Thursday, April 12, 2012

"Efficiency’s Promise: Too Good to Be True" "More Jobs Predicted for Machines, Not People"

An invisible thread connects David Owen's The Conundrum ("Efficiency’s Promise: Too Good to Be True") and Erik Brynjolfsson's and Andrew McAfee's Race Against the Machine ("More Jobs Predicted for Machines, Not People"). Both books address real -- and very important -- problems but they both arrive at false conclusions.

The "conundrum," according to Owen, boils down to a lack of commitment driven by conflicting motives about protecting the environment, "Do we honestly care?" he laments at the end, citing George Orwell's observation that, "All left-wing parties in the highly industrialized countries are at bottom a sham, because they make it their business to fight against something which they do not really wish to destroy."

Meanwhile, Brynjolfsson and McAfee prescribe the clichéd panaceas for technological unemployment of education, "flexibility" and entrepreneurship: "Our skills and institutions will have to improve faster to keep up lest more and more of the labor force faces technological unemployment."

The thread that unites the two books is an idea Marx called "the theory of compensation as regards the workpeople displaced by machinery" and Keynes criticized as the doctrine of self-adjustment. William Stanley Jevons described the doctrine as "a principle recognised in many parallel instances." Specifically, with regard to workers displaced by machinery, Jevons observed,
The economy of labour effected by the introduction of new machinery, for the moment, throws labourers out of employment. But such is the increased demand for the cheapened products, that eventually the sphere of employment is greatly widened.
Jevons related the doctrine to the consumption of fuel, arguing, "Now the same principles apply, with even greater force and distinctness, to the use of such a general agent as coal. It is the very economy of its use which leads to its extensive consumption." This rebound effect has become known as the "Jevons Paradox" and is the central argument of Owen's book.

Brynjolfsson and McAfee depart momentarily from the self-adjustment script when they seemingly consign automatic compensation to the past, "As steam power advanced and spread throughout industry, more human workers were needed, not fewer." But with their prescriptive platitudes they return faithfully to the self-adjustment fold albeit, as Keynes satirized, with "creaks and groans and jerks, and interrupted by time-lags, outside interference and mistakes."

The energy efficiency/labor productivity connection is, at least in part, not only a parallel but also an identity. To say that "the same principles apply," as Jevons did, was an understatement. All else held constant, every labor-saving device consumes less energy than the workers it replaced. Gains in energy efficiency mean that the same result can be achieved with less fuel produced by fewer workers in the energy sector.

In an imaginary, perfectly-competitive, frictionless (but one-step-at-a-time) world where final demand for commodities remained unchanged and there were no economies of scale, a one percent increase in energy efficiency would result in a one percent decline in fuel consumption, the demand for which could be met by one percent fewer workers in the energy-supplying industry. Of course such a world couldn't exist outside of thought experiments. But the point of the thought experiment is to tease out the abstract mathematical identity that remains concealed behind the "fog" of real-world contingency.

In a slightly more real world, savings from energy efficiency would make running machines cheaper and thus encourage more substitution of machinery for labor. Job losses wouldn't be confined to the energy sector, which might even expand rather than contract, depending on the relative elasticities of demand for energy and labor. And all this is while still assuming that final demand doesn't change!

Machines are both substitutes for and complements of labor, as Brynjolfsson and McAfee recognize. But the extent to which they perform as one or the other and how far they can vary independently of one another is a riddle has confounded "almost the whole body of organized economic thinking and doctrine" of the last TWO hundred AND THIRTY years. Here is how Keynes framed the problem:
Put very briefly, the point is something like this. Any individual, if be finds himself with a certain income, will, according to his habits, his tastes and his motives towards prudence, spend a portion of it on consumption and the rest he will save. If his income increases, he will almost certainly consume more than before, but it is highly probable that he will also save more. That is to say, he will not increase his consumption by the full amount of the increase in his income. Thus if a given national income is less equally divided, or if the national income increases so that individual incomes are greater than before, the gap between total incomes and the total expenditure on consumption is likely to widen. But incomes can be generated only by producing goods for consumption or by producing goods for use as capital. Thus the gap between total incomes and expenditure on consumption cannot be greater than the amount of new capital that it is thought worth while to produce. Consequently, our habit of withholding from consumption an increasing sum as our incomes increase means that it is impossible for our incomes to increase unless either we change our habits so as to consume more or the business world calculates that it is worth while to produce more capital goods. For, failing both these alternatives, the increased employment and output, by which alone increased incomes can be generated, will prove unprofitable and will not persist.

Now the school that believes in self-adjustment is, in fact, assuming that the rate of interest adjusts itself more or less automatically, so as to encourage just the right amount of production of capital goods to keep our incomes at the maximum level that our energies and our organization and our knowledge of how to produce efficiently are capable of providing. This is, however, pure assumption. There is no theoretical reason for believing it to be true. A very moderate amount of observation of the facts, unclouded by preconceptions, is sufficient to show that they do not bear it out.
Where Keynes addressed an increase in income, Marx approached the same problem from the perspective of a loss of income by displaced workers:
This theory [of compensation] implies that the £1,500 worth of means of subsistence was capital that was being expanded by the labour of the 50 men discharged. That, consequently, this capital falls out of employment so soon as they commence their forced holidays, and never rests till it has found a fresh investment, where it can again be productively consumed by these same 50 men. That sooner or later, therefore, the capital and the workmen must come together again, and that, then, the compensation is complete. That the sufferings of the workmen displaced by machinery are therefore as transient as are the riches of this world.

In relation to the discharged workmen, the £1,500 worth of means of subsistence never was capital. What really confronted them as capital, was the sum of £1,500, afterwards laid out in machinery. On looking closer it will be seen that this sum represented part of the carpets produced in a year by the 50 discharged men, which part they received as wages from their employer in money instead of in kind. With the carpets in the form of money, they bought means of subsistence to the value of £1,500. These means, therefore, were to them, not capital, but commodities, and they, as regards these commodities, were not wage-labourers, but buyers. The circumstance that they were “freed” by the machinery, from the means of purchase, changed them from buyers into non-buyers. Hence a lessened demand for those commodities — voilà tout. If this diminution be not compensated by an increase from some other quarter, the market price of the commodities falls. If this state of things lasts for some time, and extends, there follows a discharge of workmen employed in the production of these commodities. Some of the capital that was previously devoted to production of necessary means of subsistence, has to become reproduced in another form. While prices fall, and capital is being displaced, the labourers employed in the production of necessary means of subsistence are in their turn “freed” from a part of their wages. Instead, therefore, of proving that, when machinery frees the workman from his means of subsistence, it simultaneously converts those means into capital for his further employment, our apologists, with their cut-and-dried law of supply and demand, prove, on the contrary, that machinery throws workmen on the streets, not only in that branch of production in which it is introduced, but also in those branches in which it is not introduced.

The real facts, which are travestied by the optimism of economists, are as follows: The labourers, when driven out of the workshop by the machinery, are thrown upon the labour market, and there add to the number of workmen at the disposal of the capitalists. In Part VII of this book it will be seen that this effect of machinery, which, as we have seen, is represented to be a compensation to the working class, is on the contrary a most frightful scourge. For the present I will only say this: The labourers that are thrown out of work in any branch of industry, can no doubt seek for employment in some other branch. If they find it, and thus renew the bond between them and the means of subsistence, this takes place only by the intermediary of a new and additional capital that is seeking investment; not at all by the intermediary of the capital that formerly employed them and was afterwards converted into machinery. And even should they find employment, what a poor look-out is theirs! Crippled as they are by division of labour, these poor devils are worth so little outside their old trade, that they cannot find admission into any industries, except a few of inferior kind, that are over-supplied with underpaid workmen. Further, every branch of industry attracts each year a new stream of men, who furnish a contingent from which to fill up vacancies, and to draw a supply for expansion. So soon as machinery sets free a part of the workmen employed in a given branch of industry, the reserve men are also diverted into new channels of employment, and become absorbed in other branches; meanwhile the original victims, during the period of transition, for the most part starve and perish.

It is an undoubted fact that machinery, as such, is not responsible for “setting free” the workman from the means of subsistence. It cheapens and increases production in that branch which it seizes on, and at first makes no change in the mass of the means of subsistence produced in other branches. Hence, after its introduction, the society possesses as much, if not more, of the necessaries of life than before, for the labourers thrown out of work; and that quite apart from the enormous share of the annual produce wasted by the non-workers.
As the last paragraph reveals, the difference between Keynes's perspective and Marx's was not as stark as it may at first appear. Although Marx addressed a loss of income by displaced workers rather than a gain by an individual, as in the Keynes passage, "at first" the situation was one where "society" possessed "as much, if not more, of the necessaries of life than before."

There is also a discrepancy that is more apparent than real between Keynes's and Marx's accounts of how the dilemma is resolved -- if it is resolved. Keynes laid emphasis on the setting of an interest rate "so as to encourage just the right amount of production of capital goods," which, however, is not achieved automatically, as the self-adjusting school presumes. Marx stressed the necessity for "new and additional capital that is seeking investment." The key point for both of them, then, is the importance of expansion of credit, not the adjustment of price levels brought about by supply and demand.

Unlike any commodity, credit money is, in principle, infinitely expandable. However, this hypothetical infinity is constrained by an intangible: confidence. Credit thus "hangs upon Opinion; it depends upon our Passion of Hope and Fear..." The converse is equally true: unlike credit money, the production of commodities is not even in principle infinitely expandable. Natural resources and human capabilities are constrained and industrial production must operate within those constraints, despite whatever fantastic rhetorical claims economists may be inclined to dispense about "infinite wants."

Credit, not cheapness of commodities, is what sets in motion the rebound effect that both expands employment notwithstanding the displacement of workers by machines and increases consumption of fuel despite gains in energy efficiency. As pointed out above, these two rebound effects are not independent of each other. They can be stretched but they can't be decoupled. That's the bad news. The good news is that the putative desire for an "unlimited amount of work to be done" makes no sense other than for the necessity of raising revenues to service debt.

See also: Unemployment, Cornucopia and the Identity Crisis of "Green Growth" and Further Thoughts on the Use of Machines: a statistical science fiction.


  1. Tom: You make an interesting point.

    I think it would be clearer if you mentioned economic growth in the final paragraph. Credit "sets in motion" the rebound effect, as you say, but it sets it in motion by encouraging more consumption, investment, and/or government spending, which leads to more economic growth.

    When the economy slows, we lower interest rate to promote growth. Because of that growth, higher productivity of labor does not lead to higher unemployment, and increased energy efficiency does not reduce energy consumption.

    Do I understand your point correctly?

    (I actually do not agree about energy consumption, because I think the share of energy in the economy can continue to decline with energy efficiency. I agree about unemployment - the jobs are created precisely because the goal of our economic planning is to keep stimulating the economy until employment increases to acceptable levels, but we don't have a similar goal for energy consumption.)

  2. Charles: "Do I understand your point correctly?"

    Yes. The point that you make at the end is uncontroversial, I think. The share of energy in the economy can continue to decline indefinitely. But the question of the absolute quantity of consumption is inseparably linked to the goal of achieving conventional "full employment" -- that is to say rejecting work time reduction as part of the policy mix. When I say "inseparably", I mean literally that when an average job is created, an average amount of energy consumption is part of the package deal. Obviously some jobs are more energy intensive than others.

    In principle it would be possible to create only "low energy consumption" jobs. But they are unlikely to be the jobs that the market will spontaneously generate, outside of (on average) low paying jobs. This is not to say that there wouldn't be exceptions. But they would be... exceptions.