The GEM Project’s generalization of rational exchange from the marketplace to workplaces restricted by costly, asymmetric information and routinized employment is an incremental exercise. It has deep roots in the innovative work conducted in the middle 20th century by the first-generation Internal-Labor-Market (ILM) theorists, a number of whom were my teachers and colleagues at Princeton and MIT. They modeled the consequences for labor pricing and use of managerial capitalism, the subject of last week’s post.
The original ILM literature simultaneously developed with the much larger Keynesian macro program and provides a body of research more than sufficient to inform the rational suppression of wage recontracting. A great deal of toil and trouble could have been avoided if Cambridge neighbors Franco Modigliani and John Dunlop had got together for a working lunch.
Early ILM school. A loose gathering of scholars, including Clark Kerr, John Dunlop, Richard Lester, Lloyd Reynolds, Frederick Harbison, Charles Myers, and Arthur Ross, dominated American labor economics during the middle of the 20th century. Kerr named the school the “neoclassical revisionists”, identifying its ambition to use their neoclassical economics training to study “real-world” labor behavior inside large firms. Early ILM theorists modeled objectives informed by insightful specifications of employee preferences, constraints rooted in modern technology, and intentional intra-firm mechanisms exchange. They were close to providing an early solution to what became a debilitating class of labor-related deficiencies in micro-coherent macroeconomics. They uncovered the facts but ultimately failed to construct a consistent theory of workplace behavior and were, as a result, pushed increasingly outside the economic mainstream. Kerr (1988, p.21) recognized the school’s difficulty: “Perhaps the most serious problem … was that the revisionists dealt bit by bit with pieces of the puzzle and never assembled them into an integrated statement, let alone into a model or a consistent theory; and it takes a new theory to replace or change an orthodox theory.”
The GEM Project responds to Kerr’s lament, building on the work of ILM theorists to construct a consistent model that generalizes rational labor-related exchange from the marketplace to the bureaucratic workplace. Exchange generalization uniquely enables the reconciliation of micro-coherence and stabilization-relevance, permitting rigorous modeling of real-world instability.
An overview. Serious academic research on large-establishment employee behavior can be roughly dated from the celebrated 1924-32 Hawthorne experiments. Conducted at the Hawthorne plant of the Western Electric Company, the study was originally designed to be technical in nature: to determine the relation between conditions of work and the incidence of fatigue and monotony among workers. As the study progressed, however, the technical factors – rest pauses, lighting, work scheduling, and so on – fared poorly as explanations for productivity variation. Much more important were worker attitudes and the role of informal work groups.
The impact of the Hawthorne experiments, others like it, and ongoing management learning-by-doing was substantial. Practitioner understanding of labor productivity shifted from a simple to a more complex view of worker conduct, ultimately focusing on how to induce employees’ acceptance of the firm’s objectives. Part of the growing research, using a neo-classical economic perspective to expand and, where needed, modify early interpretations of Hawthorne results, was forthcoming from the early ILM theorists. Writing largely from the 1940s into the 1980s, they drew descriptions of employee on-the-job behavior (OJB) from experience with war labor boards, government adventures into wage controls, and as arbitrators and researchers with broad access to the internal workings of large firms. They investigated employee preferences and produced a detailed picture of what occurs in the large-establishment workplace.
Their central organizing idea is Kerr’s (1954) “balkanization” that identifies distinct labor-management systems contingent on firm size. As noted, work here has continued. An especially important contribution by the next generation of ILME scholars, which descriptively integrated much of the early hands-on ILME workplace analysis, is Doeringer and Piore (1971). They mapped the rules and tradeoffs used to organize large-establishment workplaces, emphasizing the roles of closed job ladders and workplace ports of entry.
Reference standards. Many interrelated insights building on market balkanization and internal labor markets were forthcoming from early ILM on-site observations and analysis. Significant “puzzle bits” included bounded mobility, the purpose and consequences of long-tenured employment, the repeal of the law of the single wage (i.e., large establishments paying more than small firms for market-equivalent labor), the practice of wage imitation and pattern setting, spontaneous and intentional industry labor-cost cartelization, the modeling of labor-union power and objectives, and the nature and organization of collective worker action. Of all the contributions, economic theory most benefits from the finding that positional concerns, especially interpersonal and intertemporal wage comparisons, play a crucial role in employee satisfaction.
Fair treatment by management, an outcome of rational employee-employer interaction substantially embodied in the maintenance of established reference standards, is the glue that holds the early ILM workplace modeling together. John Dunlop (1957) notably emphasized the power and ubiquity of intra-firm reference standards in labor compensation, naming them “wage clusters”. Arthur Ross (1948), focusing on external and intertemporal standards in the context of collective behavior, coined the more colorful “orbits of coercive comparison”. Richard Lester (1951, pp. 46-48) provided a compact ILM statement of the economic consequence of reference standards. He distinguished between time spent on the job and effort and “diligence” produced by employees. He argued, from experience, that by cutting wages employers were unilaterally altering the terms of the employment arrangement, making their firms vulnerable to retaliation in form of reduced effort, inattention to quality, and so on. The early ILM guidebook for the proper modeling of the keystone nonconvex WER in Figure 1.1 (Chapter 1) was available well before the advent of the efficiency-wage or implicit-contract model classes.
The powerful relevance to modern micro-coherent, stabilization-relevant macroeconomics would not have surprised the neoclassical revisionists. Clark Kerr always asserted that he and his ILM colleagues intended to construct a Keynesian labor economics. It is time for macroeconomists to realize that they were successful.
Blog Type: Policy/Topical Chicago, Illinois