« Robert Waldmann (and others) on Dean Baker (and hence Paul Krugman) and the burden of the debt on future generations | Main | The efficiency case for nepotism »


Feed You can follow this conversation by subscribing to the comment feed for this post.

"What, then, is the policy response to keep robots from 'stealing' our jobs?"

My suggestion may sound a bit quixotic at first. How about a pilgrimage to Farnworth, Lancashire, the home of Dorning Rasbotham, author of "Thoughts on the Use of Machines in the Cotton Manufacture" (1780)? More than anything Adam Smith wrote about the division of labour or even Marx on industry, what Rasbotham wrote in that obscure and mostly forgotten pamphlet has become the commonplace of the conventional wisdom about the effects of technology -- that one the one hand it is "the real culprit" behind job losses but that, on the other hand, it makes "the economy as a whole" better off and thus the short term pain is a price worth paying for the long term gain.

Jim was right to point out that it is not technology but power and policy that isolate and threaten workers. The confounding factor is that technology is the screen upon which power and policy are projected. We can't see the forest for the trees.

We need to worry less about "robots 'stealing' our jobs" and more about the way "our jobs" transform us into "robots" In feudal Bohemia, the "robot" was the compulsory labour rent that serfs were required to perform each year for the landlord, typically 104 days a year. Indentured servitude still prevailed in Lancashire, England at the time of the introduction of the great cotton spinning inventions of the 18th century. Rasbotham was a magistrate at a time when the most frequent offense heard by the courts was that of servants "running away" from their employment.

In short, our presumably "modern" view on the effects of technology on employment is distinctly semi-feudal. And we don't know it. We are wandering in wilderness of the 21st century guided by a map drawn toward the close of the 18th. The first thing to do would be to familiarize ourselves with the map-making conventions of the earlier day. The second would be to consign the old map to an archive and draw up a new one but here we need to be careful not to inadvertently re-draw the old map from memory and habit.

That's why the pilgrimage to Farnworth.

BTW, I journeyed to Farnworth, Lancashire, last Monday to pay my respects to the memory of the worthy magistrate and grub around in some archives. This week I've launched a virtual tour at Ecological Headstand of the "man vs. machine" Ur-myth. http://ecologicalheadstand.blogspot.com/2012/01/man-vs-machine-ur-story.html

I really think it is time to start thinking seriously about capital replacing labour. I believe it has only just begun. There are still many things that humans can do and computers and robots can't. But I, for one, do not believe it will always be so. True artificial intelligence will come someday.

Trying to run our economy the way it runs now will make no sense if that happens. Those who own capital getting paid and the rest on welfare is not reasonable.

And even today in rich countries, we are producing far more than we really need and doing far more harm to the environment than necessary, just to keep everyone employed.

We need to reduce the incentive to work, through a guaranteed minimum income, a negative income tax, or similar scheme. Take the benefits of rising productivity in the form of more leisure, rather than ever more stuff, and reduce inequality. The benefits of all the knowledge that goes into the technology we are starting to produce should belong to everybody, not just to a few people. People who create technology today and who will create artificial intelligence tomorrow do not do it all by themselves. They are just adding a little bit on top of all the knowledge produced before. We should regard all that older knowledge as the common property of everyone, and everyone should get paid for it.

Paul: "Those who own capital getting paid and the rest on welfare is not reasonable."

Is it capital, or land? Remember, if robots can produce robots, the real rate of interest (using robots as numeraire) will be pinned down by the reproduction rate of robots, minus the rate of increase of the cost of the raw materials robots need to reproduce themselves (have I got that the right way around?).

"Is it capital, or land?"

If I may be a bit unorthodox, it seems to me that capital is merely a technically useful but transient notation. What is owned is the right to a flow of revenue and in that regard "capital" has an irresistible tendency to revert to "real" property, which is to say property that has been granted by the sovereign (royal = real) in recognition of service to the sovereign. Scratch a capitalist and you'll find an aspiring aristocrat. "The tradition of all dead generations weighs like a nightmare on the brains of the living."

Nick: "Is it capital, or land?" The crucial question is if the costs of reproduction of such a perfect robotic substitute for human labor will be greater or lower than subsistence wage for humans. If at some point it will indeed be lower then humans (or their role in current economy as a labor unit) will become obsolete and redundant for capitalism. They can still retain their value for political purpose as in ancient Rome (where slaves an spoils of war served as robot substitutes) so that they would be able to use their political power to negotiate a welfare state with elites

However I am not that sure even about this. It all comes down to the question who will control a violence. In the presence of robot soldiers and police force, it is possible that human laborers could be helpless. Or alternatively, technology can give each individual massive power so that their negotiation position in a new social contract can be stronger. But then we are heading into another possible dystopia so that we replace Artificial Inteligence Oversees over humanity by the Apocalypse as a technology gives small groups or even individuals immense destructive power that will be used sooner or later by the first insane individual (group)

The Evil Effects of Robot

"The system of rent by robot or forced labour,—that is, so many days' labour without any specification of the quantity of work to be performed,—is a direct premium on idleness. A landlord wishes a field of corn to be cut; his steward sends out, by means of his Haiduks, information to the peasants to meet at such a field at such an hour with their sickles. Some time after the hour appointed a great part of them arrive, the rest finding some excuse by which they hope to escape a day's work; while others send their children or their wives, declaring some reason for their own absence. After much arranging they at last get to work; a Haiduk stands over them to see that they do not go to sleep, and between talking, laughing, and resting, they do get something done. Where horses are employed, they are still less inclined to hurry; lest they should tire them for the next day, when they use them for their own purposes.

"Now how much does the reader suppose such workmen perform in one day ? Count S-- says, just one-third of what the same men can do easily when working by the piece; and he has accordingly compounded his peasants' one hundred and four days' robot for a certain amount of labour, which they generally get through in about thirty-four clays.

"Another evil of the robot is the ill-will it begets between the masters and the workmen : their whole lives seem to be a constant effort, on the one hand, to see how much can be pressed out of the reluctant peasant; and, on the other, how little can be done to satisfy the terms of agreement, and escape punishment. Mutual injury becomes a mutual profit; suspicion and ill-will are the natural results."

Sandwichman: re-pose the question this way: "will land rents and land prices rise?" "will interest rates rise?"

"will land rents and land prices rise?" "will interest rates rise?"

You'll have to ask the King. Seriously, though, my point is that the "economic" relations are an overlay on fundamentally social relations of domination. To the extent those underlying social relations can remain concealed by a normally functioning system of exchange, rents and prices can rise and fall according to perceptions of "supply and demand" (or higgling). Replace "workers" by "robots" and the terms "interest," "rent," "profit," and "wages" are drained of any recognizable meaning or purchase, so to speak. In that circumstance rising and falling could only be by proclamation.

There are still many things that humans can do and computers and robots can't. But I, for one, do not believe it will always be so. True artificial intelligence will come someday.

That may be true, but it doesn't follow that humans will be priced out of the market. Comparative advantage tells us that even if I'm a better lawyer and typist than my secretary, it doesn't mean that I should get rid of her. In any event, even now, robots haven't priced out labour in manufacturing, they've priced out "high priced" labour (of the sort provided by CAW members). Labour intensive manufacturing is still thriving in low-wage countries.

"Take the benefits of rising productivity in the form of more leisure"

That's already happening. In the US at least, hours spent working per week have been declining steadily since the 19th century (see, for example, US data up until 1990 http://eh.net/encyclopedia/article/whaples.work.hours.us. For more recent international data, see http://www.bls.gov/opub/mlr/2009/05/art1full.pdf). This point gets lost in a lot of discussions of living standards.

"They are just adding a little bit on top of all the knowledge produced before. We should regard all that older knowledge as the common property of everyone, and everyone should get paid for it."

Right, but everying happens at the margin. If older knowledge is common property (as it clearly is) why would anyone pay for it? It's the marginal extra knowledge that has the value. Newton may have been standing on the shoulders of giants, but he was the one seeing further than others.

Bob Smith wrote,

"Take the benefits of rising productivity in the form of more leisure"

That's already happening.

Most of that decline in hours worked per week occurred before World War II and much of the subsequent decline represented demographic changes in the workforce. Since the 1980s there has been virtually no decline in usual hours of full-time employees in the U.S. This was not an "accident" or a "trend". It was a deliberate Cold War policy outcome that halted a long-term trend. The original rationale for the policies (going back to the Truman administration, Leon Keyserling and National Security Council memorandum NSC-68) was "economic competition" with the Soviet Union.

In 1947, perennial Presidential advisor, Bernard Baruch, who had directed the War Industries Board during World War I, appealed fervently for a longer workweek of 44-hours to boost national output, jobs and consumption. "Unless we work," Baruch warned, "we shall not be able to maintain our claim to power. That would be the greatest blow we could receive, for it would strip us of our strength to preserve our way of life." Ten years later, after the Soviet launch of Sputnik in October 1957, Texas Senator Lyndon B. Johnson again called for a longer workweek to combat the Soviet threat.

In the 1950s and 60s, Clyde Dankert was a widely respected expert on the economics of the hours of work. Although he acknowledged that shorter hours could lead to higher productivity and improved worker welfare, he concluded that in the context of the Cold War, the priority should be on maximizing total output.

Most of the decline in hours worked comes at the end of our lives. We like to bunch up our leisure time and call it "retirement". We don't all die at 65 any more. (Maybe at the beginning too, but I'm not sure whether school counts as work or leisure!). It's the total ratio of hours worked divided by hours alive that matters. Plus the quality of working conditions.

People have been worrying about technology replacing people for as long as there's been technology. Has it happened? No.

Instead, people get technology to do something, and then go find something else that's more interesting to do. Such is the story of human progress. Invent agriculture; use free time to build better houses, markets, empires, etc.

Humans will always be needed, because humans provide the creative element, the one that can improvise and come up with a new idea. Anyone concerned about artificial intelligence should go do some work with an advanced computer system. These things become increasingly complicated, but they're never more than data-in->process information->data out. Computers can beat humans at games because they can flawlessly calculate probabilities. But they can't invent a new game.

And I'm not even convinced that the individuals who lose their job to technology are worse off. Maybe if there 50-60 years old and have spent their whole lives moving a widget from one tool to the next. But young people who have the time to be retrained may well find themselves in a better position by the time they retire. Certainly my experience, and what I've seen amongst friends in the last recession, is that losing a job is a kick in the pants to find a better one, instead of just settling for the status quo.

"Since the 1980s there has been virtually no decline in usual hours of full-time employees in the U.S"

Which is why I also included the international data since 1980, which points out that the US is an outlier in that respect.

Also, I wouldn't read too much into the rhetoric of Johnson or Truman administration flunkies to explain the relatively slow decline in US working hours since WWII. It's interesting to note, for example, that within 3 years of Bernarch Baruch's appeal to increase the standard work week from 44 hours, the US government reduced the maximum standard work week from 44 to 40 (which, despite Johnson's rhetoric, was not reversed - in fact, it was extended to additional industries in 1966 under, wait for, a certain President Johnson). If that's a deliberate policy to increase working hours, it's a spectacularly stupid (and ineffective) one.

I think a more plausible explanation for the slower decline of US working hours since WWII relative to Europe would be a "catch-up" story (the US, being richer than Europe, started working less, sooner, than the Europeans did, and the Europeans spent their post-war years catching up) and a demographic story (the Eureopeans have aged badly while the Americans have not).

I used to worry about this. A lot. I make 'robots' that replace people out of work, and it has always bothered me. This is the story/lie I tell myself to salve my conscience:

We might also simply choose to not use so many robots. I know, I know, but maybe it isn't totally crazy. Assuming we can hold-on to democracy (big if, but stay with me), at some point mass unemployment, people with ZMP, extreme inequality, etc visits enough marginal disutility on enough people relative to the (declining) marginal utility of yet another robot for the owners of the robots, that even FoxNews won't be able to sell serfdom packaged as freedom. We'll vote to use less robots.

People like to be useful to each other. There is 'utility' in our relationships and connections to other people. I have to believe that at some point our utility to each other outstrips the utility of yet another house, Italian sports car, electronic gadget, or a few cents less for our broccoli.

Thinking about how things might look as technology displaces labour gets us into a zero divided by zero problem. Cheaper and cheaper goods and services, but lower and lower wages. What this might look like depends on the rates of decline of the two things.

I think Nick has a point in that, as capital becomes able to reproduce itself, it will no longer be a scarce commodity. What would be scarce would be raw materials. The cost of capital would just be the cost of the raw materials in it.

I hope we don't go here. I hope we can agree at some point that more production is unnecessary and harmful, and that an incentive to work is no longer needed. Dividing up the output of the economy equally would then be reasonable.

I don't understand much of this. Employment rates and real wages were increasing as manufacturing employment fell. The shift away from manufacturing is part of a sectoral re-allocation of resources.

"I don't understand much of this. Employment rates and real wages were increasing as manufacturing employment fell. The shift away from manufacturing is part of a sectoral re-allocation of resources."

Yes, absolutely.

Mike had a minor typo in his last paragraph "demanders" not "suppliers" have the market power. That is quite correct and there really is no policy response. Nor should there be an expectation that the rest of us should pay to make up the gap between the promises made by the union and reality. However, more importantly, we need to think about the way that market power is exercised. I really like Mintzberg's characterization in the Economist.

Whoops... will fix that. Thanks!

Stephen: It's true that for the last 200(?) years technology has improved, labour has been replaced in one job after another, and yet employment grew in other sectors, and real wages increased (not in every country in every year obviously, and depending on the measure of inflation). So far, the fear has turned out to be misplaced. But I think it's an open question whether this will always continue. The demand for horses fell when we invented robot horses that could do everything a horse could do with less maintenance and cheaper fuel per horsepower.

I dunno. I should probably look up what has been written on the economics of Star Trek, but if we ever get to the point where we can produce something by dictating instructions to a replicator, then it seems to me as though we'd all be in the service sector. The Enterprise even had a hair stylist on board.

But they had only one psychologist ( or maybe Deanna Troy was for the officers only..)
I always thought that neuroses were the only infinite resources in the universe.

This discussion reminded me a interesting blog I read last year: http://jeffweintraub.blogspot.com/2011/01/adam-smith-on-poverty-progress-wages.html Perfect replacement for human labor by technology may yet again remind us why masses benefit from free market economy. It certainly seems logical that it is a unique properties of human labor among means of production that made this happen.

Current day naval vessels have ship's barbers, Stephen. And the tablets that were used in the case of books are simply iPads. How about all those cellular phone? Or the famous example of 3.5" floppies, already an obsolete technology?

Mr. Stanford is being disingenuous when he states that "real wages in Canada are no higher today than a quarter-century ago, despite a 35-per-cent increase in labour productivity in the same time."

I'm sure he knows, as an economist, that the labour productivity gains flowed through to the benefit of consumers which raises the question as to why the unions think they are entitled to a portion of these productivity gains. They, in their capacity as consumers, have already benefited from these gains.

Second, Caterpillar is not a victim of globalization, it's one of the few sectors that aren't. The Caterpillar Plant is former Electro-Motive Division of General Motors. EMD was formerly located in LaGrange, Illinois but that plant closed. Its only competitor is General Electric Locomotives in Erie, PA.

GE and EMD have been a duopoly in North American locomotive manufacturing since the 1960's. North American railways face unique challenges and have distinct needs and regulatory frameworks. That's not a contradiction. North American railways are freight tonnage haulers. North American locomotives are designed to haul large loads over long distances. Braking is just as important as accelerating and North America has developed a particular design focus on braking. European locomotives are not designed for North American conditions and their braking capacity is generally far less. The UK doesn't even use the kind of air brakes that have been standard in NA for 140 years. Regulations reflect this.

There are only 6 Class I (long-haul) railways left in North America, the market for locomotives is steady.

There is nothing philosophical or academic about Caterpillar's actions, they are crass corporate greed. The crudeness of their tactics is painfully apparent in this case.

As Nick implies, Land values and rents will surely rise. It seems reasonable (to me) to imagine a future in which robots are in unlimited supply (except for the finite materials required to produce them). Human labour will be worthless, or more precisely, the total economic value of a human will be about the same as a small pile of iron, copper and a bit of silicon (or whatever a physical robot will be made of - probably flesh). So, if you want to live off the rents of your "comparative advantage" in that world, Bob, knock yourself out. I'm with Paul (and presumably this is also the view of Sandwichman): sharing the rents of the earth as a citizen's dividend is the only sane solution.

What the interest rate will be, is a good question, Nick. My short (dumb?) answer is roughly the rate of increase in output from Land. But I think you'll have a *very* hard time in defining a consumption basket that is stable enough to serve as a numeraire in a world that is changing as fast as the one we are envisioning. So it might be hard to define "the rate of increase of output." Which means that it might not be possible to have inflation targeted money, which means that "interest", in terms of whatever constitutes the numeraire, might have a very different meaning. Maybe money will represent a claim to the total Land value. "Interest," will be the rate at which we split those units. Debt, as in "consumption basket forward contract" might also fall by the wayside.

[All of the above depends, of course, as Patrick points out, on machines that are clever enough to supplant our labour, yet docile enough to serve us. Who knows?]

As far as Mike's post goes, I think it brings interesting questions about the correct IP policy in a country of "hewers of wood and bearers of water." There is no doubt, that in the brave new world discussed above, the last human inventors of robots would try to claim infinitely long patent rights ("This patent describes a robot, that invents a robot, that invents..."). But why enforce the rights of others to use capital to enslave us? By no means do I suggest that we cease to enforce IP law. I am just wondering what the optimal relative strength of our laws should be compared to the main producers of IP. China, for example, has certainly benefited greatly from taking a relatively less strict approach to enforcement.

One more interesting question is whether this world going to happen any time soon. First observe that it would probably take about 10% of our current labour to produce the output that we consumed 100 years ago. Assuming we weren't that much less happy then than today, we could pretty much be done today if we weren't engaged in a zero-sum state of perpetual competitive consumption. So we don't need to work 40 hours/week until our comparative advantage is reduced to a small pile of base elements (and the robots kill us for parts). When this future occurs is pretty well up to us.

"There is nothing philosophical or academic about Caterpillar's actions, they are crass corporate greed."

It's something apart from greed. As one high ranking CAW person mentioned to me "This is the only company I know that hate unions more than they love profits." It's hard to see how Caterpillar's actions are helping their bottom line in the short run.

There's a difference between being "greedy" and wanting to smash people, regardless of the cost.

You may want to argue that Caterpillar's actions are helping their profitability in the long run. But since when are corporations overly focused on the long run?

"You may want to argue that Caterpillar's actions are helping their profitability in the long run. But since when are corporations overly focused on the long run?"

But that would be an odd criticism from the left, i.e., that Caterpillar is bad because it's making decisions based on long-run profitability. The usual critique is the exact oposite, that corporations made bad decisions in pursuit of short-term profit.

If robots ever come close to replacing humans in most of their complexity, I don't think we'll have to worry about it for more than one generation. Robotic companions totally devoted to their human are quite likely to replace those messy and often unhappy human relationships.

If our reproduction rate is below replacement now, think what it will be once every man and woman can get a companion tuned to their exact needs - how many people (especially those now without jobs) will be mainlining the emotional heroin of a 'partner' for whom they are literally their companion's reason for existence.

"But that would be an odd criticism from the left, i.e., that Caterpillar is bad because it's making decisions based on long-run profitability. The usual critique is the exact oposite, that corporations made bad decisions in pursuit of short-term profit."

Agreed. I'm actually at a loss to figure out what Caterpillar is trying to accomplish, to tell you the truth.

So am I. I'm a rail fan and I know the North American rail sector more than most. This plant is a recent acquisition for Caterpillar; it used to be GM then independent until a few years ago. Breaking a union that has been in place for decades in a plant with stable market share in a stable industry is an emotional decision, not a rational, business one.

What profitability gains Caterpillar expects from gutting its labour reputation are beyond me (specifically the reputation of EMD). What will happen instead is that the small list of big locomotive purchasers will send their orders to GE. Which I might add is also unionized.

I remember ten years ago that Union Pacific ordered 1000 locomotives stretched over a decade from EMD, now Caterpillar. It was the biggest order ever placed for locomotives in North America. This is not a bad business.


My understanding is this plant faces real issues regarding Buy America(in terms of not being able to bid for US commuter rail contracts) however, it has been for many years exporting a substantial amount of its productions outside of North America. Occassional you see "funny" looking foreign locomotives being towed to the port of Halifax along the CN mainline.

Sustainability, and widespread economic participation, are going to be the issues of the future.

We are still arguing about and trying to solve the issues of the past...

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad