Relevant and even prescient commentary on news, politics and the economy.

## Sandwichman in the FT

##### Sandwichman in the FT

Financial Times: “The minimum wage wars are heating up: A new study fails to prove its claim that Seattle wage floor hurts workers” by Martin Sandbu, at Free Lunch on FT Alphaville

First, the numerical result struggles to pass an intuitive “smell test”. As the Angry Bear blog [cross posted at EconoSpeak!] points out, employment in Seattle was booming throughout the period: average wages increased by 18 per cent (!) in the time covered by the study; as did the number of hours worked at all wage rates. It is important to note that the researchers have data on jobs, not on individual workers — so even if there are fewer low-paid jobs than before, it does not follow that workers have lost as many jobs rather than moved into better ones.

Fifth, there is a simple arithmetical issue with the methodology. If “low-paid” is defined as below a fixed threshold wage (\$19/hour in this study), higher average wage growth in one group relative to an initially similar control group will necessarily take more people out of the low-wage bracket. Such wage bracket creep means that the higher wage growth will “cause” a “loss” of low-wage work through mechanical arithmetics even if nothing causal is happening (indeed even if everyone stays in the same job).

## Arithmetic is Hard: Wage-Bracket Creep

There has been a lot of very good critique of the methodology of the University of Washington’s study of the minimum wage increase in Seattle. However, I want to repeat and emphasize a very simple point that jumps out.

A static low-wage cutoff point, whether it be \$19 or \$100, automatically reduces the size of the treatment group (Seattle) if wages in the treatment group are increasing faster than the wage of the control group.

This is not erudite statistical methodology. This is simple arithmetic. If the wages in the treatment group increase at a higher percentage rate than the wages in the control group, more workers are lifted above the \$19 threshold in the treatment group than in the control group.  This is true if the treatment group and the control group are otherwise absolutely identical. This is what I call wage-bracket creep. The extremely simplified example below shows how this looks, the yellow cells represent workers whose jobs and hours would be “lost” (to the study) as they pass the \$19 threshold:

See how much worse off the treatment group is than the control group? The yellow cell occupants haven’t lost their jobs, they have simply been excluded from their respective groups because their wage now exceeds the static cutoff amount.

Of course, I wondered if the study authors could be making such a simple arithmetic mistake. So I reached out to one of the authors, who generously replied but appeared to confirm that they relied on a static threshold. I say appeared because some of the replies were, shall we say, “ambiguous” but did not disclaim use of a static threshold when I sought explicit confirmation or denial.

## Seattle Minimum Wage

Words cannot describe the torment experienced by the data before they confessed what the University of Washington team got them to confess. I can only urge readers with an open mind to study Table 3 carefully. The average wage increase, from the second quarter of 2014 to the third quarter of 2016, for all employees of single site establishments was 18 percent. Eighteen percent! That is an annual increase of almost 8 percent. For two and a quarter years in a row. Not bad. And the number of hours worked of ALL employees of single site establishments? Up 18 percent in a little over two years. That too is an increase of almost 8 percent per annum.
Now multiply that wage by those hours and the total payroll for all employees rose 39.5 percent over the course of nine quarters. An annual rate of increase of 17.5 percent. These are BIG numbers. They are freaking HUGE numbers.
It must have taken a team of at least six academics to extract a 9.4 percent decline in hours from the 86,842 workers (out of a total of 336,517) earning under \$19 dollars an hour at these single-site establishments. Look at the Table and weep.

Now, as I mentioned in a comment on Peter’s post, bracket creep alone could do away with at least 7 percent of the missing hours of workers earning under \$19 and hour. That is unless we assume that everyone making between about \$18 and \$19 got approximately zero wage increases while the rest of the crew were getting 10 percent raises. Look at the God damned table. This isn’t rocket science.

Minimum Wage Increases, Wages, and Low Wage Employment: Evidence from Seattle

## Asking The Man in the Street: Research v. Rhetoric

Richard Layard, How to Beat Unemployment, 1986:

“If you ask the man in the street (not Wall Street) what has caused our unemployment, nine times out of ten he will say that it is machines displacing people. In fact for this reason he is often deeply pessimistic about whether we could ever have full employment again.”

Dear Professor Layard,

In your 1986 book, “How to Beat Unemployment,” you wrote: “If you ask the man in the street (not Wall Street) what has caused our unemployment, nine times out of ten he will say that it is machines displacing people. In fact for this reason he is often deeply pessimistic about whether we could ever have full employment again.”

I am curious. Did you ask “the man in the street”?

Cheers,

The Sandwichman

## “If There Is Any Such Thing”: Why read Hoxie on theory?

Unionists are not theorists; unionism is an eminently practical thing. — Robert F. Hoxie

Theory and trade unionism are almost contradictory terms. — Edward M. Arnos

In accordance with this theory it is held that there is a certain fixed amount of work to be done… David F. Schloss

Paul Samuelson once wrote that it takes a theory to kill a theory. He didn’t say it had to be a better theory. What would it take to kill a theory that never was?

The Sandwichman’s summer project has been to consolidate my research and blog posts on the lump of labor from the last ten years into something like — and yet unlike — “the archaic stillness of the book.” Sometimes, when cross checking old sources, new sources spring up out of the archives and one of the most astonishing was Robert Hoxie’s commentary on what he called “fixed group demand theory.

The term appears elsewhere only in a few sources: a dictionary entry on the lump-of-labor theory in What’s What in the Labor Movement: A Dictionary of Labor Affairs and Labor Terminology (1921) by Waldo Ralph Browne, in The Settlement of Wage Disputes (1921) by Herbert Feis, whose discussion mainly centered on Hoxie’s analysis, and Warren Gartman made a brief, parenthetical reference to the theory in a 1950 report on Longshore Labor Relations on the Pacific Coast, 1934-50. By  far the most substantive treatment of fixed group demand theory was in Edward M. Arnos’s 1915 article, “An Interpretation of the Working Rules of the Carpenters’ Unions of Chicago.” Arnos was a doctoral student at the University of Chicago at the time when Hoxie was conducting his research on organized labor’s views on the Taylor method (“scientific management”) and Hoxie engaged his students in the research project. Hoxie also wrote on the concept of fixed group demand previously without using the terminology. I reproduce both Arnos’s and Hoxie’s discussion below.

Hoxie’s novel method was to ask people why they did something. Appendix II of his Trade Unionism in the United States contains an 18 page outline and summary of  the “students’ report on trade union program.” Appendix VIII of Hoxie’s Scientific Management and Labor presents over 100 pages of questions used by Hoxie in that study. In the latter study, Hoxie prepared preliminary statements based on extensive reviews of the literature, summarizing the labor claims made by scientific management and the objections to scientific management by unions. He then circulated the summaries to proponents of scientific management and labor leaders, respectively, for their revision and approval. By his own account, Arnos’s investigation followed similarly thorough methods.

The point of such rigorous investigation was not to vindicate or invalidate the theories in question but to examine their claims in the light of experience. The outcome was not a triumph for one theory and a defeat for another — a sorting into economic laws and economic fallacies — but an assessment of the extent to which each of the competing theories had merit and their respective limitations. Hoxie operated in the spirit of ethical debate as latter proposed by Anatol Rapoport.

There are two aspects of Hoxie’s discussion of fixed group demand theory I would like to emphasize. The first is his explanation of unions’ restrictive rules as pragmatic, opportunistic measures adopted locally and retained through trial and error rather than in accordance with some overarching “theory” of how the economy works.

The second is a subtle but devastating critique of the pretension of economic theory to apply simultaneously to both the universal long run and to local immediacy. In Trade Unionism in the United States, Hoxie rhetorically affirmed the validity of the classical economic analysis “when applied to society as a whole, if there is any such thing, and in the long run” while objecting that for workers, “there is no society as a whole, and no long run, but immediate need and rival social groups.” A few years later, Maynard Keynes echoed the assessment that “this long run is a misleading guide to current affairs.”

In a brief essay on “The Theory of Unionism: Principles of Uniformity,” Hoxie thinly muzzled a searing critique of economic orthodoxy by presenting it as the employer’s naïve conclusion:

Apparently it rarely occurs to the employer that this analysis is not complete. Having assumed that definite laws determine the manner in which income is shared among the productive factors, he apparently concludes, somewhat naively, that just as the laborers in society will in the aggregate profit by increase in the social income, so also will the laborers in any individual establishment profit by increase in its income.

Hoxie’s “employer” is simply parroting the old “Say’s Law” truism that, as Alfred Marshall put it, “the demand for work comes from the National Dividend; that is, it comes from work: the less work there is of one kind, the less demand there is for work of other kinds; and if labour were scarce, fewer enterprises would be undertaken.” Marshall’s “national dividend” was an updated and sanitized label for what a decade earlier in The Economics of Industry, he still referred to as the “wages-and-profits fund,” which was too close to the discredited wages-fund to escape scrutiny. The bottom line, though, remained that “there is no such thing as general overproduction.” There is only ever “commercial disorganization; and that the remedy for it is a revival of confidence.”

The chief cause of the evil is a want of confidence. The greater part of it could be removed almost in an instant if confidence could return, touch all industries with her magic wand, and make them continue their production and their demand for the wares of others. If all trades which make goods for direct consumption agreed to work on and to buy each other’s goods as in ordinary times, they would supply one another with the means of earning a moderate rate of profits and of wages.

Although Marshall didn’t mention this, it follows from his analysis of the impossibility of overproduction that in a crisis entrepreneurs commit the lump of confidence fallacy (or the fallacy of the fixed Confidence-fund). If only they understood how the “magic wand” of confidence works. Nor did Marshall happen to mention that the employers’ stock remedies for hard times of cutting wages and/or laying off workers simply reflects their obliviousness to the fact that “there is no such thing as general overproduction.”
Why worry about what Alfred Marshall wrote or didn’t write 136 years ago? Because it is the dogma echoed down through the ages, such as in this 1986 gem by Richard Layard, How to Beat Unemployment:

The one fatal heresy in economic analysis is to take output as given. That is the ‘lump of output’ fallacy. You must always have a theory of how output is determined and you must never say, ‘Higher output per worker reduces employment, because it reduces the employment needed to produce a given output’. Likewise you must never say ‘More people cause unemployment’, unless you can explain why output will not grow.

Along with Richard Jackman, Layard recycled the archaic and bogus analysis the next year in a pamphlet, “Innovative Supply-Side Policies to Reduce Unemployment” and yet again in 1991, adding Stephen Nickell to the team in Unemployment: Macroeconomic Performance and the Labour Market. This “analysis” became the basis of Tony Blair’s and Gerhard Schroeder’s miserable “New Supply-Side Agenda for the Left.” Jonathan Portes’s proudest accomplishment was explaining the lump-of-labour fallacy to successive cabinet ministers. And so the magic wand of confidence waves on…
But enough about the magic confidence wand (if there is any such thing). Below is some true grit from Hoxie and Arnos.
Robert F. Hoxie “The Theory of Unionism: Principles of Uniformity,” in Readings in Current Economic Problems, 1915

The third charge against the unionist which we have undertaken to examine states that while he is struggling for increase of wages he is at the same time attempting to reduce the efficiency of labor and the amount of the output. In other words, while he is calling upon the employer for more of the means of life he is doing much to block the efforts of the employer to increase those means.

There is no doubt that this charge is to a great extent true. In reasoning upon this matter the employer, viewing competitive society as a whole, assumes that actual or prospective increase in the goods’ output means the bidding-up of wages by employers anxious to invest profitably increasing social income. It follows that in competitive society laborers as a whole stand to gain with improvements in industrial effort and process. In the case of the individual competitive establishment it is clear that the maximum income is ordinarily to be sought in the highest possible efficiency, resulting in increased industrial output. At least this is true where there are numerous establishments of fairly equal capacity producing competitively from the same market. Under such circumstances the increased output of any one establishment due to “speeding up” will ordinarily have but a slight, if any, appreciable effect on price. Each individual entrepreneur, therefore, is justified in assuming a fixed price for his product and in reckoning on increase of income from increase of efficiency and industrial product. Apparently it rarely occurs to the employer that this analysis is not complete. Having assumed that definite laws determine the manner in which income is shared among the productive factors, he apparently concludes, somewhat naively, that just as the laborers in society will in the aggregate profit by increase in the social income, so also will the laborers in any individual establishment profit by increase in its income.

To this mode of reasoning, and to the conclusions reached through it, the unionist takes very decided exceptions. To the statement that labor as a whole stands to gain through any increase in the social dividend he returns the obvious answer that   labor as a whole is a mere academic conception; that labor as a whole may gain while the individual laborer starves. His concern is with his own wage-rate and that of his immediate fellow-workers. He has learned the lesson of co-operation within his trade, but he is not yet class-conscious. In answer to the argument based on the individual competitive establishment he asserts that the conditions which determine the income of the establishment are not the same as those which govern the wage-rate. Consequently, increase in the income of the establishment is no guarantee of increase of the wage-rate of the worker in it. Conversely, increase in the wage rate may occur without increase in the income of the establishment. Indeed, in consequence of this non-identity of the conditions governing establishment income and wage-rate, increase in the gross income of the establishment is often accompanied by decrease in the wage-rate, and the wage-rate is often increased by means which positively decrease the gross income of the establishment.

The laborer’s statements in this instance are without doubt well founded. The clue to the whole situation is, of course, found in the fact that the wage-rate of any class of laborers is not determined by the conditions which exist in the particular establishment in which they work, but by the conditions which prevail in their trade or “non-competing group.” With this commonplace economic argument in mind, the reasonableness of the unionist’s opposition to speeding up, and of his persistent efforts to hamper production, at once appears.

“An Interpretation of the Working Rules of the Carpenters’ Unions of Chicago,” Edward M. Arnos, 17th Report of the Michigan Academy of Science, 1916

Theory and trade unionism are almost contradictory terms. The trial and error method of testing rules, the ever changing conditions of the trade, the large number of men concerned in the agreement, the different nationalities represented in the union personnel, and the triennial agreements have left the carpenters’ rules marked as if they are in a process. The constant changes in the agreements evince the carpenters’ struggle to get control of the trade, first by one method or rule and then by another. This trial and error method has removed at least the trace of theory as a controlling force in the construction of the joint agreement. Journeymen are seldom conscious of any underlying theory of the rules in explaining their demands, methods, policies, and aims. Although the development of the rules has been free from the control of theorists, development has been in harmony with certain theories of business and human relationship. The theory of standardization, the theory of undercutting, the fixed group demand or lump labor theory, and the standard of living theory, are vital to the carpenters’ rules. Journeymen may not realize the presence of any theories, nevertheless the officers interpret the rules in the light of these theories. To illustrate, one business agent said the rule prohibiting journeymen from taking their tools on the job before they were employed was to prevent men from gathering around the places of employment prepared to work, because the employers used their presence to intimidate the journeymen on the job; i. e., according to his theory of life, men who were out of employment would place themselves where they could underbid their fellows who were employed. To illustrate the underlying force of their fixed group demand theory, one of the officials said that they were in favor of a raise of wages to 70 cents per hour because there was a certain amount of work to be done and the carpenters could get 70 cents per hour as well as 65 cents. Thus consciously or unconsciously, the carpenters supported all of their rules by some of their theories of life. Let us consider these theories and their significance after careful analysis.

…

The presence of an unemployed group and their theory of undercutting necessitates standards and uniform units of measurement. Thus the first of the hypothetical theories is accounted for. This assumption of the constant over-supply of labor also presupposes that there is a fixed group demand for labor, thus their theory of a fixed group demand or “lump of labor” theory. The third theory to be considered is that of the fixed group demand. This fixed group demand is usually approached through the desire to share work among their members, which they accomplish by limiting the supply of labor. Their rules on apprenticeship so limit the number of apprentices that it is said that only the sons of the most prominent journeymen are indentured. The number of apprentices range from one to two per cent of the number of journeymen. Rushing and excessive work have the same effect upon the supply of labor, through the limitation of the amount of work to be done in a certain time. The eight hour day and holidays limit the number of working hours and thus limit the labor supply. The fixed group demand theory is supported by their experience of unemployment. The leaders contend that the unemployed are as numerous under low wages as they are under high wages. The hypothesis is that there is a certain amount of carpentering to be done in Chicago. This is fixed by the number of persons who live there. To quote an official, “a man wouldn’t live in a tent if wages were high nor in two houses if they were low.” Of course this opinion would not bear strict interpretation nor do they claim that for it. The constant increase in the scale of wages and the accompanying decrease in unemployment in the trade are often cited as proof of their hypothesis. Their wage slogans, “high wages breed high wages,” “no wage reductions,” “cheap wages make cheap men,” and “get more now,” have their origin in this group of facts.

## Hoxie on “Fixed Group Demand Theory” (the “lump of labor”)

From Robert F. Hoxie, Trade Unionism in the United States, 1917:

There is much scorn of unionists by economists and employers because of this lump of labor theory with its corollaries. This scorn is based on the classical supply and demand theory and its variants. Supply is demand. Increased efficiency in production means an increase of social dividend and increased shares, which in turn increase production and saving. Therefore, the workers cut off their own noses when they limit output or limit numbers. The classical position is undoubtedly valid when applied to society as a whole, if there is any such thing, and in the long run. But the trouble is that, so far as the workers are concerned, there is no society as a whole, and no long run, but immediate need and rival social groups.

The fixed group demand theory is as follows: The demand for the labor of the group is determined by the demand for the commodity output of the group. The community—wealth and distribution remaining the same—has a fairly fixed money demand for the commodities of a group. It will devote about a given proportion of its purchasing power to these commodities, that is, if the prices of the group commodity are higher, it will buy less units and vice versa, but expend about the same purchasing power. Therefore, the demand for the labor of the group, profits remaining the same, is practically fixed, and increasing the group commodity output means simply conferring a benefit on the members of other groups as consumers without gain to the group itself. Therefore, to increase the efficiency and the output of the group will not increase the group labor demand and group wages. Decreasing the efficiency and output of the group will not decrease the group labor demand and the group wage.

Increasing the number of workers tends to decrease their bargaining strength relatively and to lower the total wage and the wage rate. Increasing the efficiency and the output of the workers is equivalent to increasing the group labor supply, and so tends to lower the group wage and the wage rate. Decreasing the number of workers tends to increase their bargaining strength relatively and so to increase the group wage and the wage rate. Decreasing the efficiency and output of the workers tends to increase their bargaining strength relatively and so to increase the group wage and the wage rate. The introduction of labor saving devices is equivalent to increasing the labor supply and so lowering the wage rate. Limitation of output through shorter hours, etc., i.e., decreasing the supply of labor, increases bargaining strength and tends to increase the wage. Strikes and trade union insurance funds are means of temporarily withdrawing labor supply and so of increasing bargaining strength and increasing wages. In practice the group demand theory is simply the application by the unions of the principle of monopoly, admittedly valid. But this theory only in part explains union efforts to limit both individual and group efficiency and output and to limit numbers. These policies in part rest on other theories and considerations.

Robert F. Hoxie committed suicide on June 22, 1916. For an overview of his important but neglected contribution to economic thought see Charles R. McCann Jr. and Vibha Kapuria-Foreman, “Robert Franklin Hoxie: The Contributions of a Neglected Chicago Economist” Research in the History of Economic Thought and Methodology, Volume 34B, 2016.

## “One Hundred Percent” Fake News

##### “One Hundred Percent” Fake News

No, uh-uh, false, wrong.

Trump: I’m ‘100 percent’ willing to testify under oath — The Hill-2 hours ago

Trump: I’m willing to testify under oath about Comey claims — CNN-3 hours ago

Trump willing to testify to counter Comey under oath — Talk Media News-35 minutes ago

Trump Says He’d Testify Under Oath About Comey — Featured-The Atlantic-2 hours ago

Donald Trump did not say he was one hundred percent willing to testify under oath. He replied to a question about testifying with a word salad of obfuscation that contained the phrases “one hundred percent” and “under oath” but did not connect the two in any coherent way. Read the eff’in’ transcript:

Trump’s “one-hundred percent” is free floating. His two uses of the phrase “under oath” indicate that a misinterpretation or pretended misinterpretation of the question as being whether he asked Comey to pledge his loyalty under oath. So his “one hundred percent” is simply a one hundred percent denial that he demanded that Comey pledge allegiance to him under oath. There is no commitment in the exchange to testify under oath.

Of course, even if Trump had committed one hundred percent to testifying under oath, there would be no way to compel him to honor his commitment and he almost certainly would not do so.

Trump will not testify under oath and he will not release his tax returns.

## “It Depends on How We They Value Time”

Peter Dorman calls attention to a NYT Upshot column by Neil Irwin about the cost of climate change. For Irwin, the question can be framed as a matter of discounting, “A dollar today is worth more than a dollar tomorrow and a lot more than a dollar in 100 years. But what discount rate you set determines how much more.”

As Irwin admits, the discount rate is a “business concept.” His conclusion, then, follows exclusively from a business concept of “how, as a society, we count the value of time.” Why are we compelled, as a society, to count the value of time in accordance with the business concept of discounting? Because there is no other concept of time? No, there are other concepts of time. More specifically, there is a concept of time directly opposed to and critical of the business concept of time. Labor time.

What discounting is to the business concept of time, alienation is to the labor concept of time. Alienation refers not to “feelings” of alienation but to the sale of one’s own time — and consequently autonomy — to another.

For every human being — as for the wage worker — there are 24 hours in a day,  168 hours in a week, 8760 or 8784 hours in a year. These are fixed amounts. You can’t put it in a bank and get it back in 20 years with interest. You can’t take it with you and you can’t convey it to your heirs in a will. Today is here today and gone tomorrow.

The discount rate concept has nothing to do with the qualitative experience of time by humans and everything to do with the quantitative accumulation of money by property owners. Framing the cost of climate change as a contest between different discount rates is totalitarian. We live in a totalitarian society in which the non-business concept of time is invisible. Neil Irwin sounds like a thoughtful person. It simply didn’t occur to him that there was any other relevant concept of time than the business concept.

That is why the climate is changing. And that is why not enough will be done about it. Because it all depends on how capital values time.

## Fighting Zombies with Zombies

##### Fighting Zombies with Zombies

Larry Mishel and Josh Bivens enlist zombie government policy ponies in their battle against “the zombie robot argument“:

Technological change and automation absolutely can, and have, displaced particular workers in particular economic sectors. But technology and automation also create dynamics (for example, falling relative prices of goods and services produced with fewer workers) that help create jobs in other sectors. And even when automation’s job-generating and job-displacing forces don’t balance out, government policy can largely ensure that automation does not lead to rising overall unemployment.

The catch here is that the displacement of workers by technology and the investment that re-absorbs workers displaced by technology are largely, but not entirely, independent factors. “Government policy” in the quoted paragraph is just another name for investment. Hans Neisser observed in his 1942 article on technological unemployment that “it is impossible to predict the outcome of the race between the two [investment and displacement] on purely theoretical grounds.”

The conclusion is inevitable: there is no mechanism within the framework of rational economic analysis that, in any situation, would secure the full absorption of displaced workers and render “permanent” technological unemployment in any sense impossible.

The “robot apocalypse” is neither impossible nor inevitable. It is probably unlikely, but unlikely things do happen, especially when people become complacent about the impossibility of unlikely things happening.

## Output Optimum and the Roller Coaster of Immiseration

Following up on my post from two weeks ago, Immiseration Revisited, I built a spreadsheet replica of the marvelous Chapman diagram. In addition to lines on the page, the replica provides me with tables of numbers that I can add, subtract, multiply and divide in accordance with the conceptual logic of the diagram.

The chart below shows the results of some of these calculations. The red curve graphs cumulative gross “output” and green curve subtracts the value of foregone leisure and the pain cost of fatigue and wear and tear from output to calculate net “income” (green). The length of each vertical line measures the values of output and income, respectively for a work week of the length indicated by the scale on the x-axis.

 “Big Dipper”: the Roller Coaster of Immiseration

I have set the hypothetical “output optimum” work week at 48 hours in deference to the diagram’s 1909 vintage. Assuming such an optimum and taking the conceptual diagram’s proportions literally, the ideal length of a work week for a laborer would be 36 hours. That is the point at which the value of foregone leisure and the pain cost of additional work begin to outweigh the additional earnings from the longer week. A workweek of 40 hours marks the threshold beyond which the value of foregone leisure alone exceeds the additional wage earnings.

If the optimal output workweek was 40 hours, the corresponding ideal length of workweek for the worker would be 30 hours, again assuming the reasonableness of the diagram’s proportions. There is, of course, only impressionistic evidence for the general shape of the curves and not for the accuracy of the proportions depicted. Nevertheless, the derived calculations indicate a steep acceleration of the discrepancy between output and worker welfare beginning well in advance of the output optimum.

Calculations based on the diagram suggest that by working 34 percent more hours per week, the employee can look forward to “enjoying” 29 percent LESS net benefit. If the actual cost to workers of working longer is even half or a third of those estimates, this still would represent a significant deviation not only from what Lionel Robbins dismissed as “the naïve assumption that the connection between hours and output is one of direct variation” but also from the equally indefensible premise of a consistently proportional relationship between work effort and reward.

(Most) Economists Balk

In a recent article, “Whose preferences are revealed in hours of work,” John Pencavel noted the “radical change in economist’s thinking about working hours” following the 1957 publication of H. Gregg Lewis’s article, “Hours of Work and Hours of Leisure,” Earlier textbooks attributed reductions in hours to pressure from trade unions, either directly through collective bargaining or by legislation promoted by organized labor. The earlier textbooks also addressed the effect that hours of work have on productivity, with reductions in hours usually leading to increases in hourly output and sometimes even to “no decline in total daily output.”

In later textbooks, the orthodoxy followed Lewis’s explanation that workers choose their own hours, based on their preferences for income or leisure. The connection between output and shorter hours vanished, as did the role of trade unions in achieving reductions of working time. But, Pencavel wondered, “If ’employers are completely indifferent with respect to the hours of work schedules of their employees,’ [as Lewis had posited] why did employers oppose so resolutely workers’ calls for shorter hours?”

In a footnote, Pencavel also mentioned that in Lewis’s 1957 model, employers face no obstacle “to replacing shorter hours per worker with more workers.” This is an interesting point because many economists’ arguments against the employment potential of shorter working time rest on claims that workers and hours are not suitable substitutes. That conclusion is reached by smuggling back in the output/hours relationship concealed in a Cobb-Douglas production function with the Robbins/Hicks “simplifying assumption” that the current hours of work are optimal for output, so that any reduction of hours would result in a reduction of output. It is difficult to imagine how both of these things can be true at the same time.

Although the earlier textbooks and economists acknowledged the connection between hours of work and output, most were silent on the discrepancy — or at least the magnitude of the discrepancy — between an output optimum and worker welfare. Cecil Pigou, Philip Sargant Florence, Lionel Robbins, John Hicks and Edward Denison treated the output optimum as the economic ideal. Richard Lester and Lloyd Reynolds, authors of “institutionalist” labor economics textbooks, showed more sympathy to trade union arguments but did not emphasize the discrepancy between the output optimum and worker welfare.

Sydney Chapman clearly distinguished analytically between worker welfare and the output optimum but his presentation was obscured by digressions that dwelt on shift-work as a palliative and on the philosophical necessity of paying more attention to the non-tangible aspects of culture. Clyde Dankert clearly distinguished between the output optimum and worker welfare but had the rather eccentric view that although “maximization of worker satisfactions” rather than output should be the social objective, shorter hours would have to be postponed “in view of the current cold war situation.” Only Maurice Dobb clearly and concisely stated what was at stake (although he left out the increasing value of leisure):

…trade unionists in the nineteenth century were severely castigated by economists for adhering, it was alleged, to a vicious ‘Work Fund’ fallacy, which held that there was a limited amount of work to go round and that workers could benefit themselves by restricting the amount of work they did. But the argument as it stands is incorrect. It is not aggregate earnings which are the measure of the benefit obtained by the worker, but his earnings in relation to the work he does — to his output of physical energy or his bodily wear and tear. Just as an employer is interested in his receipts compared with his outgoings, so the worker is presumably interested in what he gets compared with what he gives. A man who works longer hours or is put on piece-rates, and increases the intensity of his work as a result, may earn more money in the course of the week; but he is also suffering more fatigue, and probably requires to spend more on food and recreation and perhaps on doctor’s bills.

To compare “what s/he gets” with “what s/he gives” requires above all some way of estimating the value of what is given relative to what is being received. One may even suggest that constructing those estimates was the job economists should have been doing instead of castigating trade unionists and other advocates of shorter hours for adhering to a vicious “lump-of-labor” fallacy. Heck of a job, economists!