Imagine you're a country that has oil refineries but no crude oil. You need crude oil as a production input for your refineries, and you need refined petroleum products as am input for your cars and trucks and such. Now one day your refineries double their efficiency, and you have way more refined petroleum products than you used to. Productivity has surged—not just in the refining sector, but throughout the economy since everyone likes gasoline and jet fuel. Life is good. But oh no, it turns out that's not actually what happened. Instead of getting more efficient, someone just invented a better oil tanker so importing oil got way cheaper. That's why output of refined petroleum products went up, and that's why there's been plenty of gasoline and jet fuel for everyone. So what? What's the difference? Or what if it's actually not better tankers, but it's just that the country you used to get your oil from had some crazy policy that kept supply low and prices high and now they've changed their tune and the slick stuff is flowing. It doesn't matter, it's all the same.
That, it seems to me, is the basic problem with these arguments that measured productivity growth is off because "really" it was just cheaper imports.
Productivity is not a word I'm thrilled with. It implies something like Stavros is lazy (i.e., unproductive) and Hans is diligent (i.e., productivity) when in reality Greeks work much harder than Germans precisely because they're unproductive. Being productive means you have more dollar value of output per hour worked. If your inputs get cheaper, you're now more productive. Like magic. And good for you. That US firms improved their productivity by using the "technology" of the shipping container rather than upgrading their capital equipment is an interesting fact, but it doesn't mean the increases were fake. The interesting question is what happens moving forward as Chinese wages rise sapping away this trade-driven increase in productivity. I see four options:
One: The political and organizational technology of export-led manufacturing growth is now so well-established that cheap labor jobs will migrate out of China to Vietnam, Bangadesh, and India and life in the US will continue unchanged.
Two: Once wage pressure in China creates incentives to deploy efficiency-increasing new technologies here in US factories, it will turn out that there's plenty of bubble technology (perhaps already in use in high-wage northern europe) ready to be deployed.
Three: Wages and productivity will stay flat in the US manufacturing sector, but the quantity of workers will rise as people transition out of low-wage low-productivity service sector jobs. That will create wage pressure in the service sector and speed the deployment of labor-saving technologies like self-checkout machines, touchscreen quick service order-placing machines, and Amazon's fun new warehouse robots.
Four: We sit around crying into our soup because we've run out of technology.
Scenario four sounds terrible, and we should hope it doesn't come to pass. But if it comes to pass, it should be for some forward-looking reason. The fact that gains in the recent past have come mostly from trade factors rather than technology factors should make it more likely that there's a "stockpile" of labor saving technology ready to be rolled out if demand for domestic labor rises.