Low wages in other countries, especially China, is one of the factors often cited as the cause our enormous trade deficit. If you’ve followed this blog, you’ve heard me often counter that, in fact, low wages have virtually nothing to do with the balance of trade. After all, if low wages are to blame, how does one explain even larger trade deficits (in per capita terms) with high-income nations like Japan and Germany?
But I was going by gut feel more than anything else. For some time now, I’ve been promising a complete study on the subject to find out what the truth might really be. I’ve completed that study and can report to you that both I and the economic experts are wrong about the relationship between wages and balance of trade. There is, in fact, a relationship, but one you’d not expect.
Methodology: First, some explanation of my methodology is in order. It’s simply not possible to compare actual wages from one nation to the next for the full range of manufactured products, so I did the next best thing and utilized per capita “Purchasing Power Parity” (or “PPP”) for each nation. Per capita PPP is roughly the gross domestic product (or GDP) of a given nation divided by its population, and it tends to track pretty closely with average income. For example, U.S. GDP of about $14 trillion, divided by a population of 300 million, yields a figure of about $46,700 per capita – pretty close to the average family income.
On the trade side, I tabulated by SIC code the imports from and exports to each of 163 U.S. trading partner nations. I then identified each manufactured product by SIC code and totaled them to arrive at a balance of trade in manufactured products. By dividing that figure by the population of that nation, I arrived at a per capita balance of trade in manufactured products. It’s important to track only manufactured products, since that’s where jobs are concentrated.
I then constructed “scatter charts” of the data, looking for any relationship that might be evident. By having the computer generate an equation that describes the relationship (if any), along with a coefficient of correlation, it became easier to determine the existence of any relationship. A scatter chart that yields a shotgun-like pattern shows that no relationship exists, while a scatter chart that has the data points more or less organized along a line does show a relationship. It’s important to note that a relationship isn’t necessarily evidence of cause and effect. For example, just because your alarm clark happens to coincide more or less with the rising of the sun doesn’t mean that the sunrise causes your alarm clock to go off or vice versa.
Summary: When all 163 nations are taken as a whole, there is a relationship between PPP and the U.S. balance of trade. As PPP increases from a very low level (near zero), the U.S. balance of trade in manufactured products tends toward a surplus that grows larger as the PPP of its trading partners rises. Lower PPP tends to drive the balance of trade toward zero.
On the one hand, this intuitively seems to be correct. Higher PPP nations produce more and consume more, creating greater opportunities for trade and thus the potential for greater imbalances. Lower PPP nations produce and consume little. Thus, there’s little opportunity for trade, and little chance of running a large surplus or deficit.
However, the fact that higher PPP tends to drive the balance of trade toward a surplus seems counter-intuitive. How can this be, when the U.S. actually has an enormous trade deficit? Closer examination of the data points reveals that 116 of the 163 data points fall on the trade surplus side of the chart, while only 47 fall on the trade deficit side. Regardless of whether one looks at only the trade surplus nations or the trade deficit nations, PPP has the same effect – it tends to amplify an imbalance in trade. Among the trade surplus nations, the surplus tends to grow with rising PPP. The same is true among the trade deficit nations. The deficit tends to grow larger with rising PPP.
There is only one possible explanation for the seeming contradiction between the fact that the U.S. has a trade surplus with 71% of nations but overall has an enormous trade deficit in manufactured goods: the latter category of nations represents a far higher proportion of the world’s population. Those 47 trade deficit nations, while representing only 20% of the earth’s populated land surface area (Antarctica is excluded), represent 65% of the world’s population. The average population density of the 116 nations with whom the U.S. has a surplus of trade in manufactured goods is 58 people per square mile. The average population density of the 47 nations with whom the U.S. has a trade deficit is almost eight times greater at 440 people per square mile.
Analysis: The following scatter chart plots the PPP for each of 163 nations vs. the per capita U.S. balance of trade in manufactured goods (the balance of trade divided by the population of each nation). Tiny island nations were excluded from the analysis for the sake of simplicity and because their economies are almost universally based upon tourism. Tiny city-states like Singapore, Luxembourg, Vatican City, Monaco, etc. are also excluded.
As you can see, the data points hug the zero point at the low end of the PPP scale, but steadily fan out as PPP increases. There is a very strong correlation coefficient (0.57) with the trend line equation describing the relationship. (“0” is no correlation while 1.0 is perfect correlation.) And, as you can see, the trend line lies on the surplus side of the chart. That’s because there are more data points (116 vs. 47) that fall on that side of the chart. But aside from the number of points, the trade deficit side of the chart is virtually a mirror image of the surplus side. Trade deficits tend to increase with rising PPP, just as trade surpluses tend to do.
I then split the data into two groups – trade surplus nations and trade deficit nations – to try to understand this dichotomy. The following is a scatter chart for the trade deficit nations:
For those nations with whom the U.S. runs a trade deficit, there is a definite relationship between PPP and the size of the trade deficit, but it’s exactly the opposite of what you’d expect. The trade deficit in per capita terms tends to grow with rising PPP, and trends toward zero as PPP declines. This flies in the face of what is said by those who blame our trade deficit on low wages. Consider China. In per capita terms, our trade deficit with China is relatively low, falling right in line with other trade deficit nations with similar PPP. But the overall deficit with China is large, not because of low wages but because it’s such an enormous country. Take Thailand as another example. Their PPP is very similar to that of China’s, as is our per capita trade deficit in manufactured products with Thailand. If Thailand were twenty times larger (like China), our trade deficit with them would be the same. Yet no one cares about Thailand. In per capita terms, our trade deficit with wealthy, high-wage countries like Japan, Germany, Switzerland and others is significantly worse than our deficit with China.
On all of these charts, I had to turn off the data labels to make them more legible. But I’ve identified a few of the more notable data points on this chart – nations that are significant trade partners with the U.S. As I scanned down the list of 47 nations that appear on this trade deficit side of the scatter chart, one common characteristic quickly stood out: these are very densely populated nations. In fact, of the 47 trade deficit nations, only four are less densely populated than the U.S.: Armenia, Laos, Finland and Sweden. In addition to those data points I’ve identified on the chart, also included in this group are Bangladesh, El Salvador, Hungary, India, Italy, Malaysia, Mexico, Pakistan, the Philippines, South Korea, Taiwan, the U.K. and Vietnam among others – all far more densely populated than the U.S. This group of 47 nations represents 65% of the world’s population, but only 20% of its land surface area (uninhabited Antarctica excluded from the calculation). The population density of this group of nations as a whole is 440 people per square mile.
Now let’s take a look at the trade surplus nations and see what common characteristics they might share. The following is a scatter chart of those 116 nations with whom the U.S. had a surplus of trade in manufactured goods in 2009.
Once again, we see that as PPP increases, the trade surplus tends to grow. The surplus trends toward zero as PPP declines. The correlation to a linear, upward-trending equation is very strong at 0.61.
I’ve identified some of the more significant trading partners in this group – some very large countries like Canada, Australia and Brazil. This group of 116 nations represents 80% of the world’s land surface area, but only 35% of its population. The population density of these 116 nations taken as a whole is only 58 people per square mile.
Low wages do not cause large trade deficits. Low wages tend to result in small trade imbalances when expressed in per capita terms. High wages tend to result in large trade imbalances – both surpluses and deficits. China is no exception. With a relatively low per capita PPP, the trade deficit with China expressed in per capita terms is relatively low – exactly as the chart would predict. Our trade deficit with China is large because of its sheer size. If China were a cluster of many small nations, our deficit with each would be relatively small.
Population density is clearly the overriding factor in determining whether the U.S. will experience a surplus or deficit in trade with any particular nation.
The reason for the relationship between PPP and balance of trade being similar (but opposite) for surplus vs. deficit nations is a matter of speculation. It is my belief that, among the surplus nations, higher PPP tends to produce a greater surplus of trade for the U.S. because these nations are wealthy in natural resources and they trade those resources for manufactured products. Thus, their wealth is the cause of the U.S. trade surplus. Conversely, among the deficit nations, higher PPP produces a larger trade deficit for the U.S. because these nations use manufacturing for export to generate their wealth. Thus, their wealth is the result of the U.S. trade deficit and a surplus with the rest of the world as well.
Further discussion of the evolution of the trade deficit with China and the implications for U.S. trade policy will be covered in subsequent posts.