Liberal Party, labour market reform and the unions’ minimum wage myth
Gerard Jackson
There is nothing new in the unions’ ferocious and dishonest opposition to the government’s labour market reform proposals. We got similar nonsense in late 1996 when they pushed for more job-destroying increases in labour costs. To justify their claims they used a 1992 US survey that purported to show that a rise in minimum wages actually raised the demand for labour.
This led unioncrats Grant Belchamber and Tim Harcourt (members of the ACTU’s Research Section) to write in The Australian Financial Review that “orthodox theory” has been over thrown [by the Card-Krueger survey] and that effective minimum wages are “consistent with continuing . . . employment growth”. They naturally followed with the claim that the alleged findings of the American survey supported “the ACTU’s Living Wage claim”. None of this is true.
To overturn a theory one must first understand it. The case against effective minimum wages (effective means priced above the market clearing level) rests on the marginal productivity theory of wages. This theory explains the demand for labour in terms of the value of its product; that is, labour services are priced according to consumer evaluation.
If consumers value a person’s product in a particular line of production at $400 a week, then pricing him at $440 a week will put him out of work. This goes for any factor. Therefore, regardless of Belchamber and Harcourt’s mischievous claim to the contrary, the services of labour are valued in exactly the same way as the services of any other factor. From this it follows that when labour is overpriced relative to the value of its product unemployment will emerge.
On the other hand, when it is under priced a shortage will appear. This pair claimed that “hard evidence to support the theory has always been lacking.” In other words, labour costs are indeterminate. Yet at the time they made their absurd assertion Australia had 8.5 per cent registered unemployment, and rising, plus 27 per cent youth unemployment.
If employment rose in a particular area after the minimum wage has been increased then demand was already rising. For example, the minimum wage hike that Card and Krueger surveyed proved ineffective in Manatee County, Florida, because the demand for labour had already driven employment down to below 4 per cent, and the labour market was is tightening. This situation underlined the fact that for the minimum wage to be effective it must be set above the market rate. This leaves us, then, with flawed data and search for more “hard evidence.”
Let us start with a little medieval history. When the Black Death struck England in 1348 it rapidly slashed the labouring population. This led to a rapid rise in real wages as employers competed against each other to hire labour, despite a determined effort by the Crown to keep wages down to their “historic level.” By cutting the working population the plague increased the marginal value of the survivors’ product by raising the ratio of land and capital to labour, just as marginal productivity theory would predict.
The theory also predicts that those countries that invest more per head of the population will enjoy higher real wage growth than those who do not. Lo and behold, this is just what happened in Hong Kong, Taiwan, Singapore et al. It is also what happened in England during the nineteenth century when the output of consumerables expanded by 1600 per cent, the working population grew by 400 per cent and real wages quadrupled. But still, according to these union quacks there is no “hard evidence.”
The experience of black American provides a bleak and striking case against effective minimum wages. Charles Frederick Roos, a director of the NRA (National Recovery Act), estimated that its minimum-wage provisions put “about 500,000 Negroes on relief in 1934.” Roos was highly critical of the NRA’s job-destroying wage codes. However, the war put America back to work. In 1948 the minimum wage was 40 cents an hour and the unemployment rate for white males age 16 to 7 was just 9.4 per cent, while for black males it was 10.2 per cent.
These unemployment rates remained more or less steady until the minimum wage was raised from 75 cents to $1 an hour ( a 33 per cent increase) in 1956. Within two years unemployment for black youths had leapt to 24 per cent and 14 per cent for whites. By 1995, meddling politicians had raised the respective unemployment rates to 37 per cent and 15.5 per cent. Congress did this by continually raising the minimum wage. Of course, this is not what Belchamber and Harcourt call “hard evidence.” (Incidentally, it has not gone unobserved that the employment rate for black males tends to move in tandem with changes in the real minimum wage).
Then there is the 1981 congressionally-mandated Minimum Wage Study Commission which concluded that a 10 percent increase in the minimum wage would reduce teenage employment by 1 percent to 3 percent. The estimate in studies by David Neumark of Michigan State and William Wascher of the Federal Reserve Board, and Kevin Murphy of the University of Chicago, and Donald Deere and Finis Welch of Texas A&M confirmed the Commission’s estimate. The Commission also concluded that the minimum wage rises between 1977 and 1981 destroyed 644,000 jobs that would have gone to teenagers.
(Incidentally, during the 1980s, the House Democratic leadership suppressed a Congressional Budget Office study predicting the loss of up to 500,000 jobs if the minimum wage was increased to $5.05. The Democrats sent the study back, demanding the new version omit all references to the bill’s potential unemployment effects. Grant Belchamber pulled a similar stroke with an OECD report*)
In 1983 the U.S. General Accounting Office surveyed earlier studies on the unemployment consequences of minimum wage laws; it “found virtually total agreement that employment is lower than it would have been if no minimum wage existed.” Research by Donald Deere and Finis Welch of Texas A&M and Kevin M. Murphy of the University of Chicago found that groups with larger proportions of low-wage workers suffered significant employment losses following the 1990 and 1991 minimum wage increases. Just as marginal productivity theory would predict.
A 1995 study by Kevin Lang of Boston University concluded that raising minimum wages reduces job opportunities for low-skilled adults. In keeping with “orthodox” economic reasoning and supported by empirical evidence he found that higher minimum wages induces more and better qualified individuals to apply for minimum wage jobs thus denying marginal workers employment.
He calculated that 1990-91 minimum wage jump to $4.25 and hour cut jobs for these workers by 9 per cent. Lang’s research confirmed the findings of Bruce Fallick, UCLA, and Janet Currie, MIT. Their 1993 study found that teenagers who had been working between the old and new minimum rates were three to four per cent more likely to lose their jobs than teenagers already paid above the minimum rate.
However, according to the brilliant Belchamber and Harcourt, Card and Krueger had completely overthrown marginal productivity theory and discredited all previous economic studies and masses of empirical evidence that found a negative relationship between minimum wages and employment, despite the fact that such economic luminaries as Brozen, Tobin and Samuelson had attacked minimum wages for pricing teenagers out of work.
As Samuelson wrote: “What good does it do a black youth to know that an employer must pay him $2 per hour if the fact that he must be paid that amount is what keeps him from getting a job?” (Economics, 10th edition, 1976)
The result of Card/Krueger’s study, for which so much had been claimed, was based on crummy data. To begin with, they only looked at major fast-food chains. Now economic theory does not claim that an effective increase in the minimum wage will cause a uniform fall in the employment of marginal workers. Thus a survey of supra-marginal firms could result, depending on the size of the increase, in the apparent paradox of employment rising even after the minimum wage has been increased.
The explanation is simple. Supra-marginal firms exist because of disequilibrium. In other words, these companies are making economic profits which mean they will have few if any marginal workers. Hence, even an effective increase in the minimum (over a certain range) will not cause them to lay off labour. Indeed, because the minimum will have the greatest impact on marginal firms, major-fast food chains could increase their sales and employment opportunities at the expense of marginal outlets which are either forced to curtail their activities or close down.
To ascertain if this had happened would require a highly detailed survey. Instead, Card and Krueger based their data on telephone interviews and the questions were not even properly defined. For example, while the survey ignored the number of hours worked it asked about the number of employees as well as shifts. This might seem minor except that it led to extraordinary anomalies that immediately alerted economists who studied the results.
In one instance, a Wendy’s outlet went from zero full-time employees and 30 part-time employees to 35 full-time and 30 part-time employees! A Burger King went from 6.5 full-time and 20 part-time employees to 30 full-time and 25 part-time employees; yet another Burger King saw its full-time staff fall from 50 to 15 and its part-time staff drop from 35 to 18. (And this is what Belchamber and Harcourt called “meticulous and exhaustive”).
No wonder Gary Becker (Nobel laureate) was driven to say that, “The Card-Krueger studies are flawed and cannot justify going against the accumulated evidence from the many past and present studies that find sizable negative effects of higher minimums on employment”.
It should be remembered that Card made similar claims for California. When it was pointed out those differences in market conditions could also account for the variations in teenage employment, he agreed that was so but then claimed that “there is no indication of an adverse employment effect [caused by the increased minimum wage] in the low-wage states....” In short, economic activity could have cancelled out the unemployment effects.
Nevertheless, looking at the 1987-1989 period he still suggested “a gain in [California] employment following the rise in the minimum wage,” despite the fact that California’s booming growth impaired his conclusions. The body blow, however, was delivered by Lowell Taylor of Carnegie-Mellon. After examining changes in employment in those counties and retail sectors most affected by the increased minimum, he found that they suffered the worst unemployment effects. Card’s findings stood refuted.
Employers can deal with increases in the minimum in a number of ways. If they may have sufficient advance notice of the increase they can make gradual adjustments in advance. Card’s paper ignored this possibility, retorting that quick-food outlets can sack at short notice. He does not seem to realise that employers do not hire in order to sack. A cost is involved. He also ignored changes in man-hours, fringe benefits, training time, etc. On top of that, he only looked at employment changes over a one-year period, ignoring the role of time.
The period 1963 to1995 provides an excellent example of how employers can deal with effective minimum wages. The US retail trade is probably more affected by the minimum wage than any other sector and this sector saw average weekly hours fall from 37.3 to 28.9 during the 1963 - 1995 period, while hours in mining and building, for example, increased. Therefore the industry adjusted to increases in the minimum by reducing hours worked.
There is no escaping the fact every major study on effective minimum wages (with the rare exception Card and Krueger’s) has always found that they destroy jobs. Furthermore, Professors Robert Meyer of the University of Chicago and David Wise of Harvard, pointed out that abolishing minimum wages would raise the aggregate income of youth in the US. This is something Professor William H. Hutt pointed out decades ago. (The Theory of Idle Resources, LibertyPress, 1977, The Keynesian Episode: A Reassessment, LibertyPress, 1979, The Theory of Collective Bargaining 1930-1975, Institute of Economic Affairs, 1975).
People who fail to see the reasoning for this are obviously confused between wages and payrolls. Despite all that has been said and done, Belchamber and Harcourt expected economics to throw out marginal productivity theory on the basis of a discredited study that absurdly claimed that the demand for marginal labour was increased by raising its costs by $US1800 a year!
Did they really think physicists, for example, would abandon a fundamental scientific law on the findings of such a study? If scientists acted this way, science would be in a state of chaos. No: what they would do is subject the counter-theory to rigorous testing. In short, a process of falsification. They certainly would not listen to politically motivated interests who know nothing of science.
However, let us finish with Mr Card’s rejection of Belchamber and Harcourt’s call for the Living Wage on the basis of Card’s own work. It was Card who stated that advocates “overreacted to the New Jersey study.” In other words, Card and the others only suggest that it is when the minimum wage is ‘low’ that rises will have little effect on employment. Nowhere did Card actually say that raising wages above the market level will not cause unemployment. Therefore, it was not “orthodox” economics that was pushing a myth, but the likes of Belchamber and Harcourt.
The Impact of Safety Net Adjustments on Wages and Jobs, December 2004, co-authored by Belchamber, Donna Hristodoulidis and Andrew Watson, asserted that according to “Neo-Classical economics” increases in award wages since 1996 “should have resulted in a fall in employment especially for award reliant workers”.
Neo-classical economics makes no such claim. What modern economics does say is that if wages rates are raised above their market level unemployment will emerge. What Belchamber and company overlook is that increases in productivity can lower unemployment so long as the increases are not offset by increased labour costs. Moreover, a loose monetary policy also reduces unemployment by raising the price of the product relative to the value of the worker’s marginal product.
(Those in the Austrian school of economics note that this process does not require a change in the CPI and therefore a cut in real wages. This happens because the expanded money supply releases withheld capacity [William Harold Hutt, The Keynesian Episode: A Reassessment, LibertyPress, 1979]. Keynesianism, however, teaches that a rise in the price level is necessary to cut real wages [John Maynard Keynes, The General Theory of Employment, Interest and Money, MacMillan, St Martin’s Press for the Royal Economic Society, 1973, Book I, p. 8).
Their argument against marginal productivity theory was outrageous economic quackery. Unfortunately, Belchamber is still with the ACTU and still trying to play havoc with Australia’s labour markets. However, Tim Harcourt was appointed in 1999 by the Liberal Government to be Chief Economist of the Australian Trade Commission. No wonder somebody once quipped that the Liberal Party should be called the Stupid Party.
*When the OECD issued a report that was quite critical of our heavily regulated labour market, Belchamber contacted Crean’s department and suggested that it would be better if the government used a review carried out by the OECD’s Labor Market Directorate instead of its Economic Department. Ironically, it seems the Directorate was just as critical of Australia’s labour laws as the Economic Department.
Gerard Jackson is Brookes’ economics editor
BrookesNews.Com
Monday 1 August 2005