Tag Archives: robots

Brad DeLong sends us to another hush warning about mass disemployment and future inequality, this time from the excellent Jim Tankersley. As frequent readers know, I don’t think we’re in for some techno-dystopian future and, while I fall on the right (ish) side of Marc Andreeson’s dichotomy – “The spread of computers and the Internet will put jobs in two categories: People who tell computers what to do, and people who are told by computers what to do.” – I think there’s a lot of cognitive dissonance in this conversation.

Let’s be clear. It is theoretically possible – yet improbable – to have either a fall in median standard of living or natural rate of employment. It is, in every way, impossible to have both. Understanding why isn’t all too complicated, and requires only a simple knowledge of supply and demand. Let’s think of the future utopian society operating within a three-class system: unemployed masses, educated technocrats, and the rich capital owners.

Tankersley points to work from Frank Levy and Richard Murnane suggesting that America isn’t in a great position to educate people in the would-be “unemployed masses” to the “technocrats”. That’s a red herring. While in the realistic framework from which they argue – the medium term – more education of the poor is exactly what we need, in the long-run the demand for technocrats is likely to be low.

But let’s take a moment to think what a fall in the standard of living means? Forget my dollar income, for a minute. A stereotypically middle class American derives his high quality of life from free public education, good shelter, incredible consumer choice, copious quantities of food, and access to cheap but effective consumer durables. Americans also enjoy excellent public services like parks, free roads, cheap energy from other institutional arrangements.

Americans own these goods and services  because – by and large – Americans are involved in the production and distribution of these goods and services. (No, not the direct assembly, but most everything else). But let’s say we enter a robot future. (Interlude: have you seen the Tesla factory?) Let’s say minimum wage workers in McDonalds are replaced with wage-free robots and truckers are replaced with automated engines.

The Neo-Luddites tell us that all these middle class workers will overtime be disemployed and hence observe a crash in their standard of living. Notwithstanding the fact that redistribution mechanisms solve this problem (which does not make it theoretically impossible, per se) – this cannot happen. If robots replace workers who are in the service of providing mass market goods – toys, teaching, everything we listed above – that means those goods are being produced, and hence exist. Of course, we get the Keynesian problem of overproduction if for a time period there is excess supply. But if robots keep producing these goods there is no way they will not find their way into the hands of American masses.

Okay, you say, “what if robots don’t keep producing these goods”. When Kevin Drum linked my previous article on Mother Jones, I found all kinds of liberals (a group with which I otherwise identify) telling me that I’m naive for not knowing that the bad capitalists will hog all the robots, that they won’t share their wealth, that political redistribution is impossible, and such. That means the goods aren’t being produced. But let’s say there’s a town – Detroit? – of the “mass unemployed”. If robots supposedly “replace” their jobs and somehow they don’t get the rewards, two agents aren’t just going to sit and go “oh. shit. I’m unemployed. Too bad”.

No. It’s not like the demand for consumer goods vanished. The mass unemployed will create their own economy. That means since I don’t have a car – because of, you know, “robots” – and you don’t have a microwave, I’ll specialize in microwave production and sell you microwaves and trade you for a car. And before you know it, unemployment solves itself, because demand for goods creates a demand for labor creates employment.

Of course, this situation is highly unlikely to arise. There will be some disemployment, but it will not be associated with a falling standard of living. That’s impossible. This situation is unlikely to arise because:

  • It’s sad. It’s like a hark back to an older economy. We should be using robots to consume more, and that should be the dream of every technologist: not the exclusive ownership by a capitalist minority.
  • The first situation, where we have disemployment, but higher standards of living – indeed, “post scarcity” – is more probably. Profitable corporations need a large consumer base. The mass market is the best and, indeed, only such market. While some commenters tell me that “the rich will own the robots for their own uses”, there’s not much profit if all the one percenters create goods only for other one percenters. For one, there would be a huge excess capacity of robots, compounded by the fact that rich people like artisan work. Not crappy, mass-market Tupperware.

The former point is my philosophical belief and I hold it as axiom. The latter has good empirical foundations. The 2000s – a decade of globalization, capital-biased technical change, and rising income inequality – was actually accentuated by a fall in consumption inequality. Over the past four decades, income inequality increased 237% as much as consumption inequality. And even this is just a first derivative of my point, that is lower-quintile consumption increased far more than income. (Many also suggest we overestimate consumption inequality as it fails to capture surplus deriving from the Internet and other free goods).

I think I’ve somewhat convincingly argued (if only to myself) that both a fall in broad living standards – or even a deceleration in the rate of change thereof – is unlikely to be coincident with disemployment. But this leaves open the possibility of either happening exclusive of the other. Let’s consider the first pair, “fall in living standards, but no disemployment”. I think this is the trickiest possibility, because I can’t really see a way in which this can happen, but also can’t see a way in which it’s impossible – outside of political-democratic institutions which, quite honestly, are failing. All I can argue is that technology is definitely going to increase the consumption from the wealthy which almost certainly will translate into higher wages – if at a far, far lower rate – to the poor. A decrease, or even stagnation, then seems improbably: just a substantial disconnect between capital and labor.

The second case, “disemployment without any fall, and perhaps even increase, in living standards”. Everyone fears this. But we think about unemployment in the narrow sense of U3. But Americans are one of the most “overworked” countries in the world. We’re the richest country not only because of our incredibly high “output per hour” [productivity], but also just “hour” [work ethic]. This works well for some people. Doctors who slave it out as residents and then earn a criminally-unfair killing later in life. (Oh what I’d give to replace every damn doctor sucking money from the American middle class with a robot!)

But think of the single mother working two jobs just to have heat in the winter. Disemployment without a fall in living standards will do her well. She can spend more time with her family. In a previous iteration of this post, I wrote that we’d have a “cornucopia of thought”, or something. I still think I got the principle right, but I was fairly bashed by a bunch of readers for being an idealistic idiot. (“Not everyone has a high IQ”, etc.) Regardless, there are a plethora of studies showing engaged parents are crucial for someone’s future success, in society as well as in the economy. Today this works in advantage of the affluent (if not the ultra-rich). But in the future I do hope that a middle class worker can live very comfortably on 6 hours of work, getting more time to spend time with his family – whether reading books or at a barbecue – and sleep, relax, and ponder. Oh shit I’m getting “idealistic” again.

That’s not to say that economic mobility or condition for the poor in America today is anything great. Only that a sharp fall, or even an absence of elevation, of living standards is very, very unlikely. That is the fundamental dissonance inveighed by the anti-technologists. They assume a utopia in which robots provide everything, but at the same time a dystopia in which they provide nothing.

Abenomics, move over.

Kevin Drum isn’t as optimistic about the massive capital-biased technological change that robots promise:

Nor is it enough. Even if we can immerse ourselves in the web all we want for low cost, we still need to eat, clothe ourselves, live somewhere, and so forth. Until our future robot paradise arrives, this is a big deal. If you lose your job to a robot, your net economic position is going to be sharply worse than it used to be.

Drum doesn’t buy my argument that this structural change will almost-definitely be a good thing. At least not in the short-run. I want to clarify my last point, which I’ve clearly obscured, vis-a-vis the emphasized text. If mass labor is no longer needed to produce most goods and services (an assumption I doubt – but that later), there is definitely reason to believe in disemployment effects. But my point wasn’t about unemployment, per se, but standard of living. A rise in total output will allow much sharper redistribution of income.

Long ago, J.S. Mill noted that the humanitarians of the day needn’t criticize the Smithian free market in its “natural” tendencies. Because production was an economic question, and distribution was political. Of course, he ignored the intimate link between the two. If high-earners are taxed at sufficiently high rates, they will definitely be disincentivized at the margin.

But to the extent Drum’s suppositions are true – i.e. that robots replace most wage labor – production has, largely, been solved. (Drum’s whole point is there isn’t much work to be done in “running” the robots, material production will be on auto-pilot). The vast amounts of capital income would lend themselves to more redistribution, and I doubt that standard of living would fall much at all. Remember, the capitalists who own all the robots won’t have any profits without a huge and growing consumer base. Falling cost of production implies increase in supply, and hence more surplus –  consumers included. I cannot emphasize this point enough – without healthy consumption, robots are useless. I’ve read comments that the rich would “use” the robots “for themselves”. This is highly unlikely for a plethora of reasons but, if it comes to pass, will mean everyone else can create their own, relatively labor-intensive economy. Problem (kinda) solved. Capitalists paying huge sums for fixed quantities, like land, make a new Ricardian tax program a perfect choice.

This is tangential to the main point, because I don’t think we’ll face severe disemployment effects at all. For all the talk about “offshoring” to China, the American labor market adjusted remarkably well over the Clinton and early Bush years. The post-industrial apocalypse that is Detroit aside, the current crisis isn’t, at least not directly, one of structural unemployment.

On that note, this whole argument has been framed in an unfortunately Western point of view. Most wage labor across this world is treated in dismal conditions. They don’t want to work, but have no other way to put food on the table (or on the mud floor, as it might be). Just calculate, how many slaves do you have? I have 29. In a robotic world, that will be nil. I know there are many poor women across India who would rather put their ten year-old son in school than on the hot field as a brutally-treated farmhand.

Drum seems to think that in our robotic future, the demand for (most) labor will vanish:

I very much doubt [that millions will enter the ‘thinking classes’]. The vast majority of humans have neither the skills nor the desire for this. Rao may be right about “millions,” but that represents just a tiny fraction of the human race. What about the rest of us?

I was probably a bit idealistic about the future being “a cornucopia of thought, refinement, ideology, and science”. As I noted, if incorrectly applied, such jobs are by definition restricted to the smartest who, by definition are few. But an important part of the argument remains. Many people today who feel financial pressures either drop out of the educational system early, or undertake a purely vocational education, against their will. And this isn’t a bunch of intellectual snootballing either – I’m more than happy if people turn to “lower” pleasures (this was in the spirit of Mill, after all!)

But in the future we’ll see more youngsters confident enough to study english, philosophy, or the lab sciences. Noah Smith’s argument against all non-econ PhDs will become a nonstarter. Kids who want to pursue their time as “starving artists” will have all the means to do so. I think Drum underestimates the huge demand for the more “thinking” pursuits of life we’d see if the whole world – billions not millions – was educated.

As Tyler Cowen linked to, earlier today, a Master’s in Computer Science from one of the best universities is now available for $7,000. TotalThis is a story about how not even the most exclusive admission committees can keep the gate closed on learning itself.

And there’s another narrative here. The huge increase in material wellbeing (unequal as it might be) will itself create demand for all kinds of new things we can’t even imagine. Here’s one prediction. As a % of total employment, the sex industry will grow relatively rapidly. Because no robot can replace, how to put it.. a beautiful woman.

And this pattern carries across. We’ll have more people watching plays at the local theatre, and spending time with their grandparents. Or tending to a garden. Michael Pollen will no longer be chastised as a sexist pig.

Just because myself and Kevin Drum can’t predict the future demand for employment doesn’t mean it won’t exist. I don’t think most of the jobs today would have been predicted ten years ago, let alone fifty. I place incredibly stock in human ingenuity.

This just isn’t about the “long-run”. There are two ways about thinking about the immediate effects. Disemployment might be relatively prominent during “structural adjustment”, as an economist would say. But this doesn’t always have to be so painful. America became a service-based economy rather quickly. Also remember, at the hark of the industrial revolution Luddites were saying the same thing (and burning down buildings in the process). Drum can look retrospectively to analyze why they’re wrong and it’s different this time. But who is the say the future Kevin Drum won’t be saying the same thing!

And, in the meantime, unemployment can always take two forms: of bodies, or of hours. We have good reason to believe a good part of America is “overworked” – the same people that Drum thinks will be most hurt by this change. Moving to fewer working hours will allow working-class parents to spend more time with their kids and mothers to enjoy better maternity leave and benefits.

The smaller working-day will create more social interaction, perhaps at a local cafe. In fact, along with the sex business, I’d be bullish on fancy restaurants – in the future.

I hope I’ve made more clear the obscurities in my previous post. But I also want to get more specific. What does Drum think unemployment rates will look like over the next ten years? Fifteen? Fifty? What about labor force participation?

As for the former, I don’t think natural unemployment will ever hit more than 8%. (I don’t think it will get near, but I can see a few possibilities to the contrary). I think labor force will fall among the elderly – and I think this would be a damn good change. Working-age adults participation has been on secular decline, but I don’t think robot technology will do anything to put this fall on steroids. 

More specifically, over the next twenty years, I don’t expect participation to fall below 70%, adjusting for all hysteresis effects of low AD.

I do think Drum gets one thing absolutely right, robonomics has been given “surprisingly little attention among economists”. This will change. Because, if there’s one thing we liberals hate to admit, this is a supply-side revolution. We have to fundamentally reconsider what income is, and how it’s allocated. I think consumer surplus has already done a lot of that (I mean yeah, the Internet isn’t “free”, but it’s about as close as it gets. That’s pure surplus).

There is the more daunting question of work and dignity. As John Steinbeck beautifully put it during far worse times:

“The last clear definite function of men — muscles aching to work, minds aching to create beyond the single need — this is man. For man, unlike anything in the universe, grows beyond his work, walks up the stairs of his concepts, emerges ahead of his accomplishments. This you say is man — when theories change and crash man reaches, stumbles forward, painfully, mistakenly sometimes. This you may know when the bombs plummet out of the black planes on the marketplace, when prisoners are stuck like pigs, and the crushed bodies drain filthily in the dust. If the step were not being taken, if the stumbling forward ache were not alive, the bombs would not fall, the throats would not be cut. Fear the time when the bombs stop falling while the bombers live — for every bomb is proof that the spirit has not died. And fear the time when the strikes stop while the great owners live — for every little beaten strike is proof that the step is being taken. In this you can know — fear the time when manself will not suffer and die for a concept, for this one quality is the foundation of manself, in this one quality is man, distinctive in the universe.

Cutting words. But I wouldn’t doubt the creative capacity of Man. We’ll create movies, delve into the craziest depths of quantum mechanics, solve the question of P and NP, and “create beyond the single need”. We, after all, are human.

That doesn’t mean there won’t be inequality, or discontent. It won’t be the end of ideology, passion, or thinking. There will be an underclass: of who, what, and where I don’t know. History, after all, won’t be over.

But those in it will be a lot better off than they once were. And that’s why the future is a good place.

Update: Brad DeLong has aggregated many responses to this question. Most notably, Keynes himself:

I draw the conclusion that, assuming no important wars and no important increase in population, the economic problem may be solved, or be at least within sight of solution, within a hundred years. This means that the economic problem is not – if we look into the future – the permanent problem of the human race…. Thus for the first time since his creation man will be faced with his real, his permanent problem – how to use his freedom from pressing economic cares, how to occupy the leisure, which science and compound interest will have won for him, to live wisely and agreeably and well. The strenuous purposeful money-makers may carry all of us along with them into the lap of economic abundance. But it will be those peoples, who can keep alive, and cultivate into a fuller perfection, the art of life itself and do not sell themselves for the means of life, who will be able to enjoy the abundance when it comes… 

…Meanwhile there will be no harm in making mild preparations for our destiny, in encouraging, and experimenting in, the arts of life as well as the activities of purpose. But, chiefly, do not let us overestimate the importance of the economic problem, or sacrifice to its supposed necessities other matters of greater and more permanent significance.

My post – in other words – is almost a century too late.