Blogging for the New York Times, Nobel laureate Paul Krugman recently referred to a Times article on the possible return to the US of manufacturing once outsourced (or “offshored”) to Asia.  In “Rise of the Robots,” Mr. Krugman suggests that much (all?) “reshored” manufacturing will be highly capital-intensive.  Factories will be run by robots, with only a few, highly educated, highly paid human supervisors finding being employed.  Therefore, he concludes, reshoring isn’t the job creation panacea some might think.

I have several comments:

1.  This is not new news.

For over two decades, tech businesses like semiconductor manufacturing have been very highly automated.  Component assembly is increasingly so.  In  the semiconductor case, only a few process engineers watch over $3 billion installations that may generate billions in annual operating profit.  Units of output are tiny and weigh next to nothing, so transport costs aren’t that important.  As a result, tax incentives for building and the rate of tax on corporate profit are the two main determinants of where a plant will be located.

One reason there aren’t more fabs in the US is that income tax rates here are relatively high.

2.  At least some of the current reshoring is either in response to political pressure or to creating a more favorable corporate image in the media.  AAPL, an example cited by the Times, has pledged to invest $100 million to make Mac computers in the US.

Sounds good, doesn’t it?

But $100 million is less than 2% of the company’s annual capital spending budget.  So it’s just a drop in the bucket.  If we pluck a number out of the air and say the investment will generate $1 billion in annual sales, which I think would be an awful lot, that wouldn’t amount to even 1% of the $160 billion or so in sales that AAPL will ring up this year.  Plunk! (=the sound of a drop hitting the bucket)

3.  Stuff that’s very heavy, spoils easily or that faces strong protective barriers against imports, normally must be produced in the same country where it’s sold.

4.  For a brief time I owned shares of Osaka-based manufacturer Sanyo Electric in my portfolios.  I bought it despite its collection of ugly business lines because at the time it was by far the dominant global maker of cellphone batteries.  That business was growing like a weed.  It alone was, in my view, worth far more than the entire market capitalization of the company’s stock.

Because Japan was a high labor cost country, Sanyo had created a highly automated operation.  For each 20,000 units of annual battery production, it had installed machines worth $1 million, which were  watched over by six employees making $50,000 a year each (these are not the real numbers, but that’s not important  for my point).

Business was great–until Chinese competitor BYD emerged.  If the name sounds familiar, it became famous years later when Warren Buffett “discovered” it.  BYD didn’t have the highly educated workers available to it that Sanyo did.  So it couldn’t use the highly automated machinery that its Japanese rival employed.  Instead, it bought simpler, locally made machines that were manned by a larger number of less skilled workers.  To produce equivalent output to Sanyo’s, it installed $500,000 worth of its simpler machines, run by 20 people being paid $7,500 a year each (again not the real numbers–like the real Sanyo figures, those are in notes which remained the property of my employer when I left) .

…and?

The really stunning thing about this example is that:

–BYD made its batteries with both less input of capital cost and less input of labor than Sanyo.  In the textbooks that’s not supposed to happen.  You’re supposed to have to choose between capital-intensive or labor-intensive production methods.  And you’re supposed to be able to compete using either approach, depending on your local labor cost structure.  Not here, though.

A little arithmetic–

Assume that we write the cost of the machinery off in equal installments over ten years.  Then Sanyo’s costs are raw materials + electricity + water, etc. + $100,000 in depreciation + $300,000 in salary.  That’s $20 for each battery + materials…, maybe $25 for each in total.

For BYD, the figures are raw materials etc. + $50,000 +150,000.  That’s $10 + materials… for each battery.  That’s maybe $14 in total.

True, the BYD batteries were probably only 90% as good as the Sanyo ones.  But they cost only a little more than half as much to make.

Lots of medium-tech stuff is like these batteries.  Note, too, that the Chinese salary I quote is less than half the minimum wage in the US.  So the Chinese business model won’t fly here.

5.  As the NYT pointed out in a follow-up, wages in eastern China have more than doubled since I owned Sanyo Electric–meaning that, all other factors being equal, BYD’s labor cost advantage has almost completely eroded.  I presume, but don’t know, that, if so, BYD has shifted production into western China, where wages are still low.

If this business follows the pattern of other industries I’ve followed, like the textiles, at some point the battery industry will shift out of China in search of lower costs.  Machinery will be shipped to another low labor-cost country, India?  Bangladesh?, where production will be resumed.  In fact, BYD may enjoy considerable local tax breaks for doing so.

But wherever the machinery ends up, it’s almost certainly not going to end up in the US.  That would just recreate the company’s situation of too expensive low-skilled labor.  Also, its plants may not be particularly welcome in a country where the firm has no political clout.  More than that, it could be that being Chinese-owned would make it a target of adverse political action.

My take:

This is a big issue, one without a clear solution.  Contrary to Mr. Krugman’s suggestion, I don’t think we’re seeing a reprise of the 19th century, when holders of large amounts of capital had a gigantic (unfair?) edge over people born into families of modest means.  Rather, the 21st century reality is that the market price of unskilled labor in an increasingly global world is under $10,000 a year.

A country can try to protect politically powerful but non-competitive industries, as Mr. Obama has recently done with tires, but that leads to disaster–enriching a small group of political favorites at the expense of everyone else (see my posts).

If all the good manufacturing jobs are robot-driven, then not all highly educated workers will find jobs there. That’s also not a great surprise, since the manufacturing sector in the US has been shrinking for decades.

But, of course, poorly educated workers will be excluded from manufacturing employment entirely.

In the service sector, where all the job growth has been in the US, the field seems to belong to highly educated, computer-savvy entrepreneurs.  Again, the poorly educated need not apply.

I don’t think that in the US a good education is a sufficient condition of personal economic prosperity, but it is a necessary one.