Scott Sumner has an excellent new post at EconLog titled “Conservatives and the mythical third way .” I encourage you to read it.
Shyam, who is one of the commenters on Scott’s post, wrote in response the following:
It seems to me that the angst around the trade deficit is more about bringing back well-paying manufacturing jobs than the overall balance. I think a smarter conservatism would focus on that instead of slapping tariffs on Canadian aluminum.
Shyam’s view about “bringing back well-paying manufacturing jobs” – a view implying that such jobs have been unfortunately lost – is shared by many people. But it is flawed both empirically and theoretically.
Empirical flaw: Adjusting for inflation using the Personal Consumption Expenditures deflator , the real average hourly wage for production and nonsupervisory manufacturing workers  is today (July 2020) at an all-time high. Today’s wage ($22.87 in 2020 dollars) is 10 percent higher than it was ($20.71) when China joined the World Trade Organization in December 2001; 12 percent higher than it was ($20.37) when the U.S. granted to China Permanent Normal Trade Relations (PNTR) status in October 2000; 21 percent higher than it was ($18.83) when NAFTA took effect in January 1994; and 60 percent higher than it was ($14.31) in December 1973 – the year middle America is alleged by many to have reached its economic zenith.
Also, contrary to popular misconception, there’s nothing special about manufacturing employment. On average, manufacturing employment pays less than service-sector employment. While the average hourly wage for production and nonsupervisory manufacturing workers is today $22.87, the average hourly wage for production and nonsupervisory service workers in the private sector  is $24.45.
Theoretical flaw: A job in America – say, one producing widgets – will be ‘destroyed’ by trade only if foreigners supply Americans with widgets at a price lower than is Americans’ cost of producing widgets. And Americans’ cost of producing widgets includes the cost of American labor. This cost, in turn, is determined by the wages that widget-factory workers can earn in alternative employments. The higher the pay in alternative employments, the more likely are imports of widgets to destroy American widget-making jobs. Therefore, jobs ‘destroyed’ by trade are generally ones at which workers’ pay in alternative employments are higher than is their pay in the ‘destroyed’ jobs.
It follows that it would be just as accurate to describe American jobs destroyed by trade as jobs destroyed by improved alternative employment opportunities for American workers. But the latter description does nothing to drum up support for turning more power over to the state.
Protectionists, of course, refuse to accept this conclusion. But their only plausible response is one that relies upon labor-market frictions and imperfections – such as, for example, the reality that many workers acquire firm-specific skills the values of which are much lower in alternative employments.
That such frictions and ‘imperfections’ are commonplace in reality is undeniable. Equally undeniable, though, is the fact that international trade plays no unique role in causing these frictions and ‘imperfections’ to operate to the immediate detriment of some workers. Any change in spending causes demands for some outputs to fall while causing demands for other outputs to rise. Labor-market frictions and ‘imperfections’ are no more likely to result in some workers suffering job or income losses when imports increase than they are to result in such losses when labor-saving technology improves or when consumers simply change their tastes.
The economic case for protectionism will begin to make sense only when protectionists identify in international trade some relevant feature that is absent in purely domestic trade. Yet in the entire, centuries-long discussion of trade policy – and despite a great deal of effort – no such feature has been identified.
No protectionist has ever convincingly explained why jobs lost to imports are unacceptable while jobs lost to simple changes in consumer preferences are acceptable. No protectionist has ever persuasively demonstrated that job losses at home caused by foreign-government subsidies are any different from job losses at home caused by improved efficiency of foreign suppliers. No protectionist has ever compellingly shown that the same government officials who are incapable of intervening productively in purely domestic economic affairs are able to design tariffs and subsidies that increase the country’s economic welfare. No protectionist has ever believably revealed that labor-market frictions and ‘imperfections’ play a larger or more-destructive role when shifts in the demand for labor are caused by international trade than when caused by domestic changes.
It’s not that protectionists haven’t tried to offer explanations of how commerce that crosses political borders differs in essential economic ways from commerce that doesn’t. It’s that all such explanations have failed. When probed, these explanations make no more sense than would explanations of how, say, commerce among men and women differs in essential economic ways from commerce of men only with men and of women only with women. Such differences can easily be built into hypotheticals, but none can credibly be shown to exist in reality.