Neil Irwin had an interesting article this week in the New York Times about trade deficits. The piece was triggered by comments on the dangers of deficits by both Donald Trump and Bernie Sanders.

Trump has been arguing that the fact that the United States buys more goods from other countries than it sells overseas means that the US has become a loser while others, like China and Mexico, have been winners. Sanders has been claiming that trade deficits prove that trade agreements are bad for American workers.

Irwin nicely argues that deficits, on their own, are neither automatically good nor bad. Getting rid of a deficit, moreover, also involves important tradeoffs and consequences that may be much worse.

New analysis prepared by Moody's Analytics, at the request of the Washington Post, shows that Trump's plans to raise tariffs on China and Mexico would drive both countries into recession as both countries would likely retaliate. The US would also fall into recession, with millions of American workers at risk of new job losses.

There are, of course, additional problems with raising tariffs, such as getting Congress to agree to raise tariffs (since the President cannot do this alone); as well as the minor obstacle of such tariff increases being a clear violation of American commitments at WTO. Nor, of course, would higher tariffs somehow bring jobs back to the United States. In fact, the net result of jacking up tariffs as Trump suggests would be mostly to raise prices for consumers least likely to be able to afford them and to cause a lot of needless economic damage. It would be an "own goal" of epic proportions.

The stakes would be high then, on a proper assessment of whether or not the country actually has a trade deficit and whether or not such a deficit matters.

Depending on what statistics are used, many people often focus only on merchandise trade figures. This ignores trade in services completely. For a country like the United States, services trade is usually in strongly in surplus.

Services are also embedded in manufactured goods. For many supply chains, services make up 30-70 percent of the total value of the final product. These services could be logistics, retail, research and development, cleaning, legal services, customer support and so forth.

Many of these services may not show up at all in current trade statistics.

Officials at the OECD and the WTO, as well as their counterparts in many countries, have been hard at work to revise the way that trade is recorded, partly to better reflect the importance of services in manufacturing.

Better measurement is also critically important to handle some of the claims about trade “deficits.” Trade is no longer so much about “exporting” a finished item from one country into another where it is recorded as an “import.” Instead, in a world of increasingly complex supply or value chains, items are rarely created entirely in any one country nor do they move across a border only once. Instead, they may include hundreds or thousands of parts from multiple locations managed by globally or regionally dispersed teams with the final items crossing borders many, many times before reaching the final customer.

The classic example, used everywhere now, is the iPhone. Traditional statistics may record the full price of the phone as an export from China to the United States. Hence this would show the full amount of the phone as an American import from China and another example of “how China is eating our lunch,” as Trump would say.

But a careful unbundling of the phone shows clearly that this simple story is just not true. The bulk of the value of the phone remains in the United States, contained in the profit, research and development, intellectual property and marketing owned by Apple. The rest is distributed elsewhere, including to Korean companies that make critical, high-value components for the phone and Taiwanese companies that manage most of the manufacturing. Very little of the actual value of the phone resides with Chinese companies.

Depending on which Apple product is being described, the figures basically show that of a $300 product, about $3-6 should be properly attributed to China. The rest goes somewhere else and most is American. Hence what looks like a -$300 deficit for the US and +$300 surplus for China, is actually a sizable surplus for the US and a very small surplus for China.

It’s not just high-tech products that fit this description. Even fairly simple food products are made up of items shipped around the globe. The Financial Times has just run a fascinating (if deeply disturbing) piece. The point of the article is about food fraud, but highlighted at the bottom of the article is a bit about the global travel of fish, caught in waters off Scotland, shipped to China for filleting, sent to South Korea for storage in massive freezers, and eventually sold back in Europe.

Think a bit about the statistics for such fish sticks (hopefully not adulterated with unmentionable items along the way). The logistics, filleting, warehousing, shipping, distribution and retail sales are all services. Tracing the supply and value chains attached to a fish stick would likely also result in shifting patterns of imports and exports from all of the countries involved as well, moving some from “deficit” to “surplus” for fish sticks, for fish, for trade in goods or maybe even for overall trade.

It is a rare company that can obtain all the raw materials, components, parts, and services it needs from purely domestic sources. In order to create products, most companies have to import. This is true even for firms that will export components, parts or final products.

Changing the statistics dramatically alters the trade “deficit” story. It does not eliminate a “deficit” somewhere, of course, but it does shift the magnitude of a deficit with some countries.

Of course, it is also important to remember that country accounts do not work like personal bank accounts. It’s normal for people to assume that negative account numbers must mean bad things. After all, after balancing a checkbook, a negative number is not normally a good sign.

But trade figures do not work the same way. Imports are not automatically a bad deal.

In short, whenever politicians promise solutions to “fix” problems related to trade deficits, citizens ought to be extremely concerned. Most likely, the remedy will be much, much worse than anticipated.