You need not be a fan of Donald Trump to say that trade has had a big impact on manufacturing jobs, you really just need to be someone in the reality-based community. Unfortunately, a lot of people who should, and probably do, know better are insisting that trade is not a big deal. The story is that we lost the jobs due to productivity growth, not trade.
There are three points worth making here. The first is a simple logical one, we have a trade deficit of around $500 billion a year, a bit less than 3.0 percent of GDP. This is basically all due to a deficit in manufactured goods (we have a surplus on services). Does anyone believe that the extra imports associated with the trade deficit are not associated with jobs? Can $500 billion worth of manufactured goods be produced without hiring people? (This matters much more in a context where we face secular stagnation, meaning there is not enough overall demand in the economy.)
The second point is that our trade deficit has not always been this large. Our deficits had been around 1.0 percent of GDP through most of the period from the late 1970s until the East Asian crisis in 1997. Following the crisis, the value of the dollar soared and the trade deficit did also. It eventually peaked at almost 6.0 percent of GDP in 2005–2006. (I should be giving the non-oil deficit, but I'm too lazy to look that up just now.)
Anyhow, this explosion in the trade deficit coincided with a sharp decline in manufacturing employment.
Jobs in Manufacturing
Source: Bureau of Labor Statistics.
As can be seen, manufacturing employment stayed close to 17.5 million from the early 1970s to 2000. We had plenty of productivity growth over these three decades, but little net change in manufacturing employment, in spite of cyclical ups and downs. It was declining as a share of total employment, which almost doubled over this period. Then, as the trade deficit explodes, we see manufacturing employment plummet. Note that most of the drop is before the Great Recession in 2008.
The final point is that much of the gains in productivity in the last two decades are illusory. Susan Houseman points out that the bulk of the reported gains in productivity growth are not in industries like autos and steel, but in the computer sector. So a pickup in productivity growth cannot explain the decline in manufacturing employment in most sectors.
I should also add that even the productivity growth we do see is in part due to the trade deficit. When jobs are lost due to import competition, it is generally going to be jobs in the least productive plants. By eliminating low productivity jobs, average productivity will rise even if no plant has actually increased its productivity.
Anyhow, we should not look to combat Donald Trump by following his tendency to ignore reality. Yes, trade has cost manufacturing workers jobs. We can propose different remedies (mine begin with getting the value of the dollar down against other currencies), but let's not deny what is true.