Keynes’s Jetpack

In the long run we are all dead, but not all at the same time.

I draw the conclusion that, assuming no important wars and no important increase in population, the economic problem may be solved, or 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 . . .

—John Maynard Keynes, “Economic Possibilities for our Grandchildren,” 1930

The short run had never looked as bleak to Keynes as it did in 1930: “a slump which will take its place in history amongst the most acute ever experienced.” In Britain, the twenties had not so much roared as spluttered. But when Keynes visited Madrid to give a lecture, he chose to adapt one he had been trotting out to students for more than two years, attacking both revolutionary and reactionary doomsayers. Pessimism had made people “blind to what is going on under the surface.”

In the long run we are all dead, but not all at the same time. Keynes’s message was simple: extrapolate conservatively the economic growth rate of the modern age so far, and imagine the wonders one hundred years hence: 2030. His audience would not live to see it, but many of their grandchildren would. The great-grandchildren, born in the last quarter of the twentieth century, would climb a stairway to heaven and bask in unknown pleasures from middle age. The generation after that would be born into paradise.

In the best traditions of science fiction, the author fudged the precise workings of the technology behind the wonders. Keynes evidently had little growth theory to draw on: he talks in monetary terms of the wonders of compound interest. Investments simply grow at around 2 percent a year — ask not why. Technology improves, for an improvement in “technical efficiency” of 1 percent a year. Making generous room for more of the “disastrous mistakes” that had brought forth the depression, he predicted that living standards would “in the progressive countries” be four to eight times higher a century on. There would be a phase of “technological unemployment” as labor productivity outpaced the finding of new uses for labor, but ultimately we would work out how to spread the dividends so that everyone worked an average of three hours a day.

Science fiction tends to be at once too radical and too conservative in its visions of the future. We don’t have jetpacks, a moonbase, robot butlers, or a mission to Mars, but we do have the internet. Keynes turns out to have been on track in his numerical guesses. According to the long-run data assembled by the late Angus Maddison, per capita real income in the United Kingdom was in 2008 4.4 times that of 1930. Extrapolating that average 1.9 percent annual growth rate forward to 2030, and Keynes’s great-grandchildren (not literal ones; he had no kids) would have on average 6.6 times the real income of his contemporaries in 1930. The United States is right at the upper bound of Keynes’s estimate, having grown on average 0.2 percentage points faster per year: 5 times the real income of 1930 in 2008 and 7.9 times by 2030. (It works out to 7.6 times if we extrapolate from the slower average growth rate since 1970.) Extending the arithmetic to all the countries Keynes considered “progressive” — Western Europe and the Anglo colonies (Japan was still a long way behind in 1930 and Latin America a mixed bag) — we find real per capita income 5.5 times that of 1930, heading towards a 9-fold increase by 2030. Maddison’s figures give us estimates in the same ballpark for the world as a whole, but extrapolating from the different 1950–2000 data in the Penn World Table, Fabrizio Zilibotti finds a more astonishing 17-fold rise, due to a few extra tenths of a percentage point in annual growth.

And yet, could “the economic problem” seem any less solved? Where is our fifteen-hour workweek? Had Keynes been right about how society would use its productivity growth, we would not find his predictions borne out in the per capita income statistics — these measure only the production of commodities. It can be argued that we have taken some of the dividend in leisure at the end of our individual long runs: life expectancy has risen while retirement ages have fallen since 1930: the average American male enjoys an extra thirteen golden years. But even including this, the average European works almost double the proportion of their lifetime waking hours that Keynes predicted, and Americans two-and-a-half times. People everywhere are being told to lower their expectations and buckle down to years of austerity. Europeans are rioting, Americans occupying.

Keynes represents the strange intersection of two traditions. On the one hand, he is  an heir to a Victorian stream of aesthetic, moralistic anti-capitalism — with roots in an aristocratic worldview whose genealogy has been traced by Raymond Williams in Culture and Society. He is contemptuous of the “strenuous purposeful money-makers,” suffering from “a somewhat disgusting morbidity, one of those semi-criminal, semi-pathological propensities which one hands over with a shudder to the specialists in mental disease.” He looks forward to a

[ . . . ] return to some of the most sure and certain principles of religion and traditional virtue — that avarice is a vice, that the exactions of usury is a misdemeanor, and the love of money is detestable, that those walk most truly in the paths of virtue and sane wisdom who take least thought for the morrow.

On the other hand, Keynes belongs to a tradition of technocratic liberalism, which he played no small part in modernizing himself. Here, the economic order is justified not on the grounds of private property and tradition, but on utilitarian ones. It is a system that works, and to the extent it doesn’t work, experts need to tune up the machine, ignoring the misguided short-term self-interest of labor and capital. Thus the speed at which the economic problem is solved depends on “our willingness to entrust to science the direction of those matters which are properly the concern of science.”

This combination of attitudes explain how Keynes can be seen as both conservative and radical. “Economic Possibilities” is not the only work in which Keynes seems to wax socialist. There is, for instance, the “euthanasia of the rentier” passage at the end of the General Theory in which he looks forward to the end of capital incomes and the “functionless investor” who earns them. There too, he stresses that the transition “will be nothing sudden, merely a gradual but prolonged continuance of what we have seen recently in Great Britain, and will need no revolution.” Actually existing socialists, and sometimes the working class itself, disgusted him. In 1925 he asked, “How can I adopt a creed which, preferring the mud to the fish, exalts the boorish proletariat above the bourgeois and intelligentsia who, with whatever faults, are the quality of life and surely carry the seeds of all human advancement?” He never had much interest in distribution or inequality, except insofar as they were of functional importance to the system: how much would the poor spend out of increments to their income relative to the rich?

The second, liberal technocratic side to Keynes remains a vigorous strain in economic thought, but the aristocratic anticapitalism has been bred out. For the modern neoclassical it is obvious why Keynes’s predictions went astray — he was simply wrong about preferences.

He projected the ethic of Bloomsbury or the Cambridge Apostles onto the public at large. In the future, everyone would be an aesthete: everyone will learn from “those people 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,” those people most “able to enjoy the abundance when it comes.” Keynes believed that people had “absolute needs” which could be sated, and could be morally improved to stop caring so much about “relative needs,” which are only enjoyed to the extent that we have more than others. As the marginal utility of commodities fell, people would react to the income effect of productivity gains by steadily expressing a preference for more leisure.

Wrong on all counts, says the neoclassical: our indifference curves are evidently not shaped in that way. People have taken their income gains mainly in commodities. For every satiable desire, new desires have arisen, and leisure is too costly in terms of these new needs. We could have done otherwise. Nothing prevents us from working a few hours a week and enjoying all the middle-class comforts 1930 had to offer. It is an affordable option by the very definition of “real income,” but we choose not to take it.

This is the baseline argument in a 2008 book of essays by eminent economists looking back on Keynes’s 1930 predictions. Around the baseline, a range of respectable opinion offers variations around it. Some contributions are interesting mainly as examples of Keynes’s continuing ability to provoke frothing at the mouth among the Right. Edmund Phelps’s essay seems to have been ghostwritten by a pompous secretary of the chamber of commerce in a mid-sized town, scattering names from Aristotle to Cellini to Cervantes to Maslow’s hierarchy of needs, giving his sections titles like “Keynes Disdainful of the Quest for Wealth” and “Keynes Blind to the Intellectual Satisfactions in Business Life.” Michele Boldrin and David K. Levine are gracious enough to allow that Keynes was “perhaps not a fool . . . perhaps, indeed, he was brilliant, possibly so much so that he never had to bother with logical consistency and facts.” He was, “we are told” a “giant of economics,” and standing on his shoulders they at least “learned something about how not to theorize about human needs and their determinants.” If this is your kind of thing, you are in luck: technical progress delivered an infinitely elastic supply curve when it gave us the conservative blogosphere.

People may complain about their First World problems, but this shows only a lack of perspective on their part, or an expressed preference for bitching. As Benjamin Friedman points out toward the end of the collection, this way of thinking has a very long history: Adam Smith, in The Moral Sentiments, wrote that “all men, sooner or later, accommodate themselves to whatever becomes their permanent situation.” Alfred Marshall, more than a century later, remarked that “after a time, new riches often lose a part of their charm.”

With standard neoclassical choice theory it is easy to show that people are materially better off than they were a year ago if they could still afford everything they bought back then, and then some — and it is not much of a stretch to put a value on the “then some,” because market prices do it for us. If you can do it from one year to the next, it seems no problem to string the gains together across any number of years. Extend the chain long enough, and it seems you can quantify objectively how much better off the average person is today than the average person in 1930. You might argue that real income as a whole comes with diminishing returns of satisfaction — surely we are not really five times better off. But if that were so, would people not be moving along their indifference curves to enjoy more leisure? Evidently free time is not much more valued relative to the commodities that can be bought with a wage than it was eighty years ago.

Once you start to think about the trend to ever higher living standards, Chicago economist Robert Lucas once said, “It is hard to think about anything else” — anything trivial like stabilization or distribution. If the pie grows fast enough, even the people with the least will see their pieces expand faster than they could hope from a bigger share. Who cares about inequality in the rich world when just about everyone is better off than the upper classes of a hundred years ago and much better off than their contemporaries in poorer countries?

Socialists would surely agree with Keynes that while there is no cure for the human condition, the economic problem can indeed be solved. We would disagree that the solution will emerge smoothly and naturally out of capitalism. Jacobin writers have repeatedly made the case that both equality and liberation from toil should be at the center of the socialist program. But how do we respond to the optimistic and pessimistic liberal positions?

The tendency of capitalism to awaken new wants even as it satisfies old ones has been discussed by Marxists since Marx himself. But rarely have we taken neoclassical choice theory seriously enough to dispute its implications. One exception is G. A. Cohen, deep in his Karl Marx’s Theory of History. He draws a distinction between the pursuit schedule — the standard neoclassical demand schedule, which reveals itself in actual market behavior — and the satisfaction schedule, which orders bundles of goods (including leisure) in terms of the satisfaction the individual would actually get from them — which may be unknown to the individual. What individuals demand is not necessarily best for the individuals.

Neoclassical demand theory abandoned reliance on concepts of “utility” for some good reasons — the impossibility of observing it, measuring it, or even saying clearly what it is. It was replaced by a conception of ordinal preferences: we can talk of what someone prefers to something else but not by how much they prefer it. Cohen’s suggestion looks like a throwback, and one with the unpleasant connotation that the researcher knows better than the individual about what’s good for them. However, the purported measurement of satisfaction had already been smuggled in the back door with the treatment of rising living standards in terms of real income per capita. The relative amounts of money individuals prove themselves willing to pay for each good implicitly becomes a measure of satisfaction. The rigorous neoclassical must admit that choice theory provides the grounds for no such move. All it can say is that in any given period, assuming rationality, everybody chooses the available bundle of goods and leisure which they most prefer.

The theory deliberately says nothing about the motivations for such choices. In practice, people express their preferences for certain things over others not necessarily because they bring more pleasure, but also perhaps out of habit, impulse or to avoid dissatisfaction. The other side of the Cohen argument is to emphasize that these reasons do not originate from some mysterious well deep in each individual’s soul, but arise in a social context which has formed the person’s norms. How the other people around us live affect our expectations for our own lives. It is difficult to resist the pull of norms around us, because we feel not only the satisfactions of the things we have but the lack of the things we could have. To live with the 1930 commodity bundle in 2012 is not to get 1930 satisfactions, but to feel a lack of everything we have come to expect since then — a lack which the extra leisure will not make up for. So hardly anyone chooses to do it.

Finally, capitalism is structured with an extremely strong bias to redeploying productivity gains towards output expansion. The other possibility, increasing free time, “threatens a sacrifice of profit associated with increased output and sales, and hence a loss of competitive strength.” It is no iron law. Differences between Europe and the United States make clear that social norms over vacations and working hours can diverge. But there is certainly a strong tendency towards productivism, which has nothing to do with individual preferences. It is imposed on individual firms by competition and on national policy by macroeconomic considerations: witness the pressure on France’s thirty-five-hour working week, or the drive in Australia to increase labor force participation, tapping hidden reserves of labor-power among mothers of young children, would-be retirees, and the disabled.

As my colleague Peter Frase has argued, drawing on the work of labor economist Lonnie Golden, many workers report a preference to be working fewer hours than they do, but the jobs on offer make it difficult to do without sacrificing income more than proportionately. A gigantic marketing effort aims to intensify feelings of want for commodities, while there are no ads for time off.

Cohen’s distinction between the pursuit schedule and the satisfaction schedule implies a possible gap between what we choose, given our social and historical positions, and what we would ultimately enjoy most. But it hardly entails a paternalistic imposition of new choices: he is simply urging a persuasive effort. It differs from the moralistic “over-consumption” argument in not blaming the individual, by recognizing that the shift will be a collective one or it will not happen at all.

Finally, what of the argument that rising living standards make inequality irrelevant? It follows from the above that inequality is directly relevant, because people’s conceptions of how they could and should live are influenced by the lifestyles of those around them — the standards of the wealthy filter down to the rest of us. But even if you believe that  argument to be bogus, and that living standards arise objectively from goods consumed regardless of context, inequality matters a great deal.

The quantification of growth simply in terms of “real income” misleads by giving the impression of a homogeneous mass of stuff spewed out by the black box of the economy. Ultimately, income is a claim on the produce of our collective labor and on the natural resources we appropriate. Ever increasing abundance — even if it were ecologically sustainable, and even if it did bring ever-rising satisfaction — is no answer to inequality. Why shouldn’t people have their share of what they help to produce?

Further, because it depends on augmenting the efficiency of labor and capital inputs, technical progress is quite uneven in what it cheapens relative to the average wage. Some labor processes are much more susceptible to productivity increases than others. Economic development has meant large changes in relative prices as well as a general rise in real incomes. It is unlikely, for example, that the average person will ever be able to recreate Keynes’s own lifestyle of the 1920s, no matter how high real incomes grow. (At around £230,000 in today’s pounds, Keynes’s average annual income that decade of £5068 would still place him just within the British top 1 percent.) The reason is simply that it involved several large houses in central London, fully staffed with servants. The affordability of servants depends entirely on the ratio between their wage and their employer’s income. The value of large houses in Bloomsbury tends to rise with incomes: all the productivity improvement in the world will not cheapen them, but make them dearer.

More seriously, in places like America where middle and lower-class real incomes have stagnated, households have enjoyed not simply a boring but at least comfortably stable basket of commodities. As Larry Summers fretted in the Financial Times in January, thinking simply in terms of real income misses the impact of major relative price changes:

Measured via items such as appliances or clothing or telephone services, where productivity growth has been rapid, wages have actually risen rapidly over the last generation. The problem is that they have stagnated or fallen measured relative to the price of food, housing, healthcare, energy, and education.

To the extent that people care more about the cost of food, housing and healthcare than appliances and clothing, this is bad news. And the trend is likely to continue, because as manufactures become ever cheaper, demand shifts into areas where productivity growth is much slower. The technological cavalry is becoming weaker. This matters even in countries like Australia, where the real wage has been growing: the cost of food, housing, utilities, petrol, healthcare and education have grown faster.

Thanks to the income effect of the cheapening of other goods, households have not on average had to cut back on necessities, but it is a close-run thing. Many households have reacted by supplying more labor. It is hardly surprising many Australians don’t feel as well-off as the commentariat say they should.

Keynes’s utopian vision of the future was also a  call for restraint in the present. It was a call for moderation and patience addressed to a generation tempted by radical challenges to capitalism:

But beware! The time for all this is not yet. For at least another hundred years we must pretend to ourselves and to everyone that fair is foul and foul is fair; for foul is useful and fair is not. Avarice and usury and precaution must be our gods for a little longer still. For only they can lead us out of the tunnel of economic necessity into daylight.

We can now see how this works out. The same message could be given to us today. William Baumol, in his essay in the 2008 retrospective, pretty much does so: “I can now ask the audience to suppose that real US income will once again increase sevenfold in the next century. Can you imagine what luxuries average-earning Americans will have at their disposal?” Will we then say the economic problem is solved? More likely, it is not a problem that will solve itself.