jobs annually over the past few years.
Moreover, there are two reasons to worry that the damage from our failure to include enforceable rules against currency management will go beyond these few countries. First, the TPP is touted as a framework that other countries will want to sign on to. China has used its large trade surpluses to buy almost $4 trillion in recent years, keeping the value of the Chinese yuan artificially low relative to the dollar. According to our estimates, that has been the source of 1.3 million to 1.6 million jobs lost (see our data note below as to how we make these calculations).
In fact, a full accounting of the jobs lost to the set of countries that have suppressed the value of their currency relative to the dollar over the past five years comes to 2.5 million to 3 million jobs per year.
Second, although we strongly disagree, the administration has quite clearly argued that we are helpless in the face of these interventions. We can do what we’ve always done, they say — try to persuade other countries not to manipulate their currencies — but we cannot enforce rules against such actions without putting our own Federal Reserve in the cross-hairs of the international tribunals set up by the TPP.
Not only is this wrong in terms of the substance, as Simon Johnson explains here, but it’s in effect a declaration that the administration doesn’t intend to take any serious measures to address currency imbalances.
In a speech last week at Nike’s headquarters, President Obama lashed out at opponents of TPP, saying they’re flat wrong on the substance. But although we can quibble over the precise jobs numbers lost to currency management, the vast majority of trade economists, including those at the White House, agree that this practice has led to larger trade deficits that have hurt our manufacturing workers.
We don’t understand how going to Nike, a firm that employs 26,000 workers in the United States and 1 million abroad, and telling those of us who’ve carefully tracked the impact of globalization for decades that we just don’t understand the benefits of trade deals is supposed to persuade us to support the TPP, a deal we’ve not seen because it is negotiated in secret.
What we do know is what’s not in there: enforceable currency rules. And as the recent manufacturing and trade deficit data remind us, exchange rates matter. It would be much better for American workers if the White House would recognize that fact and engage with us on a solution rather than telling us how we just don’t get it.
Our job loss estimates from currency management use methods described in Bergsten/Gagnon and Scott. Bergsten/Gagnon list countries with records of buying dollar reserves to boost their current account surplus and they estimate the share of the change in the dollar holdings that maps onto the U.S. trade deficit. Using five-year averages of the change in dollar holdings, we take these values as a share of gross domestic product and, as per Scott, assume a 1 percent increase in the U.S. trade deficit as a share of GDP is equivalent to the loss of 1.2 million jobs.
Our estimates are conservative in various ways. First, we assume no multiplier (Scott applies a negative jobs multiplier of 1.6 to the increase in the trade deficit). Second, our method assumes only the change in holdings of dollar reserves affects the exchange rate and increases our trade deficit. In fact, the stock matters too, e.g., even if China does not add to its almost $4 trillion holdings, its massive dollar reserves place upward pressure on the relative value of the dollar. Bergsten/Gagnon write that “China’s heavy past intervention has lingering effects on the level of the exchange rate, keeping it considerably lower than it would otherwise be.” Our method ignores this effect.