The use and misuse of (certain) statistics

It seems, as Shuda Senoy notes, that consumer goods are falling as a share of US imports:   "the proportion of consumption in imports has been falling, i.e., the proportion of industrial inputs has been rising."  (Hat tip Felix Salmon). 

Hmm.  Why might that be.  Oh yeah.  Crude oil is an industrial input -- see Exhibit 8.  Refineries turn an imported industrial input (crude oil) into a consumer good (gasoline).  The US generally doesn't imported the refined product directly.

The price of many industrial inputs -- especially crude -- has gone up recently.  They are not quite as high as they were over the summer, but they are still up significantly relative to where they were a few years back.  And with US production of crude oil falling both absolutely and relative to US demand, the US has had to import more of it over time.  Senoy's analysis just conveniently forgot to note that fact.

Commodity prices are a wee bit higher than they used to be.  Thank China.  Prices of imports of consumer goods haven’t moved much recently.  Thank China, which subsidizes the rest of the world’s consumers by holding down the RMB’s value.  

There is a more general problem with the argument that consumer goods, or impotrs from Asia, have fallen as a percent of US imports..  Imports have been rising as a share of US GDP.  So something can fall as a percent of total imports and still be rising as a share of US GDP.

I haven't looked at the data on imports of consumer goods carefully, so I don't know for sure what has happened.  But the argument that consumer goods imports are falling as a percentage of total imports seems to be a variant of an argument that I have looked at, namely that Asian imports are falling – or at least not rising – as a share of total imports.   Most US imports from Asia are consumer goods.  And I can assure you that US imports from Asia are not falling relative to GDP.

There is a variant of the same argument that floats around – Mickey Levy of the Bank of America made it at the JEC hearing.  It goes as follows:

Consumer goods imports are falling as a share of US imports.    Capital goods imports are rising as a share of imports.   Ergo, imports are not the product of a “consumption boom” – they are the product of a boom in capital spending.

Alas, the broader data doesn’t support that argument.  Consumption has risen as a share of national income (as household savings has fallen).   And business investment remains below its 2000 levels.    Overall investment hasn’t fallen by as much, but that is because of a surge in residential investment.

There is an alternative way of looking at the rising share of capital goods imports in total imports.   The US may not have as big a competitive advantage in capital goods production as it once did.

Capital goods imports also come significantly from Europe and Japan – not China.   So their prices have – I suspect – risen even as China has kept down the price of consumer goods imports.   German, Swiss and Swedish engineering isn’t quite as cheap as Chinese made DVD players stuffed with Japanese, Taiwanese and Korean components.

Bottom line: don’t trust any argument that looks only at imports from Asia as a share of total imports.  Or consumer goods imports as a share of total imports.   It is an easy way of slanting the data.   With imports rising rapidly as a percent of US GDP, something can fall as a share of total import and still be rising relative to GDP.  It is sort of like any argument that looks at China in per capita terms to argue, see, China really isn’t that big a deal.    China’s per capita current account surplus, you see, isn’t that big.   That’s true.   No one matches the per capital current account surplus of Kuwait, the Emirates and Saudi Arabia.  Their surplus is the counterpart to the rise in US imports of “industrial supplies” like mineral oil.  It also is irrelevant for the debate over global adjustment.

One note: I edited this for clarity/ crispness after initially posting it.