Sunday, March 24, 2013

Where Did All the Jobs Go? 

Harlan B. Miller

We are in what we all hope is a fairly short-term economic crisis, one that was triggered by financial greed and by the irresponsibility of lenders, borrowers, speculators and regulators. As a result of this economic slowdown, many people are out of work, and many have just given up looking for work. But, important as this transitory unemployment is, it isn’t what I want to address here.

 Instead I am interested in a jobs problem on a much longer time scale. Jobs are disappearing not just here but around the world. In the long term, if we don’t change our ways, we are headed for a rising and permanent labor surplus. We already have significant numbers of adults who have never been employed. Without substantial changes we can expect the never-employed to become the majority, perhaps in less than a century.
Most of the things I will tell you, you already know. But you have probably not put them together.


Let’s take a very, very, long view. When first we came down from the trees we weren’t really well equipped to meet our needs. Our needs were quite simple, enough food and shelter so that enough of our children could live to puberty and maintain the numbers of our group. Everybody worked full time hunting and gathering, and probably it took, on average the labor of, say, 101 or 102 humans (or proto-humans) to meet the needs of 100 humans. So most groups failed, and either died out or were absorbed by other groups. Some were lucky enough or lucky to be hardworking or smart enough, to bring the ratio down to even, 100 working could support 100. These groups became the ancestors of us all. By the way, I’m just making up all these numbers.

Then we got even smarter and invented spears, and stone tools, and fire. Now maybe just, say, 96 at work could support 100, at the previous level of need. However, we are no longer at that level of need, now we need firewood gatherers, and fire tenders and starters, and people to chip the flints. Our raw productivity has increased, but so has our level of need, so we are still just about even.

Eventually we create agriculture and start living in larger groups. Our raw productivity zooms up, but so do our needs. Now we need food storage and digging sticks, and a lot more medicine men because we have much more disease when we are more crowded and less mobile. And very soon we are going to need traders and bookkeepers.

 Fast forward past bronze and iron and efficient fishing, and so on. By the Classical Age of Greece we are enormously more productive, worker for worker, than the hunter-gatherer, but our needs have increased enormously. We need temples and priests, theatres and playwrights, money, warships, lots of bookkeeping, law courts, marriage, and so on and on.

Very approximately the increased curve of productivity and the increased curve of needs stay together. Really important inventions, like the wheelbarrow, kick the productivity curve upward. Social developments such as the transformation of women of the wealthier classes from workers to status-symbol birds in gilded cages can push the overall productivity down. But more or less, on the whole, by and large the supply of productivity and the demand of needs (or at least wants) ran side by side.

Then came the Industrial Revolution, and the Transportation/Communication Revolution and the Information Revolution. Beginning early in the 19th Century productivity per worker began to accelerate, and to pull away from aggregate needs (wants). Needs greatly accelerated, too. We needed highways and railroads and telegraph systems and telephones and to explore the planet, and eventually at least the solar system. Now we all need fast internet connections, and dentists, and insurance, and at least an automobile apiece.

Still, despite our seemingly endless desires, our ability to meet those desires has far outstripped them. If, with my imaginary numbers, it has almost always in the past taken just about all the workers in a population to support that population, so if in the past a population of 100 could meet its needs with full employment, then today probably no more than 75% or 80% of the workers are needed. This percentage is dropping and will continue to drop.

We have, and will continue to have a structural labor surplus.

Now let me pitch things a bit less abstractly and come to the present day.


First consider the jobs that left the U.S. (and the UK and Germany and Japan, etc.) and went elsewhere. This is not a new development. American shoe manufacture left New England for the lower labor costs of the Carolinas, then left the country for Spain, then left Europe for Indonesia. Now clothing is made in sweatshops all over the world including such unlikely places as Nepal and the United Arab Emirates. Everything else is made in China, much of it in factories owned by Taiwanese.

I’ve been struck for several decades by the manufacture of manhole covers. One would think that the ratio of labor cost to material cost is very low for manhole covers, and since they are quite heavy, that shipping would be a significant part of the end price. But for quite some time almost all the manhole covers in the US have been made in India. It doesn’t seem that that could make sense, but obviously it does.

The factory jobs that have left are just not coming back (unless the world economy collapses). In time wages and conditions for Chinese workers should improve, but barring catastrophe American workers will never be cheaper by the hour than Chinese, and almost certainly they will never be cheaper per unit of production.

There are still areas of production in which we and the other prosperous countries dominate, including construction machinery and some types of instrumentation. But there is no reason to think that will be so forever.

A decade or so ago some of the futurist literature told us, or anyway people like me, not to worry about all the factory jobs moving away, since there would always be jobs for us “symbolic analysts” who design the systems and write the software, and so on. Workers would work, but we would direct.

Of course this was always nonsense, and some of it was racist nonsense. The populations of India and China, put together, are about eight times the population of the US. So there are about eight times as many smart Indians and Chinese, in any of the many varieties of smart you choose, as there are smart Americans. Fortunately for us, it will be some time before their education systems catch up. But already we outsource lots of programming, medical diagnosis, and so on.

With all the fondness for “distance learning” one could surely outsource the teaching I used to do.

These jobs are gone or going, and they’re not coming back.


Second are the jobs that simply disappeared.

My first job was a high school summer job in a large bank. It was 1953 or 1954. I was given trays of cancelled checks, in order by account holder name. There were no account numbers. I checked the signatures against the account cards and filed the checks in chronological order in folders for each account. Other people had already checked signatures and deducted amounts, and yet others prepared statements. The back offices of banks were well populated.

Today almost all of this is done without human intervention once a teller has scanned the check. And, of course, since there are, relatively, many fewer paper checks in the system than there were a half-century ago the number of tellers and cashiers is also proportionately much lower. The transformation of the financial world has multiplied the possible varieties of fraud, but radically decreased employment opportunities.

There are essentially no more telephone operators, and very few gas station attendants. There are still gas station attendants in Oregon and New Jersey, which illustrates the fact that quite unnecessary jobs can be created or preserved by government intervention in the market.

There are still court reporters, but there are now very few stenographers (that there are any is due to the persistence of ego).

It is probably impossible to specify how many jobs have disappeared because of the transformation of logistics. Containerization has revolutionized shipping, radically cutting costs and pilferage. Hundreds of thousands of low-wage longshoremen world-wide have been replaced by thousands of well-paid, highly skilled crane operators and directors. And the number of these expert technicians will continue to decrease as robotics improves.

More visible to most of us is the effect of bar codes and radio-frequency IDs and other machine-readable tagging. Try to imagine how many check-out lanes your favorite grocery store would need if the cashiers had to read prices on the products and enter them manually. Then add in the needed additional pricing and stocking personnel, inventory takers, warehouse workers, and so on. Thousands and thousands more people would be employed, consumer prices would necessarily be much higher, and you would no longer get customized coupons.

In agriculture changes in technology and especially in market organization have largely eliminated the family farm. In almost any non-corporate farm in our part of the country at least one adult has, of necessity, an outside job. As I once overheard “In the Depression people could go back to the farm, but now there ain’t no farm to go back to”.

Not only are there many fewer farmers, there are now very few citizen farm hands. Consider hay bales. I’m certainly no authority here, but it seems to me that in the 40s and 50s and probably earlier baling machines produced a package, roughly contained by twine, about a foot square in cross-section and perhaps three feet long. Such bales could be hauled and stacked by humans. (Not easily. They were heavy and awkward.) One rarely sees such bales today. Instead hay is stored in enormous plastic tubes, or in gigantic wheels. In any case it is optimized for mechanical, not muscular, moving and storage.

In the US the radical reduction in agricultural labor was spread over perhaps half a century, but in Mexico, whose smaller peasant farms were even more vulnerable, it was a matter of at most half-a-dozen years, and there is no telling when that economy will recover from the shock.

When, in the first half of the Nineteenth Century, weavers and others were displaced by the new and enormously more efficient mills, many protested and some tried to destroy the machinery. In England they were called Luddites from a leader named Ludd. On the European continent they were “saboteurs” because they put their sabots (wooden shoes) in the gears.

The Luddites and saboteurs lost, obviously. We’re simply not going back to spreading manure by hand, or reading and writing price tags, or shipping towels or televisions one box at a time. Almost none of the technologically-displaced jobs are coming back. (There are a very few possible exceptions.)


Third are the jobs that we just won’t take. Varying combinations of low pay and difficult, dangerous, or just very unpleasant labor result in jobs that most American citizens just won’t do. Agricultural stoop labor, such as picking strawberries or cabbages is out-sourced in-country to foreign laborers who may or may not be legal. The same is true of much of what little domestic service remains, and of animal slaughter. Our apples are picked by Guatemalans; English apples are picked by Moroccans.

Eventually even some of these jobs will disappear. Machines harvest corn and potatoes. Perhaps someday even berries will be delicately gathered by robots.


The forces that produce the labor surplus are effectively both universal and irresistible. In time fewer and fewer Chinese workers will be needed. A dozen years ago, at earlier stages of miniaturization, thousands of 12- and 13-year old Malaysian girls assembled and soldered tiny components on circuit boards. They were valued for their small and nimble fingers and the superior vision often found in the pubescent. Now they are in their late 20s and need glasses. No one replaced them. All the components are already on the chip.

So, barring catastrophe, we face substantial labor surpluses, effectively forever. If we look at this as meaning that larger and larger percentages of the population will be unemployed and require assistance, it sounds pretty bad. But if we look at it differently what it comes to is that keeping things running will require proportionately fewer and fewer hours of work thus providing more and more hours of leisure, then it doesn’t sound bad at all.

 In fact what is before us is the realization of the dream of some Nineteenth Century utopians. No one will have to work unless they want to. We can rest and play and read and write and contribute to productivity as a hobby. A dream come true.

But in fact we have so organized our world that the dream is a nightmare.

The nightmare we can still try to avoid is already reality in much of the Third World and what used to be the Second World. For millions and millions in the megacities of the global South there are no real jobs and no real relief. There is at best what is euphemistically referred to as the “informal” economy. 


Just imagine a world in which we all share the work and share the leisure How do we get there?

Right off the bat let’s mention some things that are definitely NOT part of the problem. The Social Security retirement age should be lowered, not raised. The best thing I ever did for the job market was to retire and get out of the way. Liberal education is not a waste or impediment to employment, it is preparation for a good human life with adequate leisure. Liberal education should not be seen as competing with vocational preparation but as primary. Perhaps we should extend K-12 education to become something like K-14. There’s no rush to get young folks into the job market.

There are a few steps in the right direction we could take right away. As things stand our tax structure provides real disincentives to hiring. It costs less to give a $1,000,000 raise to the CEO than it costs to give $1,000 raises to 1000 production employees, and much less than it costs to hire 25 new workers as $40,000 each. The reasons are the cut-off of Social Security contributions above $106,800 and, in the case of new workers, the cost of medical insurance for new hires. Just removing the cap and collecting at the full FICA rate on all income would both strengthen the long-term prospects of Social Security and level the cost to the company of raises. Going to a universal single-payer health care system would really encourage additional hires.

We already know some other near-term steps that would help the employment situation. Here and now we could create at least hundreds of thousands of public or semi-public jobs to repair our roads and rebuild our bridges and put power lines underground. It wouldn’t employ as many people as the CCC did in the 1930s; we don’t need millions of shovels, but thousands of machines with skilled operators.

This would require lots of training and retraining, but retraining is no panacea and certainly cannot return us to what we used to call full employment. Thousands of former fabric and furniture workers of the Piedmont of Virginia and the Carolinas could in principle be retained for other work - but there is no other work. And of course the Indians and Chinese can retrain just as well.

The only long-term solution to the labor surplus problem, the only real way to change it from problem to pleasure, is a dramatic shortening of the work year. Varying mixes of shorter days and/or shorter weeks and/more vacations could be used. The nominal work year in the United States is 2000 hours (40 hours a week for 50 weeks). In practice the average is closer to 1800 hours. In Germany the average is about 1400 hours. And Germany, unlike the United States, and despite its relative poverty of natural resources, maintains a positive trade balance. They export more than they import.
In the long run even Germany’s 1400 hours will be too much.

Predictions are just predictions and maybe I’m wrong. Maybe this will all become irrelevant when climate change, population growth, and fanaticism produce disintegration of the global system and we have jobs for everyone defending the seawalls of our North American fortress from the starving multitudes. But if we are reasonably lucky progress will continue and less and less labor will be needed to run the world. It is already very late to start thinking about this. Widespread leisure not something to be feared.