Do imports account for the decline of US production?

One of the components of GDP is "net exports". As you know, net exports went seriously negative since about 1980:

Graph #1: Balance on Current Account
See also this comparison at FRED

Stable until 1970, at least at this scale, and no evidence of negativity until after 1980. Interesting, though: the first big downtrend, in the 1980s, slows suddenly in 1986 and turns upward by 1988. That was also when the US saw a big slowdown in the growth of private debt. Two good things at the same time, and probably no coincidence.

But how big is that trade imbalance, anyway? What if we could zero it out? What would GDP look like if we don't reduce it for net exports? I mean, we can add the value of those imports to GDP, as if we produced the stuff here. It will bring the GDP number up -- but how much? What would Real GDP growth look like, then? Would it still be trending down? In other words, do imports really account for the decline of US production?

Here's how the Balance on Current Account looks in comparison to GDP:

Graph #2:Balance on Current Account (blue) and GDP (red)

At this scale the Balance on Current Account looks stable to the latter 1990s. And even where it's big, it is small in comparison to GDP. But it was a fair percentage, before the crisis there.

If we subtract our growing trade deficit from GDP we get a number that's larger than GDP by the amount of the trade deficit. Graph #3:

Graph #3: GDP (red) and GDP with the Trade Deficit Subtracted (blue)

The red line is GDP. The blue line is higher, as if we produced the imports here.

It's higher, yes, but not a lot higher.

These things always surprise me.

So now, if we look at the growth rates of those two lines...

Graph #4:Growth Rate Comparison

... there is hardly any difference at all between them. If domestic production is lagging, it's not because of the balance of trade. Oh, we import a lot, yeah. Too much. But our trade imbalance is another of those things that's a result of the problem -- a contributing consequence, maybe, but not the cause.

And if I show Graph #3 after taking the natural log of the values, the two lines follow a similar path. Both show a slowdown in the early 1980s:

Graph #5:Log Scale Comparison

Nothing if not similar.


Two more graphs to look at, and we're done. I want to strip out the inflation, and compare the two versions of "real" GDP:

Graph #6: "Real" Values, Log Scale Comparison

No. The trade deficit doesn't make a big dent in GDP.

Graph #7: "Real" Values, Growth Rate Comparison

Nope. Can't say GDP declined because of the trade deficit. I'd say this instead: The thing that's causing the decline of GDP is also responsible for the trade deficit.