Labor Economics

Though Paul spent much of his time discussing ideas with fellow students and faculty members, only one joint paper was published while he was at Harvard. His work with Marion remained unpublished, and his work with Stolper on international trade was not published until 1941.¹⁶ His co-author on the paper published in 1940 was Russ Nixon, a couple of years older than Paul and who had been an instructor at Harvard since 1936.¹⁷ John F.

d. The technique they had used was Volterra integral equations. Lotka, who typically responded by explaining to Paul that something had been covered in his earlier work, said that while Paul's method had expositional advantages, he did not consider it more general.

e. See chapter 13 this volume.

Tsuru.) Not surprisingly, Nixon's thesis, submitted in 1940, had been on employability and the labor market.

Samuelson explained that he became involved in writing the paper “Estimates of Unemployment in the United States” with Nixon because he learned that “a third-party go-getter was elbowing Nixon out of a joint ven­ture,” and he sympathized with the underdog.²¹ Friendship clearly provided an important motivation for Samuelson: they kept in touch when he was vis­iting Washington during the war, even though they were no longer working together.

The heart of the paper was their discussion of five measures of unemploy­ment from 1929 to 1940 during the Great Depression. Taking into account statistical issues, such as the way different censuses treated unpaid family labor, they concluded that there was no uniquely correct measure of unem­ployment. However, though the paper was primarily an exercise in statis­tics, its main interest in relation to Samuelson's intellectual development lies in the theoretical arguments used to make sense of different ways in which unemployment could be measured. They started by defining full employ­ment in terms of a short-run labor supply curve: “Employment is full when individuals work as much as they would be willing to work at a given real wage (or structure of real wages)."^22,f Such a definition, though it might seem complicated, was essential in a world in which workers had different prefer­ences and if one family member became unemployed, other family members tried to find employment. Employment might be less than full because the labor market was imperfect, perhaps because those who were unemployed held back from bidding down the wages received by those who were work­ing, or because employers refused to hire workers at less than what they believed to be a “fair" wage.²³

This was the conventional analysis. It rested on the tacit assumption that a cut in wages would reduce or eliminate unemployment, as implied by the theory of supply and demand in a competitive market, and it was at this point that they tackled the ideas of Keynes.

f. The real wage is the money wage rate adjusted for changes in prices; it measures the quantity of goods and services that the wage rate will buy.

Leontief, they were skeptical of Keynes's argument that workers might suffer from an “optical illusion,” whereby they resisted cuts in money wages but did not resist cuts in real wages brought about by prices rising faster than wages. They argued that, according to Keynes's own theory, employment and the real wage would change only if the interest rate, investment, or the pro­pensity to consume were changed; it was thus possible to have a downward spiral of wages and prices without any change in employment. They were equally dismissive of Keynes's attempt to provide an alternative definition of full employment as the level of employment where rising effective demand caused wages to rise; wages often rise, they argued, when unemployment was still high. Noting that “an increase in unemployment is usually associated with declines in production and the real national income,” they accepted that “fluctuations in these magnitudes are the result of fluctuations in the level of ‘effective demand.' ”24 However, they were clear that this idea was not tied to any particular theory of the cycle; they did not see it as a specifically Keynesian idea.

A concept to which they attached particular importance was “disguised unemployment,” on which they cited Joan Robinson.²⁵ When people lost their jobs, they might engage in subsistence farming or house-to-house selling, or they might accept lower-paid work than they were qualified to do. This was a significant problem in the United States, for agricultural statistics showed unpaid family workers and subsistence farmers as being employed, and their reclassification would raise the unemployment count by millions. The importance of agriculture was no doubt the reason why they had discussed the topic with Harvard's main agricultural economist, John Black, whom they thanked for many suggestions and several of whose papers they cited.

Their discussion of this point raised the question of where the boundary was between problems of microeconomic resource allocation and problems of the business cycle. “Some readers,” they claimed, “may believe that prob­lems of non-optimal allocation of resources belong in the field of welfare economics and value theory, and that consideration of such problems in con­nection with a study of unemployment is tantamount to permitting welfare economics to swallow up business cycle theory completely.”²⁶ Against this, they countered,

We would argue that the concept of disguised unemployment should not include all deviations from optimal allocation, but only those which are due to cyclical variations in the level of effective demand. The concept of disguised unemployment is peculiarly useful because it illustrates that even in a perfectly organized labor market... large fluctuations in the level of the national income and production would still be possible.²⁷

Nixon and Samuelson were, like Keynes, making a clear separation between what would later be called microeconomics and macroeconomics. They also followed Keynes in arguing that the failure of Say's Law (the idea that sup­ply creates its own demand and so there cannot therefore be a shortage of demand) is not simply the result of rigidities in the price system. However, their use of Keynesian ideas was selective, for they had clearly accepted some of the criticisms that their teachers, notably Leontief, had made of Keynes's General Theory.

It is also worth noting that this article is the first indication of how the war, which had broken out in Europe, though the United States was not yet fully involved, was beginning to affect thinking. British authors (including Keynes) were turning to a different concept of full employment—maximum labor resources or “national potential”—which, if used, would result in much higher estimates of unemployment. “It is paradoxical,” they wrote, “that the true economic realities of scarcity and choice become apparent, even to the uninitiated, only in time of war, when the problem of effec­tive demand becomes nonexistent and the monetary veil is pierced.”²⁸ War had made people aware that, although the immediate problem was lack of demand, the basic economic problem remained that of scarce resources. This led Samuelson and Nixon to think of full employment as a ceiling that could never be reached.²⁹ They were thinking seriously about the supply side of the economy, and not just about aggregate demand.