Wednesday, December 1, 2010

The Coming Post-Job Economy

According to The St. Louis Fed, American manufacturing employment, as a percentage of total employment, has been falling for more than sixty years.


Because the American economy grew so rapidly after WWII, manufacturing employment here grew, in absolute numbers, into the 1970’s. In more recent years, however, the decline has been both absolute and relative, and there has been a tremendous increase in the number of jobs created in low-wage countries such as Mexico, China, and India. There is a natural tendency to attribute this shift to competition from low-wage workers. Some of it is, but the problem is not that simple.

American jobs are lost to the cheapest alternative to American labor. Sometimes, that alternative is labor somewhere else; sometimes it’s machinery right here. Domestic labor, foreign labor, and automation compete for the opportunity to produce things for American consumption. Unless American labor is the cheapest alternative, the jobs will be lost to one of the alternatives or the other. “Blaming” only one of those competitors sets us on the wrong path to dealing with our employment woes. If we could eliminate all competition from low wages, we would still have to compete with the machines.

The rise of automation can be seen in the way the loss of manufacturing jobs has affected manufacturing output. Over the past twenty-five years, American manufacturing production actually increased, at least until the 2008 recession hit.

Graph: Manufacturing Sector: Output


The growth has not been spectacular, but clearly, the volume of things our people make has not decreased nearly as fast as the number of our people employed in making them. Free-traders argue from these graphs that automation accounts for most of our job losses (since the volume of outputs has grown) and automation is a good thing because it frees up labor to do other things. I don’t buy either the inference or the platitude.

Clearly, automation has not displaced workers in every industry that has lost jobs. If automation accounted for all of the job loss in America, we’d still be making the same things we were making before, but with fewer workers. Instead, many of those things are being made by cheaper workers somewhere else, and we are making different things, things that rely so heavily on capital and technology to produce that labor is not a factor in their cost. Automation has not displaced workers; it has simply filled the trade vacuum created by poorer countries’ comparative advantage in labor services.

I do think, though, that many of the jobs lost to cheap labor would have been lost to automation if the cheap labor had not been available. If cheaper people had not come along to make the things we used to make, cheaper machines might well have done so. If that’s the case, the problem we face goes well beyond leveling the international playing field.

I do not mean to minimize the effect of cheap foreign labor on our economy. Globalization has eliminated natural barriers between labor pools, creating a trade in labor on a scale probably not seen since the days of slavery.

Labor is a Special Commodity

In David Ricardo’s classic example of comparative advantage, the English made cloth and the Portuguese made wine; in effect, they were trading English rainfall for Portuguese sunshine, capturing economies of scale with respect to each of those resources. Optimizing the use of these non-labor-resources enabled labor in each trading partner to flow to the local industry where it could add the most value. It may have taken a long time for the benefits of trade to “trickle down” to the average worker, but trade created jobs in both countries.

In contrast, much modern trade, especially US trade with Asia, is built on sharing the latter’s human resources, i.e., cheap foreign labor. All of the ordinary implications of trade still hold: prices are lower everywhere, and (first order) aggregate wealth is greater than without the trade. But whereas the cloth/wine trade created jobs in both partners’ export sectors, the man-made/machine-made trade creates few jobs in the high-wage country’s export industries. That sector, by hypothesis, requires fewer people, which means that former manufacturing workers have to find work making something that the high-wage country neither imports nor exports. That’d be sales, personal services and construction labor.

Free traders argue that we are endlessly creative, that we will think of things to pay each other to do that can only be done locally. All trade-based dislocations, they say, require adjustments. In making this argument, they offer an amazing example of chutzpah, citing as a reason for optimism the statistic that we have the most productive workers in the world. Look, they say, at how our productivity has grown:

But this chart can be inferred from the earlier ones: if we are making more things by employing fewer people, the ones still working in manufacturing must be very productive. This type of productivity says nothing about the displaced workers themselves other than that they were less productive than the machines and foreigners who replaced them. Someone has to push the buttons at the robotic factory. That person’s “productivity” does not reflect the skills of the American workforce. All it shows is that, consistent with Ricardo’s observation, the trading partner with a comparative advantage in capital (that’d be us) will trade capital-intensive goods for labor-intensive goods, and vice versa.

Such trade may free up people in the capital-intensive economy to do local work, but it does not create jobs for them in export industries. Every other form of trade creates jobs, but trade for labor does not. That’s why labor is a special commodity and why trade creates the current unemployment problem instead of solving it.

Why all else fails

The capital intensity of our manufacturing sector explains why traditional (Keynesian) efforts to stimulate the economy cannot reduce unemployment here.

Good, old fashioned fiscal stimulus does nothing for our local economy because stimulus spending is about the multiplier – the tendency of spending to create business for suppliers of suppliers of suppliers ad infinitum. Most of those suppliers are assumed to employ people in the economy where the money is spent. Today, however, the supply chain always leads abroad. When we say that something will have an effect on “the economy,” the economy we talking about is global. If we’re importing a lot of labor-intensive goods, we can expect our stimulus to create jobs where those imports are made.

Take a look at our trade with China for 2010, and see how well our stimulus is “working” – for China:

All figures are in millions of U.S. dollars on a nominal basis, not seasonally adjusted unless otherwise specified.

January 2010 6,888.8 25,185.1 -18,296.3
February 2010 6,855.1 23,363.8 -16,508.8
March 2010 7,403.6 24,300.2 -16,896.6
April 2010 6,591.2 25,905.7 -19,314.5
May 2010 6,752.7 29,036.8 -22,284.1
June 2010 6,715.0 32,866.5 -26,151.5
July 2010 7,344.7 33,260.0 -25,915.3
August 2010 7,253.5 35,288.5 -28,035.0
September 2010 7,168.2 34,999.2 -27,830.9
TOTAL 62,972.8 264,205.9 -201,233.1

  • 'TOTAL' may not add due to rounding.
  • Table reflects only those months for which there was trade.
  • SOURCE: U.S. Census Bureau, Foreign Trade Division, Data Dissemination Branch, Washington, D.C. 20233

Spending in the US by anyone – Americans, American governments, or the Chinese sovereign wealth fund – stimulates the economy in China (and, of course, in OPEC’s oilfields, but this post is long enough without dragging them into it). The Chinese recycle our money by lending it to us at low interest rates, but when we spend it here, we spend too much of it on imports, if not on the first go-round, then when the multiplier kicks in and the guy with the new job buys a flat screen TV.

Balancing trade – important as that may be – won’t solve our unemployment problem either. Even if our exports kept pace with our imports, the hours worked would not. By definition, a capital intensive trading partner needs fewer workers than a labor-intensive one to produce the same value of outputs.

By all means, we should try to balance our trade and to end Chinese mercantilist currency manipulation. The trade deficit transfers national wealth, which has strategic implications, and it distorts our capital markets with disastrous consequences. But doing so will not bring back a level of manufacturing employment in the US anything like that of twenty years ago, and any politician who promises that it will do so is doomed to disappoint.

What about infrastructure projects? In boom periods, labor isn’t available to work on the infrastructure, so there is plenty to do when things get tough. But infrastructure work alone is not enough. It might be enough if the multiplier effect were great, but today, it isn’t. Like reducing our trade deficit, we need to enhance our infrastructure, and we should have no qualms about borrowing to do it, but it will only create the direct jobs required to do the actual work. The multiplier will have too much of its effect abroad.

Historically, new industries have sprung up to take advantage of abundant labor. But that was when there was no hole in the employment bucket. If an American had a bright idea, he hired Americans to execute it. Now, we invent it here and build it there. It is in the nature of globalized trade that if an invention generates jobs, it generates them where the labor is cheap. Exceptions exist for things that can only be done locally, but we cannot build an economy on outliers. Every logical avenue ends at the capital intensity of domestic manufacturing. There may always be something new for actual humans to make, but there is no reason to believe under current conditions that those people will be Americans.

So, for now, the jobs that remain here – jobs outside the manufacturing sector, which has either been outsourced our automated – are in the personal service sector, where productivity is much lower than in manufacturing. Those jobs pay lower wages than US manufacturing jobs, a problem exacerbated by the oversupply of people available to do them. Training may put people back to work at these jobs, but it won’t send them and their families to Disney World.

And so we come to the place where classical economic theory meets political reality. The theory says that in the long run, we will adjust. We will find something to do. How long that will take, and with what alterations in our way of life are questions to which the theory is wholly indifferent. We cannot logically argue against the eventual outcome if the system is given enough time to sort things out. But Keynes wasn’t kidding when he said that in the long run we are all dead. There is no guaranty that we will, or should, have the patience to let the system sort things out. The system will not sort things out if social unrest destroys our democracy before it does.

When all else fails…

Having ruled out the usual job-creating suspects, we need to think outside the box. I think that means making our economic life as capital intensive as our manufacturing, to require only as much work from people as the economy has good work to offer, and to find away to share the benefits of the work others and machines are willing to do.

The recommendation that we not have to work all our lives to live all our lives does not imply that talents and energy should be wasted. But it does recognize that to get back to anything like full employment, we will have to put some very square pegs into some very round holes. Specifically, would-be manufacturing workers will have to become healthcare workers. There are other things that can only be done locally – construction and oil-drilling, for example – but the opportunities are limited. What cheap foreign labor has freed us up to do, and what we need lots of, is healthcare. But it seems fair to at least ask how well the actual human resources that America has to offer match up with these jobs that Americans need done.

The issue is fraught with gender politics. The only thing we need more of that lends itself to male proclivities is military service. It’s never a good thing when a country’s outlet for its excess testosterone is the battlefield. Temporarily, we are in the opposite situation: too many wars and not enough fighters. But if we can get past Iraq and Afghanistan, we really don’t want war to be the most attractive option for our young men.

In contrast, healthcare is essentially a feminine pursuit. Doctors have historically been men, but that’s changing, and a political and social consensus exists that women can do the job as well as men. I suppose there’s a certain amount of machismo in the operating room, but, again, the issue isn’t super-surgeons; it’s “ordinary” physicians and nurses and aides and technicians – jobs that there are a lot of, jobs whose salaries make the future of Medicare so daunting. Not having to manufacture has freed “us” up to do these jobs. What’s not clear, though, is exactly who will do them, with what implications for our social structure.

Toward a Post-Job Economy

Maybe we need to go back to the premise behind our jobs-based economy. At the end of the day, using jobs to allocate goods and services is simply one technology for doing so. There are others. Capitalism allocates goods and services to those who risk their capital. Communism allocates goods and services based on need. As a practical matter, we cannot all be capitalists: nothing would get done if no one labors. And if money needn’t be earned, nothing gets done either. So Jobs are how we make things, and jobs are how we get things.

The key, I think, is to recognize that our manufacturing sector will be permanently capital intensive. It may be capital intensive now because foreign labor is cheap, but even if foreign labor becomes more expensive, our inventors will step up and take their place. We are already seeing a concentration of wealth in the hands of those who own highly productive businesses. That concentration is unhealthy, and I believe we need to divert the flow of some of that wealth to people who have worked a “full” career, as adjusted to reflect the capital intensity of our industry.

Obviously, this is radical medicine, as it denies the ethical supremacy of the market and the associated degree of autonomy in the business sector. The anti-trust laws offer a good intellectual model for tampering with the “free” market. Those laws interfere tremendously in the unfettered activities of business titans. Why not allow a cartel? Lots of economies of scale, no destructive price wars, one-stop shopping. And yet, history teaches that the corrupting effect of power trumps the efficiencies of focus. Competition is good for consumers, but it is not good for monopolists. Competition requires meddling in the way business is done. So we meddle.

The same arguments apply to the concentration of wealth resulting from a capital-intensive economy. In a labor-intensive industry, the assets go home at night, and nobody owns them. In a capital-intensive industry, the assets are turned off at night (sometimes), and as few as one person may own them. So long as we have a mix of industry types, the workers do all right. But when workers, as a group, have nothing to sell, because foreigners or machines are under-pricing them, ownership of manufacturing assets shifts to the few who own capital-intensive businesses, and that is not a politically acceptable equilibrium.

Sadly, diagnosis does not always result in useful prescription. Marx understood the ills of capitalism quite well, but he didn’t have clue how to set up a better arrangement. The Europeans are in a mess now because they tried to do what I’m saying must be done – pay people not to work. They have proved either that it can’t be done, or that it can’t be done the way they tried to do it. I’m not sure what happens in France when they raise the early retirement age. If people work longer, other people won’t work at all, unless the austerity creates jobs, which is not how such things usually play out.

I do not know enough about the capital-intensity or concentration of wealth in Europe to draw strong inferences from those countries’ woes. I do understand, though, that a government can promise to much and tax too heavily. Still, if we are to allow trade or automation, we will have a capital-intensive economy, which means that we must do something about the concentration of wealth, and we must either find a way for low-productivity local jobs to pay what high-productivity jobs pay or for jobs to not be the way goods and services are allocated in our economy.

One way to reduce unemployment is to restore the one-earner family as the national business model. That would cut the (paid) adult workforce almost in half. The one-earner family has advantages over simply shortening working lives to thin the workforce: it allows people to work longer, which means that talents and skills are not wasted in early retirement, and it frees up one member of the family to rear the children, which is not a bad thing. But that sounds like trying to put Jeannie back in the bottle even before we look at the job the one earner would be doing. If that job is nursing, … well, as I said, the problem is fraught with gender politics. (If the one-earner thing happens, it will not be by political action but by peer pressure from out-of-work families on their two-job neighbors.)

In a labor-intensive society, people must work for a living. But in a capital-intensive society, we must all become capitalists. Since we cannot do that directly – the allocation of capital is a skill with enormous economies of scale – we must do the next best thing: tax the actual capitalists enough to keep the rest of us well fed (but not so much that they lose interest). Taxes, of course, take lots of forms, and not all taxes discourage economic activity. The Obamacare provision allowing “children” to stay on their parents’ health plans to age 26, which pays young people not to work, imposes a tax through the employers whose healthcare costs rise accordingly. Happily, 22-26 year olds are the cheapest to insure, so the tax does its job very efficiently.

I admit that what I am describing has an annoying European feel. Haven’t we seen this movie, and doesn’t it end badly? I’m inclined to a more granular view. The social democracies of Western Europe may have been too generous for their specific capital bases – too much cargo, too little engine. Giving displacement of American workers the best possible spin, let’s just say that our economy has achieved unprecedented productivity. Of course, the “natural” equilibrium of such an economy features concentrated, dynastic wealth. We must make that equilibrium something else – a level of general prosperity similar to when one manufacturing paycheck could support a family very nicely.

There is no logical or doctrinal obstacle to that result. I’m not proposing a communistic redistribution of wealth or a needs-based allocation of goods and services (although I would means-test all benefits by taxing them in the hands of high earners). People should work productively at some time in their lives, and, most important, their post-work lives (and, maybe the pre-work lives of their children) should reflect their actual contribution to the economy. I just believe that a larger portion of commercial revenues should go to compensation, not as wages, but as pensions and other benefits paid to people for leaving the playing field to the next generation of workers in the increasingly shrinking workforce. (Yes, this is wasteful of talent and skill, so feel free to support the one-earner-per-family alternative if you have the, er, courage.)

It may turn out that we cannot prosper in our two-earner model unless most of our people are employed in making the things we use. If so, we’re in trouble, because competition from cheap labor (and tariff-nullifying machines) will not go away. But everything turns on specific, contingent facts. The question cannot be answered with generalities or lazy inferences from others’ failures. Everyone who tried to invent a flying machine before the Wright Brothers failed, but none proved that a flying machine could not be built. We are not going to become less capital-intensive anytime soon. We should at least try to treat the challenge as opportunity.