Scott Sumner  

A dangerous myth about trade deficits

Can fiscal transfers solve the... The Joys of Yiddish and Econom...

In the long run trade must balance, in the sense that imports must ultimately be paid for with exports---plus interest. Thus if we buy a billion dollars in laptops from China, we might pay for those goods by exporting $1.5 billion in Boeing jets in the year 2030.

But this true fact leads many economists to falsely assume that a country cannot run measured trade deficits forever. In previous posts I like to cite the case of Australia, which sells condos and vacations to Asians in payment for cars and TVs. The Australian government will report a trade deficit year after year, which never seems to go away. But that's misleading. Overall trade in goods, services and assets may be balanced; it's just that those condos are not counted as exports of "goods".

Tyler Cowen recently quoted Jason Furman and Olivier Blanchard:

Net revenues from border adjustment taxes and subsidies will be positive so long as the United States runs a trade deficit. But if foreign debt is not to explode, trade deficits must eventually be offset by trade surpluses in the future. Net revenues that are positive today will eventually have to turn negative. Indeed, any positive net revenues today must be offset by an equal discounted value of negative net revenues in the future.
This is flat out wrong, and it's not even debatable. The official trade deficit does not measure the increase in net indebtedness; as a result the measured U.S. trade deficit can (and likely will) go on indefinitely.

A common mistake made by famous economists is to confuse statistical measures such as "the trade deficit" or "CPI inflation" with the theoretical concept that economists use in their models. In terms of pure economic theory, the US sale of a LA house to a Chinese investor is just as much an "export" as the sale of a mobile home that is actually shipped overseas. But one is counted as an export and one is not.

You might think that this is mere semantics---who cares how these terms are defined? But this confusion leads to important errors in policy evaluation, such as the incorrect assumption that the proposed border tax/subsidy would be neutral towards trade. That might be true if all imports were taxed and all exports were subsidized. But that's not what's being proposed. When a British tourist visits Disney World in Orlando, the spending is a US service export. But that service export will not be subsidized. When a Chinese buyer purchases a home in California the "export" of the home will not be subsidized. The real strength of the US economy is in the export of services and assets, which are largely ignored under this system.

It is one of the most profound consequences of China's growing wealth: Chinese investment in U.S. real estate has exploded, particularly in California and New York. Chinese nationals are now the biggest foreign buyers of American homes, purchasing at least $93 billion worth of home in the past five years, including $28.6 billion in 2015 alone. Commercial property purchases have surged as well, to $8.5 billion last year, a 15-fold increase from 2010. And, at nearly $208 billion, China is the biggest foreign holder of U.S.-government-backed residential mortgage bonds.
Screen Shot 2017-03-11 at 12.25.36 PM.png

This error leads some economists to confuse a theoretically neutral border tax that applies to all imports and exports, with the one actually being proposed, which would tax imports far more than it subsidized exports, and hence would end up being at least somewhat protectionist.

This does not mean that the tax reform proposal that the GOP is putting together is necessarily a bad idea---the proposal does have a number of good features, such as expensing investment and equal treatment of debt and equity. But we need to be aware of what is being proposed. As far as I can tell the plan is at least slightly protectionist.

COMMENTS (31 to date)
bill writes:

Their basic proposition also ignores that if NGDP is growing by X% per year, we can easily run deficits forever as long as those deficits are less than X% of NGDP. Q3 2016 current account deficit was about 2.5%, so pretty sustainable.

James writes:

This post is a great dismantling of a technical error but it doesn't address the root cause of the mistake: Some people are skeptical of cross-border economic activity.

If you explain that the Chinese will sell products to Americans and use the proceeds to buy American real estate, economic nationalists will say "How terrible! First we spend all of our money buying things from China that we used to buy from America and now land becomes unaffordable as China buys up all the real estate."

Unfortunatley, this kind of mistake is so prevalent that both major parties have no choice to but to pander to it.

In the past, economic nationalists saw exports as a benefit and imports as a cost, so they were willing to accept imports as a necessary condition to create exports. More recently, they have come to realize that exports are actually a cost. I hope they will soon see that imports are a benefit.

Gordon writes:

There was a good post over on the FRED blog about this topic as well. I especially liked it because it had a chart that showed the capital account in relation to the current account.

Sam writes:

The solution to part of the problem re: Disneyworld, would be to allow foreigners to get a tax rebate for anything they buy while in America.

It's a bit tricky to manage, but I remember when I visited Iceland that I could bring my receipts to a bank at the airport and they'd give me the tax back. Wouldn't be infeasible to set that up in the US for a DBCFT.

The house problem is why in most countries Housing Purchases are VAT exempt. Although less of a case to do so with rent.

Both the eurstat and the American statistics do count tourism revenues as exports. lists tourism as 11% of US exports.

Exports from tourism will, of course, be rougher estimates than exports from heavy industry: if a friend from abroad visits and we go out for dinner together, is there anyone checking who paid for which share of the bill? Travel shopping probably flies under the radar: whenever I'm stateside with some free time, I'll end up clothes shopping to take advantage of the large price differential compared to Europe. Will my spending at the mall be counted as tourism? I assume you can get good numbers if someone is paying with a foreign credit/debit card which results in an international transfer, but getting a bundle of cash before international travel is still customary (recently my bank changed my debit card to one which does not work internationally [outside of the EU]! Complaining didn't help: they consider this an upgrade).

Tourism is now a very large part of the world's GDP. This trade organization estimates ~10%: This sounds ridiculously high until you think about it. At least in Europe, it's not uncommon that people spend 10% of their yearly paychecks in traveling. If you then count some local travel (a day trip to a nature park or some business travel), then it becomes very plausible.

Btw, if tourism is 11% of US exports, shouldn't Trump be more careful about the economic impact of messing up the borders on exports? It's not like the US has a positive image of foreign-friendliness at the border as it is. I think the US is underrated as a tourist destination (it gets a lot of business travel, but less leisure travel than it should), but few people visit beyond NYC & Disneyland.

mbka writes:

Another aspect is the current software-as-a-service business model in IT. I wonder if and where the licensing fees for Microsoft Windows and the like are accounted for in the trade statistics. These may well be considered rents and not be counted at all. And unlike condo purchases by foreigners, service exports such as these should look harmless to even the most enraged economic nationalist.

Scott Sumner writes:

Bill, But if deficits were measured properly that would not be possible. True trade deficits must be offset by future surpluses, otherwise countries would be giving us stuff for free.

James, Yes, and what's so funny is that they'd be happy if the Chinese bought mobile homes and had them shipped to China.

Gordon, Thanks for the link.

Sam Yes, that is an option. I know that I don't bother doing that when I visit Europe.

Luis, Good points. Yes, services are theoretically counted as exports, but as you point out they are often unmeasured.

mbka, Good point.

John Thacker writes:

So that argument does equally apply to any VAT (all of which are border adjusted) and any consumption tax. Some countries allow VAT refunds to tourists for goods (generally a hassle only worth doing on expensive durable goods) but not on services or things consumed while a tourist. Does that make VATs a little bit protectionist?

The border adjustment transforms a cash flow tax into a consumption tax on domestic consumption, which does include consumption done by tourists in the country. Do you now oppose the progressive consumption tax and VATs that you used to support, or do want them to exempt and refund tourist spending on services?

AntiSchiff writes:

Dr. Sumner,

This is a very good post and so needed today. I hope you will write an editorial on this topic to appear in a major newspaper. This perspective needs exposure.

I also hope you will start pushing for a wage subsidy in the US, given the continued increases of minimum wages in various states and the political situation we find ourselves in. A wage subsidy is no magic bullet, but I think it could lower the political temperature in the country, if enacted.

Capt. J Parker writes:

If the US ran large trade deficits every year and the corresponding capital surplus resulted in increasingly larger portions of US land, natural resources and US based means of production being owned by foreigners, why would that not be something to be concerned about?

bill writes:

I think it would be possible. Example: If the US ran current account deficits of 2% of NGDP each year and capital account surpluses of 2% of NGDP each year, both forever, and NGDP grew at 4%, then the cumulative deficits (I'm not sure what that is called) would stabilize at 52% of NGDP. Current account deficits of 3% of NGDP (assuming 4% NGDP growth) would stabilize at 78%.

Anon writes:

[Comment removed for supplying false email address. Email the to request restoring this comment and your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Scott Sumner writes:

John, It seems to me that it would be a little bit protectionist in any country with a large trade deficit, so that would not apply to Japan and the eurozone, for instance.

I'd still support a consumption tax/VAT because it would be far less bad that the current income tax system. I might even support this corporate tax reform despite the border tax, as it does other things that are positive.

Thanks Antischiff.

Bill, But I think you are forgetting the effect of interest on that debt.

John Thacker writes:


Without the border adjustment the cash flow tax is neither a domestic consumption tax nor a VAT. It requires the border adjustment to get much closer to being one (though perhaps with the exceptions you note), which is why all VATs are border adjusted. A cash flow tax, applied territorial on domestic business activities, taxes investment as well. In order to tax domestic consumption, things not consumed in country must be refunded at that rate, and imports consumed in this country but otherwise exempt from the cash flow tax must be taxed at the border at the same rate. This transforms it into a tax on domestically located consumption with no tax on investment, regardless of the rate.

Now, it is true that durable goods bought by tourists while in country and taken home are certainly not domestic consumption. This is why most VATs refund those items (though sometimes the process is a hassle.) Goods and services bought by tourists while visiting but consumed in this country are arguably domestic consumption (from a geographical perspective) but arguably not, as you point out with strong validity. Most VATs do not include such in their border adjustment, but they could.

Your argument, as far as I can tell, is that the border adjustment doesn't adjust enough, but that still seems to be an argument that border adjustment is better than not adjusting. After all, it brings things much closer to a domestic consumption tax than not doing so. If you claim that you favor the reform but oppose adjustment, I can't view that any differently than saying that you would prefer not to have a consumption tax. I do think that the concerns about a quick transition are understandable.

John Thacker writes:

I certainly understand the complaint that the border adjustment misses exports located geographically domestically by treating them as domestic consumption even though they are not, being done by tourists. That is in no way an argument for adjusting less, though, rather it is an argument for adjusting more. A corporate income tax without border adjustment certainly taxes both those exports as well as more traditional exports, while exempting imports made by non US companies (and, in complicated ways, sometimes by US companies as long as the profits are left overseas.)

It does take some effort for me to wrap my head around the idea that a tax is protectionist for not taking exports enough, especially when you concede at the same time that it taxes all domestic consumption the same, whether domestically produced or foreign produced. The problem is treating some exports performed in a domestic location as domestic consumption. That may be a distortion from the ideal of pure domestic consumption tax, but "protectionism" doesn't quite seem the right word for taxing exports too much due to miscategorization, especially when the current system is even worse.

Bill Woolsey writes:

Capt. J. Parker

We can add to the capital stock. That foreigners are buying condos doesn't prohibit us from building enough so that Americans can have one too.

Your assumption would be equivalent to saying that if we keep on producing more and more goods for export in order to pay for imports, then won't that be a problems because we aren't producing anything for home consumption?

robc writes:

I made the same point back in the 90s when everyone was considered about our trade imbalance with Japan.

It was balanced, we were just exporting Manhattan skyscrapers and Hawaiian golf courses.

bill writes:

My understanding is that interest expense adds to the current account deficit itself (so it's netted out of the 2% deficit). In reality, there are large capital accounts in both directions and so far the US earnings on its foreign holdings exceed what we pay foreigners for their capital, including interest. Of course, that could change.

Jerry Brown writes:

Bill Woolsey, replace 'condos' with 'land' and maybe a problem becomes more important. We are not so good at building more land especially in prime locations.

Jim writes:

Thank you for the post.

But doesn´t this just make the argument that many people have claimed that in perpetuity china will just own everything in the US? How would this be good? Isn´t that the Trump teams argument?


Michael Rulle writes:

A short thought experiment.

Capital Surplus must equal Current Account deficit. That is an accounting identity, but has no necessary connection to a positive economic outcome.

Assume a country,A, with no employed people. But country Z wants to increase its production and sales figures. Z sells a variety of consumer goods to A. In order to do so, Z obviously needs to lend money to A.

The end result is Z owns finanicial assets and A owns goods. The cynic might say, that is a great deal for A. (Of course, we do not need foreign trading partners to accomplish the same outcome as long as some people work in our country. This drove the dot com boom.)

My point is it matters what kind of capital surplus we have, not merely that accounting requires the capital surplus equals the current account deficit. I think that our seemingly massive forward domestic deficit, which is seemingly not sustainable (maybe ponzis can last forever per Samuelson) is being funded by Govt debt and Govt guaranteed debt. Much of it foreign.

Yet our income from foreign investments equals our payments to foreign investors. But our opportunity cost still seems high to me.

I would like more savings, but we are not inventivized to save.

Jerry Brown writes:

When you get into accounting identities you get into territory where MMT economists are very comfortable. One of their favorite identities is (S-I)=(G-T)+(X-M+FNI) where (X-M+FNI) is the current account balance (CAD). (S-I) is said to represent the private sector financial balance.

This can be rewritten as [(S-I)-CAD]=(G-T). The term on the left represents the financial balance of everything not government while the term on the right is of the government financial balance and will always be equal and opposite to the left hand balance.

At this point a reasonable assumption is made that the private domestic sector can not continuously run in deficit for the simple reason that it will go bankrupt. Assuming that is true, a continuous current account deficit will entail an offsetting government fiscal deficit of almost equal size depending on just how much the domestic private sector is willing to dis-save.

The point of all this is that while Sumner is correct that the US balance of trade can and probably will remain negative indefinitely, the Federal Government is also going to be running large offsetting budget deficits indefinitely at the same time, and if you are worried about the budget deficits and the national debt well then you have to fix the current account deficit at the same time. There is no way around that part.

Jerry Brown writes:

Mostly my previous comment was more or less copied from the Bill Mitchell blog. Check out the answer to quiz question #1 if reference is needed.

Capt. J Parker writes:

@Bill Woolsey,
You said:

We can add to the capital stock. That foreigners are buying condos doesn't prohibit us from building enough so that Americans can have one too.

Yes, agree, and the fact that the US has been running trade deficits for years suggests that perhaps the US is pretty good at growing our capital stock so there hasn't been a downside to selling some of it every year to finance imports. My point, contrary to what James is suggesting, is that there is no law of economics that says there can never be a problem with selling capital to foreigners to finance imports. I think Jerry Brown's thought experiment makes this point pretty well. If we were selling off our endowment of land to foreigners to finance imports then it's not hard to imagine negative long term consequences.

Let's say you owned a business and you were growing it some every year and your hope was that your business would provide a livelihood for your kids, grandkids and so on. But, you also desired a full and rewarding life for yourself. What strategy should you follow? a)Retain full ownership, plow a portion of the earnings back into the business and use the other portion for your own consumption or b)Augment both your own consumption and your business's capital stock by selling a portion of the company every year to outsiders.
Strategy b probably gives you more consumption in the short run. It might give you a more consumption in the long run if the extra capitalization makes the business grow much faster than using strategy a. But, it would be a mistake to assume that there is no possibility that strategy b might lead to less consumption for you and your offspring in the long run.

Prakash writes:

Isn't it a straight forward implication of the efficient markets hypothesis that one cannot specialise in creating assets?

How does Prof. Sumner reconcile the same?

John Thacker writes:

If we're considering distortions compared to an economy with no taxation (foreign or domestic), then I'm not certain that it's a distortion at all to fail to subsidize these exports-- at least so long as the tourists' home nation doesn't tax their foreign consumption while traveling but does tax their domestic consumption. In fact, it may be less of a distortion to tax them.

One of the possibly distortionary effects of refunding durable goods bought by tourists is that tourists, unlike corporations, rarely pay VAT (in the form of a border adjustment) when bringing those goods back home, even after exempted from VAT in the visiting country. Arguably in that case the VAT exemption/refund would be an export subsidy; it would be more neutral if VAT were paid to the home country on the other end.

From the perspective of taxing consumption, the current tax is wildly distortionary; from the perspective of taxing worldwide corporate income, it is less distortionary but full of loopholes (that would easier to plug, but not perfect, by going after consumption.) Part of the problem, as the Tax Policy Center has put it, is that you can break things down into four categories, based on whether goods (or services) are produced domestically or foreign, and consumed domestically or foreign. No government taxes things both produced and consumed by foreigners. An income tax (ideally) taxes things produced domestically whether consumed domestically or foreign. A consumption tax (ideally) taxes things consumed domestically whether produced domestically or foreign.

I don't see the argument for declaring one sort inherently more distortionary than another (especially post any currency adjustments). As for as distorting trade goes, having a very different system from trade partners can be the most distortionary. If all countries are using similar systems then there is less distortion; if all countries tax consumption then all forms of trade are taxes similarly, and that can also be true if all countries tax income only. Most countries use a mix of VAT and corporate income tax, and so the USA is an outlier and distorts trade by relying so heavily on corporate income tax. Going entirely to consumption, as the DBCFT does, would create distortion in the other direction, but it would at least match the current world trend of shifting from corporate income tax to VAT and still be a smaller distortion of trade.

Scott Sumner writes:

John, I don't think you are disagreeing with my claim that the plan is a "little bit" protectionist, nor my comment that the current system also has flaws, which might be even worse.

robc, Good point.

Bill, Here's it's useful to go back to the "no free lunch" concept. Ultimately, we must pat for everything we import.

Jim, People had that same worry about Japan back in the 1990s, and those fears did not pan out. There are many other factors to consider, such as the immigration from China to America, which turns foreign owned assets into domestic owned assets.

Jerry, I think you are missing the point. What you say is true of the actual current account deficit, i.e. the theoretically appropriate measure. But measured CA deficits are very different, and can go on forever.

Prakash, I don't know what that means.

Jerry Brown writes:

Scott Sumner, fair enough- my longer comment went off on a tangent from what you were talking about. Sorry about that.

bill writes:

I know for sure that you know more about this than I do. My sense is that even a single year of a current account deficit or surplus is not a free lunch, because it's balanced each year by a capital account surplus or deficit. I think the second paragraph of this post explains how a country can run trade deficits for a long long time.

Neal Reynolds writes:

"If we were selling off our endowment of land to foreigners to finance imports then it's not hard to imagine negative long term consequences."

-- Capt. J. Parker

Oh Captain, My Captain! Thank you very much for stating this -- I've for years been pointing out the same thing in various comment blogs to no avail. Glad there is at least ONE person out there who sees the truth!

Steven Englander writes:

This is a very interesting debate. You take a long-run equilibrium condition and make it a driver of short-term FX and other asset markets without specifying how the equilibrium is reached. So the equilibrium condition becomes the causal driver, rather than the result of market and economic adjustments. It matters because all of these analyses are pretty much assuming that we are in equilibrium now and the BAT is the disequilibrium shock. But we may be in disequilibrium now and the BAT shock is part of the mechanism that enforces the exports = imports condition. I would not necessarily endorse this view but you can take the same approach and say that if our border adjustment taxed exports and subsidized imports, nothing much would happen because the equilibrium tells us that exports and imports will balance.

Comments for this entry have been closed
Return to top