Big question. Last week’s post sampled the broad and commanding explanatory power of microfounded meaningful wage rigidity. It is an extraordinary story. This week poses, and builds the analytic framework needed to understand, a follow-up question. Has MWR lost some of that formidable power, originally rooted in the Second Industrial Revolution, in the subsequent digital and artificial-intelligence revolutions? The guidance derived in the GEM Project is that MWR exerts a central influence, along the lines explicated in the website’s e-book, in any workplace that is (i) restricted by costly, asymmetric employer-employee information and (ii) populated by routinized jobs, as defined below.
Analytic framework. This post is deservedly labeled wonkish. Its objective is to rigorously identify another piece of the puzzle of highly specialized economies that mainstream market-centric general-equilibrium analysis must ignore. Routinized employment, named Class-I in the GEM Project, has accounted for a substantial share of the total U.S. workforce for more than a century.
Many Class-I jobs are rationally associated with the receipt of wage rents and have, as a result, long been a staple of political interest. Such employment is what Donald Trump had in mind when, in the 2016 election, he recklessly pledged to return out-migrated large-firm assembly work to the US. Macro theories must account for such labor heterogeneity. But, once again, New Keynesian market-centric modeling is not up to the task. And, once again, that failure is consequential.
The Project’s generalized-exchange theory accommodates the heterogeneous nature of jobs, further introducing useful structure into its description of labor pricing and use. The idea is hardly new. One of Adam Smith’s best-known injunctions is that wages cannot be independent of work-task characteristics. Workplace-Exchange Relations for corporate research scientists differ markedly from exchange relations of production workers in a large factory. Generalized-exchange modeling uses a bimodal separation to introduce useful job heterogeneity.
Complications rooted in employee consciousness become more significant as work tasks become increasingly routinized. Work is divorced from end products, limiting the range of tasks performed and requiring navigation of hierarchies of authority. Employers and employees exchange workplace inputs (Ỉj) for outcomes (Ỏj), respectively defined as:
Ij= IΈj+ Ifj and Oj= OWj+ ONj such that IΈj=Ej and OWj=Wj,
where IΈ stands for cooperative labor input on the job, If is job-related fixed employee characteristics, OW denotes the pecuniary outcome from the job, ON represents the nonpecuniary outcome, E is the labor input that has a 1:1 relation with production, and W denotes the wage paid.
Class-I employment imposes two significant restrictions on rational workplace exchange:
- ON/O is relatively small, i.e., nonpecuniary outcomes are an inherently minor component of total employee outcome.
- If/I is relatively large. Human-resource decision-making, whenever confronting substantial workplace information costs and asymmetries, is more informed by worker fixed inputs, particularly seniority, than by inherently imperfect measurements of on-the-job behavior.
The formal definition of Class-I jobs is ON/O < If/I. The remainder of the employed workforce holds Class-II jobs, defined as ON/O ≥ If/I. This second job class significantly, but not exclusively, includes employment that yields substantial personal satisfaction from the performance of the work tasks.
The GEM bimodal employment separation is recognizable in modern economies. Many Class-I jobs are shaped by the specialization imposed by large-scale, high-volume (goods and services) production technologies, occurring in workplaces characterized by costly, asymmetric information. Such routinized employment is “unpleasant mainly because it fails to stimulate the worker yet prevents him from seeking stimulation elsewhere.” (Scitovsky (1977, p.92)) Adam Smith drew early attention to the progressive simplification of work tasks, concluding that a significantly restricted range of on-the-job activity is a source of substantial boredom.
Large bureaucratic establishments additionally make no secret of their general practice of rewarding employees in routinized jobs more on the basis of their fixed characteristics, particularly seniority, than on inherently flawed measures of their performance on the job. Extensive evidence demonstrates the ubiquity of awarding wage increases, eligibility for desirable job transfers, and ranking in a layoff queue on the basis of employee seniority – practices that employees typically interpret as fair treatment by management.
Meanwhile, Class-II jobs tend to be more of a hodgepodge, reflecting employment characteristics that variously include substantial satisfaction from the work itself (pushing up ON/O) and cost-effective OJB supervision (pushing down If/I). Using longitudinal surveys, Kohn and Schooler identified “occupational self-direction” as especially significant in shaping an employee’s qualitative response to his or her job. Work tends to be intrinsically rewarding when it requires individual initiative and independent judgment, implying more complex workplace tasks, lack of routine, and remote supervision. They found, not surprisingly, that jobs with such characteristics tended to be higher-status positions in their organizations. Richard Florida has named workers who enjoy relatively high non-wage benefits from their jobs “the creative class”. Including scientists, engineers, artists, musicians, designers, and knowledge-based professionals, he estimates this creative class accounted for roughly 30 percent of total employment in the United States in 2000, up from 20 percent in 1980.
This is where a modern variant of Florida’s classification fits: dead-end creative-class jobs. The following is from a recent collection of young-person grievances about competitive-market economies: “Educating a generation and saddling them with debt and then not giving them jobs where they have the wage that they presume they should receive based on the amount of time they spent on education, that’s a pretty good way to turn them into radicals.” The interesting twist here is the apparent belief of some that they should not be responsible for the relative (pecuniary versus nonpecuniary) returns implicit in their choice of what type of general human capital to acquire and how much to invest in how well it is acquired. I don’t recall commonplace anger about those choices and the lack of public subsidies to redress them in earlier generations.
It is also noteworthy here that economists at research-oriented universities hold class-II jobs. Their personal experiences, focused on work satisfaction, idiosyncratic hiring processes, evaluations based heavily on publications and other easily observable measures of performance, and tenure, poorly correspond to the circumstances and behavior of workers in class-I jobs. Economists must stop, in their own research, generalizing their personal experience to the labor-force as a whole. Doing so has produced badly misleading models, especially the ever more elaborate use of frictional (voluntary) unemployment in attempts to explain forced layoffs that result from adverse disturbances in nominal demand.
Conclusion. There is nothing inherent about the digital and AI revolutions that require new jobs to be mostly Class II. To the extent they are both Class I and located in ubiquitous information-challenged workplaces, microfounded MWR will retain its extraordinary explanatory power.
Blog Type: Wonkish San Miguel de Allende, Mexico