Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


Growing out of the trade deficit

by Brad Setser
September 16, 2004

The U.S. government’s spin on the widening trade deficit is that it reflects a growth deficit in the rest of the world. This spin does not hold up well. All parts of the world other than Europe are doing quite well this year. Global growth is strong. In any case, the US does not trade all that much with Europe: US firms tend to produce in Europe for the European market, and European firms tend to produce in the US for the US market. The big trade flows are across the Pacific. Strong world growth is producing strong export growth: year over year U.S. export growth is 12.7%.Exports are growing faster than GDP, so exports as share of GDP are rising (good news — they were falling as a share of GDP from 00-02). The problem: imports are rising even faster — 14.2% overall (y/y). Even if you take out oil (up 22% y/y), goods imports up 13.4%.

The real problem is that the US already imports (1.5 trillion at the end of 03, 15% of GDP) so much more than it exports (1 trillion at the end of 03, 10% of GDP), that exports have to grow 50% faster than imports just to keep the trade deficit constant. So if imports are growing at 13%, exports need to grow at around 18% to keep the trade deficit from rising. Extrapolating based on monthly trade data suggests the 04 trade deficit (goods and services) will widen by $85 billion, to around $580 billion. A realistic estimate would add in another $10 billion to this estimate to reflect expected high oil prices, putting the overall deficit at around $590 billion for 04.

In the long-term, US exports need to rise as a share of GDP, so the US say exports 16% of GDP and imports 16% of GDP. But if imports keep rising substantially faster than GDP, exports are chasing a moving target.

Incidentally, this is one problem I have with Sebastian Mallaby’s analysis. He likes to emphasis that manufacturing is declining as a share of GDP (because productivity in manufacturing is growing faster than in services), and that it wrong to blame trade for manufacturing’s decline. That is partially right. But at the same time, if the rest of the world started spending the dollars it earns exporting to the US, i.e. the world started buying more US goods and fewer US financial assets (right now the world saves roughly 1/3 of what it earns exporting to the US as reserves), the US would sell $500-$600 billion more of goods and services abroad. Taking the current ratio of goods exports to services exports, that translates into at least $330 billion in additional goods exports — or an increase in goods production of around 3% of GDP. Trading treasury bills for imported goods on the current scale has to have an impact on the composition of US output and employment, and it must be reducing US manufacturing output below what is otherwise would be even in the context of a long-term shift out of manufacturing. This is not an argument against trade; it is an argument that the US needs to pay for its imports with current exports rather than with promises of future exports (external debt is a claim on future export revenue just as government debt is a claim on future tax revenue), and that the rest of the world needs to be spending more of its current earnings buying US goods and less buying US financial assets.


  • Posted by ed

    What is the savings rate in the Asian economies? The deficit seems to be driven more by a savings surplus than a growth deficit.

    Also, what is the process behind the accumulation of US Treasuries by foreign central banks? I understand that if a US telecomm concern, to use an example, sub-contracts with an Asian contract manufacturer there is a flow of dollars to purchase product. How do the dollars generally accrue to the Asian central bank? Is it primarily through taxes on the transaction or savings deposits in the country of production?

  • Posted by DF

    The dollars are exchanged in a bank (let’s make it simple and imagine there’s only one central bank in the country) against the local asian currency. Say Yen.
    The bank now holding the dollars could change them for yens, it should look for americans having yens to sell and possibly ask for less yens for a dollar. This would reduce the possibility to export. Instead of that it buys securities of the US government. Where do the yens come from then ? Well the bank issues more of them.
    Then there should be inflation … since there are more and more yens (and dollars) around. Well no because these extra yens are invested into building new facilities and shares in japan stock with little PER …

    In short : there’s a global asset bubble.

  • Posted by brad

    I would add one more step to the process described by DF, and to make my example work, lets replace Japan with China.

    step one — A US telco subcontracts to an Asian (say Chinese) firm, and gets paid in dollars.
    step two — the Chinese firms trades its dollars for renminbi at the central bank. It needs renminbi to pay its workers.
    step three — The chinese central bank lends those dollars back to the US, say to the US government, rather than selling them to a chinese firm that wants to import US goods.
    step four — as DF notes, the issuance of renminbi (cash) for dollars is inflationary, it is an increase in the money supply.
    step five — the central bank sells renminbi bonds for renminbi cash, reducing the money supply (sterilization)

    problem is that china’s capacity to sterilize $100 billion in annual reserve accumulation is limited, so there is some inflation. Inflation, in the context of a fixed exchange rate, will eventually lead to a real appreciation and more import demand. But that process is slow.

    And yes, Asian savings rates are very high — in china, i think is something like 40% of GDP. That lets them finance much higher rates of domestic investment than in the US and still lend some of their savings to the US, where investment to GDP is much lower than in China, and savings are even lower.

  • Posted by Khadija Khusro

    The steps given by Brad are the simplest explanation of how international trade affects economies. Two very critical aspects of these transitional dynamics are exchange rates and interest rates. Assume the Chinese central bank floats the renmimbi. Now it need not buy dollars everytime the renmimbi shows signs of appreciation. As a result it will not invest anymore in US securities. TO attract more capital the Fed will invaraibly have to increase its interest rates which in turn may result in inflationary pressures. This would make US goods less competitve globally, result in a higher deficit and currency volatility. Seems like an endless fall…?