Balance of payments issues are not macro issues
Recently I’ve seen a lot of misleading posts on international economics. Here’s one by Paul Krugman. And this one by the normally reliable Ashok Rao repeats some of the misconceptions:
Dean Baker is concerned that America’s persistent trade deficit is a result of East Asian mercantilism, contradicting the textbook story on international capital flows. Baker’s argument is correct – demand for dollars originating from an Asian “savings glut” forced us to run a trade deficit. But where Baker worries about declining employment from an overvalued dollar, I see potential for America to provide yet another international service and profit in the process.
True, a shirt imported is a shirt an American could have made, so a trade deficit costs us jobs on the margin. On the other hand, you can think of that really cheap shirt you’re wearing as a free good: a gift to America in exchange for liquidity. You’re still getting all the utility of wearing the shirt and loosing nothing in the process (unless you drive with a “Buy American” bumper sticker in which case you were probably screwed to begin with).
A stronger point still: as an American, it’s your moral duty to consume. The odd way the international monetary structure is set up gives us the responsibility of issuing the reserve currency. If India wants to buy oil from Saudi Arabia, it has to sell us something first. This guarantees that we run a trade deficit.
I’m afraid this is mostly wrong. Let’s start with the term “mercantilism.” What does it really mean? Does it mean a policy of discouraging imports and encouraging exports? Well a 10% tariff on all imports and a 10% subsidy on all exports nets out to zero. Nada. At least according to standard trade theory. Does it mean depreciating your nominal exchange rate? Yes, that may “work” in the short run (although it may not, as we saw when the US depreciated the dollar in 1933.) But even if it works in the short run it doesn’t work in the long run. The effects are offset by inflation, and the real exchange rate doesn’t change. Check out Latin America over the past 100 years if you don’t believe me.
Maybe is means a “pro-saving” policy, as the current account (CA) surplus is domestic saving minus domestic investment. OK, but I wouldn’t call that sort of policy “mercantilism.” I’d call it “smart” at least compared to the vast majority of countries that discourage saving. Contrary to Rao, Americans have a moral obligation to save more, not consume more. And it’s not even clear that the big East Asian surplus countries achieve those surpluses from government policies. In Japan the government spent enormous sums on wasteful infrastructure, which actually reduces the CA surplus. And they still has surpluses. Northern European countries tend to have larger surpluses than East Asian countries, and they aren’t accused of mercantilism. Some claim Germany benefits from the euro, but they’d run large surpluses even if they were outside the euro. The CA surpluses of Sweden, Switzerland and Norway are larger than in Germany (in per capita terms) and they are outside the euro. CA surpluses reflect saving/investment imbalances, not (in most cases) “mercantilist” policies, whatever you wrongly think mercantilism might mean.
And no, India does not have to sell anything to us to buy oil from Saudi Arabia. Nor does anyone else, if “selling” means goods and services. If we accept the odd claim that Treasury securities are somehow needed for the forex reserves of other countries (but why would that be true?) then those countries could get all the T-bonds they need by selling us assets of various sorts. It doesn’t require the US to run a CA deficit.
Nor do imports cost us jobs “on the margin.” If we import more than we export then we save less than we invest. So every extra dollar of imports means an extra dollar of investment or consumption, as long as the Fed targets NGDP. And if the Fed doesn’t target NGDP, then it’s monetary policy, not imports, which are costing us jobs. We had an even bigger CA deficit when our unemployment rate was 4.5%, and a few years earlier Germany’s unemployment rate was 10%. Trade is not the issue, NGDP is.
Then Ashok makes some excellent points about the real problem being a soaring demand for dollars, and not enough money. But then he gets off track again:
An excellent substitute for the United States dollar is a promise to pay a dollar later – known as Federal debt. Increasing the supply of the former requires us to print money and increasing the supply of the latter requires us to run budget deficits. Okay, I lied. It’s your moral duty to provide liquidity. The really sweet side effect of doing that is more stuff. Did someone say pre-school? Or solar power? In other words, the solution to the trade deficit “problem” is to allow for a larger budget deficit””U.S. government debt is one of our most important export goods.
The problem here is too little NGDP, and issuing more bonds won’t help. Otherwise Japan’s huge budget deficits would not have led to an astonishing decline in NGDP over the past 20 years. A much better solution for Japan would have been a 3% NGDPLT. Had they done so over the past 20 years their NGDP would now be almost double current levels, and the debt/GDP ratio would have been far smaller, for at least five reasons:
1. More inflation.
2. Slightly more RGDP growth.
3. Less unemployment comp.
4. Less discretionary stimulus on wasteful infrastructure projects.
5. Lower real interest rates (due to the zero bound problem with falling NGDP.)
Their monetary base would also be smaller, as a share of GDP.
As Paul Krugman asks, if we don’t run a deficit who will? Some things are zero sum. Better the issuer of the world’s most in-demand currency run a deficit than the guy who prints Rupees, which nobody seems to want. We provide the public good of liquidity in return for the free good of imports.
This argument was more problematic in the sixties when economists like Robert Triffin argued that we’d be forced to run a deficit forever causing an inflationary spiral. But in the sixties neither the domestic nor global economy was anemic, and inflation was a real problem.
Triffin was of course wrong, he confused running budget deficits with printing money. Reagan showed we could run big budget deficits and reduce inflation at the same time. Foreign central banks want T-bonds, not dollar bills. The inflation of the late 1960s did not come from the modest budget deficits, but rather from the acceleration in the growth rate of the monetary base. We printed too much money.
People should not mix up monetary and fiscal policy. Nor should they mix up international economic issues with macro issues. Yes, under fixed exchange rates they are linked, but that’s why we should have flexible exchange rates. And if you have flexible rates there’s no need for such massive holdings of international reserves.
I believe in K.I.S.S.:
1. Use monetary policy to do NGDPLT.
2. Enact pro-saving policies (a la Singapore) to offset the other anti-saving policies.
3. Let the market determine exchange rates and CA balances.
And for God’s sake let’s all stop acting like there is a qualitative difference between German workers building cars and exchanging them for Chinese built iPads versus Australian workers building condos and exchanging them for Chinese built iPads. Does geographical location of the German/Australian export really matter? Of course not. Then why do most pundits act like Germany is a muscular success while Australia is a feeble weakling?
Apologies to Ashok, most of this tirade had nothing to do with what he actually wrote. I actually agree with Ashok that we should not worry about the CA deficit. In contrast, I’d have to take some medication before addressing Krugman’s post.
Tags:
30. September 2013 at 11:03
It looks like we agree on the principal point that CA deficits are more or less irrelevant (not to mention measurement difficulties).
My point about India and Saudi Arabia is more concerned with an “overvalued” dollar due to our status as a reserve currency. (scare quotes because I don’t think something in efficient markets can be “overvalued”, but only relative to a more “egalitarian” monetary system). And I think this gives more credence to Triffin’s old worries. As the global economy expanded, gross transaction demand would follow, appreciating the demand for dollars. It seems reasonable to assume holding dollar Treasuries today is important for many emerging markets (fuel subsidies and external debt come to mind). Also if budget deficits are expected to be financed with money, Triffin’s link is valid.
Agree that something like an NGDPLT would have been better for Japan, but I’m definitely not arguing for budget deficits as a substitute for smart monetary policy as much as saying there’s little connection between a trade deficit and “secular stagnation”.
We agree about the employment effects. When I said an imported shirt costs jobs “on the margin” I meant it in the same sense as a robot costs us jobs on the margin. Neither actually cost jobs overall, but only in the microworld of the shirt factory.
Ultimately I don’t see a big disagreement. Foreign banks want to hold T bills, so why don’t we create more?
I’ve never seen trade deficits (or surpluses) as something specifically bad or good (except within non optimal monetary unions, *cough*). Ultimately, dollar liquidity strikes me (in some ways) as gold from a previous era: in which case we can either print money or run deficits. Both result in more “stuff” (or ownership of assets which is just a claim on future stuff).
30. September 2013 at 11:15
And you’ve told us interest rates, finance, the housing industry and the rest are not macro problems because EMH.
Fact.
John Paulson made $1,000,000,000 dollars in one day betting against the housing market in Feb of 2007.
Even Fama rejects your version of EMH — see Russ Roberts EconTalk with Fama.
Read Hayek’s Monetary Nationalism and International Stability to get some sense of the connection between international balance of payments and macroeconomics:
https://mises.org/document/570/Monetary-Nationalism-and-International-Stability
30. September 2013 at 11:50
Ashok, I suspected that we agreed more than your post suggested.
Greg, John Paulson big payday is 100% consistent with the EMH. If you don’t think so, then you don’t understand the EMH.
I have heard the Russ Roberts interview with Fama, and he doesn’t disagree with me on the EMH, AFAIK.
None of this has any bearing on the post.
30. September 2013 at 13:04
[…] Update: Scott Sumner reminds us that “Balance of Payment Issues are not Macro Issues“ […]
30. September 2013 at 19:05
Scott, what makes you assume that real (and nominal) interest rates wouldn’t actually be higher in Japan with better NGDP growth?
“When I said an imported shirt costs jobs “on the margin” I meant it in the same sense as a robot costs us jobs on the margin. Neither actually cost jobs overall, but only in the microworld of the shirt factory.”
I think you’re pulling a Krugman there, Ashok, I don’t think that’s what you’re readers would have understood by it.
30. September 2013 at 20:21
“The inflation of the late 1960s did not come from the modest budget deficits, but rather from the acceleration in the growth rate of the monetary base. We printed too much money.”–Et tu, Scott Sumner?
Yes, I know you used broad brush strokes.
But let’s consider if inflation was really so bad, given the tremendous positive results, and the huge structural impediments that existed then.
A 30 percent jump in real per capita income in the 1960s. Not bad, eh? Perhaps worth a little inflation?
And inflation? Well, check out this FRED chart. When measured as PCE chain-type index, inflation topped out at 4.5 percent in 1969. See also Table B-3 Economic Report of the President.
http://research.stlouisfed.org/fred2/data/PCEPI.txt
There is another broad brush stroke type of monetary history that is equally valid.
It goes like this: The Fed properly squelched inflation in the early 1980s by tight money—and then kept monetarily asphyxiating the economy ever since.
After 1980, real GDP growth rates fell below those of the 1960s and 1970s, and debt levels soared, due to artificially high interest rates. Japan-lite, let’s call it. Inflation headed into low single digits.
The Fed-zania for tight money reached a climax in the 2008 Great Recession, when the Fed clamped down at precisely the wrong moment, bringing down many borrowers, the banking system and the economy, and bringing on deflation in 2009-10.
My broad brush history is…not that far off from what really happened. And just as accurate as “inflation was too high in the 1960s.”
(Side note—Consider the huge structural impediments of the 1960s: A unionized labor force, huge market-making manufacturers, such as Big Steel and Big Auto, huge retailers, such as Sears (annual catalogue, btw), almost no international trade and thus less competition, regulated trucking trades, regulated airline rates, regulated phone rates, no Internet, no Craigslist, and a top marginal tax rate of 90 percent).
The simple dismissal of the 1960s today as one of “too much inflation” reflects the social norms of our era, and the inflation-hsyteria that defines the economic profession currently.
But would I take 30 percent real income growth in a decade, and inflation topping out at 4.5 percent?
Yes, I would, over and over again….
30. September 2013 at 21:19
The trade deficit is sustainable because American foreign investments have much higher market value to book value ratios than foreign investments in America. Since we profit so much from our foreign investments, we can use some of that profit for consumption (imports), while foreigners have to keep saving more in the us to achieve similar absolute growth in dollar profits because of their low returns, which means they have to send us even more goods and services. It would take a complete collapse of our foreign investments to have a chance of undercutting the trade deficit.
Here’s a brief post on the topic:
http://idiosyncraticwhisk.blogspot.com/2013/08/stop-hatin-on-trade-deficit-aka-america.html
1. October 2013 at 02:52
“Maybe is means a “pro-saving” policy, as the current account (CA) surplus is domestic saving minus domestic investment. OK, but I wouldn’t call that sort of policy “mercantilism.”
The ´problem´ is not saving, the problem these economists are talking about is the difference between investment and saving. A country can have a very high saving rate and an even higher investment rate, and thus a CA deficit. A country can also have a very low saving rate and an even lower investment rate, and thus a CA surplus.
1. October 2013 at 03:45
“CA surpluses reflect saving/investment imbalances, not (in most cases) “mercantilist” policies, whatever you wrongly think mercantilism might mean.”
Again, the first part of this sentence is a tautology and does not explain anything. CA surpluses do not ´reflect´ saving/investment imbalances, they are by definition saving/investment imbalances.
What are possible causes of these imbalances according to you?
1. October 2013 at 04:09
Saturos, That’s possible, but in my judgment nominal rates would have risen less than the NGDP growth rate increased, as the equilibrium nominal rate in Japan is currently far below zero.
Ben, Yes, growth tends to be faster when inflation is rising than when inflation is falling. The problem is that inflation can’t keep rising forever, without hitting hyperinflation.
kebko, Yeah, I was thinking of mentioning that but I thought the post was already running too long.
Jan, Yes I know it’s a tautology. My point is that the factors that cause the current account imbalance are the same factors that cause changes in savings/investment. In other words they are factors largely unrelated to trade policy, mercantilism, etc. I would add that I focused on savings because I never hear people accusing East Asian governments of trying to hold down domestic investment.
1. October 2013 at 04:55
Thank you.
“My point is that the factors that cause the current account imbalance are the same factors that cause changes in savings/investment. In other words they are factors largely unrelated to trade policy, mercantilism, etc.”
How could they be unrelated? It is only possible (with some exceptions, if the imbalance is located not in the balance of trade but the other components of the CA) to have a saving/investment imbalance if there is an imbalance between exports and imports. So there must be some connection with factors that influence trade, so also with policies that influence trade.
1. October 2013 at 08:16
I’d like to second Jan’s concerns about the tautology. Why can’t trade policy drive savings/investment imbalance as much as vice versa? How did orthodox econ choose which side of the equation was the independent variable?
Current accounts are useless, agreed. Kebko’s investment performance, ashok’s measurement problems, sumner’s location-of-the-purchase-doesn’t-matter are all highly persuasive. But I’m not concerned about the current account, rather real income. Historical evidence on trade policy is decidedly mixed, something not predicted in the slightest by orthodox trade economics. I personally feel this is because techology/skill acquisition, scale/network/cluster effects, organzational/institutional development all have path-dependence.
1. October 2013 at 08:39
To clarify: I agree with Sumner’s KISS prescription. My path-dependence concerns apply to domestic as well as international trade, and pro-savings policy is probably the easiest way to prevent the domestic economy being too distorted by reserve-currency status. I’m just hesitant to entirely dismiss the policy paths taken by E Asia and N Europe.
1. October 2013 at 10:26
Jan, Here is a concrete example. Suppose the government puts large import tariffs into effect. The exchange rate will appreciate. The tariffs will reduce imports and the stronger currency will reduce exports. There is little effect on the CA balance, because tariffs have little effect on desired saving and desired investment.
myb6, I’m not quite sure what you mean by the policy paths of East Asia and Northern Europe. I presume you don’t mean high trade barriers, because HK, Singapore, Sweden, Norway, Switzerland, etc., don’t have big trade barriers. So you must mean pro-saving policies. But then we agree–a pro-saving policy can indeed raise the CA surplus.
1. October 2013 at 13:46
Scott, why couldn’t your example tarriffs change desired S/I? Extreme example: we legislate autarky in a series of capital-intensive industries.
E Asia/N Euro: include exchange-rate mgmt and industrial policy. I was categorizing for brevity, honestly don’t know anything about Sweden/Norway. Fin/Neth/Ger/Switz/HK/Spore/SK/Taiwan/Japan apply. I’m not as convinced as the orthodox that their real incomes (again don’t care about CA) would be as high without the interventions.
1. October 2013 at 14:37
Well, I’m convinced by Baker’s/Krugman’s argument. Their point is simple. Having a large trade deficit implies a gap in domestic demand, which needs to be filled if the economy is going to be at full employment. In the late-1990s, the gap was filled by over investment in equipment and software; in the 2000s, the gap was filled by over investment in housing. Each of those periods of over investment ended badly.
Question for Scott: if the Fed kept NGDP at trend in the previous cycle, what goes up as investment in housing goes down? The only way the economy remains at full employment, I would argue, is if expansionary monetary policy significantly depreciates the dollar and the trade deficit swings quickly into surplus. But Baker’s point is that that process is long. No large country like the US has ever been able to swing a trade deficit into surplus overnight. This is why Baker has recommended a temporary increase in G, and then over time we need the trade deficit to slowly move into surplus. If this process is sped up from NGDPLT, then I’m all for it. But there’s no way NGDPLT makes this process happen overnight.
2. October 2013 at 06:06
My, Yes, trade barriers could change the savings investment ratio. But that’s a second order effect.
You do realize that your list includes the countries at the very top of almost any list of “small government.” Places like Hong Kong, Singapore, Switzerland, etc. These are free trading places. Obviously all countries ave some intervention into the economy but the list of countries you presented tend to have less intervention than almost anywhere else in the world. So it seems a little odd to claim that policy intervention explains their success.
When the governments that do the least amount of in intervention are the most successful, that’s a pretty good indication that intervention is not a net plus for the economy. I’m not sure what sort of interventions you think were helpful in most cases, as the phrase “industrial policy” is very vague. The US does an enormous amount of “industrial policy” if you mean encouraging in some industries and discouraging others. Many other countries do even more than we do. Some, like Hong Kong and the Scandinavian countries do less.
JSeydl, You said;
“Their point is simple. Having a large trade deficit implies a gap in domestic demand, which needs to be filled if the economy is going to be at full employment.”
That’s flat out wrong. I can’t imagine that even Krugman would agree with that statement. A trade deficit has absolutely no implications for aggregate domestic demand. It simply shifts demand from one sector to another.
You said:
“Question for Scott: if the Fed kept NGDP at trend in the previous cycle, what goes up as investment in housing goes down? The only way the economy remains at full employment, I would argue, is if expansionary monetary policy significantly depreciates the dollar and the trade deficit swings quickly into surplus.”
The period from January 2006 to April 2008, shows that you are wrong. Housing construction in America fell in half over that 27 month period and we stayed at full employment. Only when nominal GDP started plunging in late 2008 did unemployment rise sharply, up to 10%. That shows it was falling nominal GDP, not falling housing construction, that was the real problem.
2. October 2013 at 08:10
Baker writes:
“One important component of demand that has been big-time in the negative category in the last 15 years is net exports … The trade deficit in turn led to a big gap in demand that was filled at the end of the 1990s by the stock bubble and in the last decade by the housing bubble. (A trade deficit means that income generated in the United States is being spent in other countries instead of the United States.)”
http://www.cepr.net/index.php/blogs/beat-the-press/krugman-on-bubbles-and-secular-stagnation
Krugman responds by saying:
“I don’t disagree”
http://krugman.blogs.nytimes.com/2013/09/26/trade-and-secular-stagnation/
From January 2006 to April 2008, consumption was still strong, as the personal saving rate remained at historic lows. But once home prices fell so far as to push millions of homeowners into negative equity, they stopped spending.
Components matter. Nominal GDP plunged in 2008 because of the continued drop in housing/CRE construction and because of the drop in consumption.
If nominal GDP keeps rising at trend in 2008, and there isn’t a quick reduction in the trade deficit, and increases in G are off the table, then housing construction and consumption must continue to run at bubble-inflated levels. Why and how on earth could that ever possibly happen?
2. October 2013 at 09:35
JSeydl wrote: “Having a large trade deficit implies a gap in domestic demand, which needs to be filled if the economy is going to be at full employment.”
Scott Sumner replied: “That’s flat out wrong. I can’t imagine that even Krugman would agree with that statement. A trade deficit has absolutely no implications for aggregate domestic demand.”
But in the Krugman article that Sumner linked to, Krugman himself wrote: “we went from persistent small surpluses … to persistent large deficits … This meant that we needed more domestic demand, other things equal, to achieve full employment”
Assuming we ignore whatever is the actual truth of international macro, is there a significant difference in the meaning of what JSeydl summarized, vs. what Krugman actually wrote? I’m missing the distinction.
2. October 2013 at 10:52
Scott, reconsider whether HK (exchange rates, land) and Spore (pretty much everything under the sun) should actually be considered low-intervention. Switzerland is clearly not afraid of managing the franc. CATO/Heritage lists are motivated, reverse-engineered and add no value.
You’ve acquiesced (second order ignorable) to Jan’s concerns. That’s appreciated, and was the main topic, so I’ll shut up now.
2. October 2013 at 11:07
I typed (second order *NOT EQUAL TO* ignorable) with symobls, apparently the site didn’t like that.
3. October 2013 at 06:11
“Jan, Here is a concrete example. Suppose the government puts large import tariffs into effect. The exchange rate will appreciate. The tariffs will reduce imports and the stronger currency will reduce exports. There is little effect on the CA balance, because tariffs have little effect on desired saving and desired investment.”
If the exchange rate is managed, and if the consequences are sterilized, the REER will not appreciate.
But I am more interested in what are the factors that according to you do influence the CA balance. And please do not answer ´the difference between saving and investment. 🙂