Donald Trump constantly complains about our trade deficit, promising us the good life of balanced trade. I’m not sure exactly how he sees ending our trade deficit as the key to prosperity, but let’s play the game. What does the world look like if Trump, through his tariffs, threats, and promises, gets the trade deficit down to zero?

According to Trumpian rhetoric, the big prize is that we would get manufacturing jobs back. These are supposed to be good paying jobs that allow workers to enjoy a decent middle-class standard of living.

Let’s start by asking how many manufacturing jobs we can expect to get back in Donald Trump’s dream world? Our trade deficit last year was $906 billion, a bit more than 3.0 percent of GDP. This is down considerably from a peak of almost 6.0 percent of GDP in the first decade of this century, but still far from zero.

So, if we want to get it to zero, we have to decrease our imports and increase our exports by a total of $906 billion. We do export a lot of things other than manufactured goods. For example, last year we exported more than $165 billion worth of agricultural products. We also export services. Last year, our businesses collected more than $186 billion for exporting financial services, the full tab on service exports was $1,121 billion.

We could reasonably expect that a move to balanced trade would involve an increase in both agricultural exports and exports of services,  but let’s assume that the full adjustment comes through increased exports, or decreased imports, of manufactured goods. Last year our gross output of manufactured goods was $7,210 billion. This means that if we balanced trade by increased production of manufactured goods, we would need to increase output by 12.6 percent.

Last year, we employed an average of 12,817,000 workers in manufacturing. If we got Trump-balanced trade, and we assume that growth in employment was proportional to the increase in output, we would add just over 1.6 million manufacturing jobs. This is a bit more than 1.0 percent of the current total for non-farm employment of 158 million.

It’s worth noting that even this figure is likely high, since the manufacturing jobs associated with new capacity are likely to be more productive than the average, since it would be using the most modern equipment. But let’s go with the 1.0 percent number for now. Would this transform the labor market?

While manufacturing jobs in decades past were good-paying jobs, or at least better than non-manufacturing jobs, this is no longer the case. Manufacturing jobs used to pay more than other jobs because they were more likely to be unionized.

If we go back to 1980, 32.3 percent of manufacturing workers were in unions compared to just 15.0 percent for the rest of the private sector. Currently those numbers are 7.9 percent for manufacturing and 5.9 percent for the rest of the private sector. That is just not that much of a difference.

And it shows in wages. If we compare the average hourly wage for production and non-supervisory workers (a group that comprises 80 percent of employees) in manufacturing and the rest of the labor force, we find that manufacturing workers now do worse than workers in the rest of the private sector.

The average hourly wage for production and non-supervisory workers in manufacturing last year was $33.97. That compares to $35.20 for production and non-supervisory workers in the rest of the private sector. If the point of increasing manufacturing employment by an amount equal to 1.0 percent of employment was supposed to improve workers’ living standards, we’re going the wrong way.

To be clear, this is not a comprehensive comparison of wages in manufacturing and elsewhere in the economy. A full analysis would have to include non-wage compensation, like healthcare insurance and pensions, and also control for education, gender, location and other factors. But when we start out with manufacturing jobs paying 3.5 percent less per hour, the idea that we will substantially improve the lives of working people by eliminating the trade deficit does not look very promising.

We could ask about the prospects of increased unionization rates in manufacturing to restore the historic wage premium. There are two big problems with that story.

The first is that Trump’s sidekick, Elon Musk, is ferociously antiunion. He has even filed a suit calling for the elimination of the National Labor Relations Board, the agency that enforces labor law. It doesn’t seem likely we will see a huge burst of unionization under the Trump-Musk regime.

The other issue is a logical problem. Unionized workers get better pay than non-unionized workers, but this is true in every sector. It would be great to see increased unionization rates in manufacturing, but it would also be great to see higher unionization rates in healthcare, retail, and every other sector of the economy. The point is we want more workers to be represented by unions at their workplace, it doesn’t matter which sector they are in.

The bottom line here is that Trump’s dream of balanced trade is not only a longshot from an economic perspective (as long as foreigners want to invest more here than we want to invest elsewhere, it is literally impossible), it will do nothing towards restoring a golden age for America’s workers. That might be tough news to break to a 78-year-old man suffering from dementia, but it happens to be the truth.