A favorite pastime of mine is browsing through macro textbooks, identifying familiar topics that are badly explained by mainstream friction-enhanced general-market-equilibrium (FGME) theory. In particular, I am looking for phenomena that cannot be adequately understood absent the generalization of rational exchange from the marketplace to workplaces constrained by costly, asymmetric information and routinized employment. (What follows uses the loose Project convention of naming the workplace sector the large-establishment venue (LEV) and the marketplace sector the small-establishment venue (SEV).) A problem with the hobby is that it is too easy. Examples of failed applications of market-centric theory, accumulating since the Second Industrial Revolution, are ubiquitous – the low-hanging fruit of modern macro research.
Labor-supply elasticity is this week’s subject. LSE is defined as the ratio of the percent change in labor hours to the percent change in real wages. While the concept sounds boring, it has had a lively history. The ratio, assuming causation from wages to hours, was infamously used to argue that cyclical unemployment can be made consistent with FGME, becoming a flashpoint of debate during the acrimonious macro wars of the late 20th century. The LSE argument triggered many scholarly efforts to estimate its size. That attention joined the empirical literature already being produced by the longstanding public-finance interest in the efficiency of taxing labor income, largely focused on the degree to which higher levies suppress hours worked. While various micro studies have for the most part produced consistent empirical results, there has been much less success grounding those findings in rational behavior. It is a tenet of the GEM Project that rational-behavior modeling is necessary to adequately understand empirical evidence.
Here’s what FGME textbooks overlook. Useful analysis of LSE, whatever its focus, requires explicit recognition of the role of labor rent. The phenomenon became especially critical in modern thinking once the GEM Project demonstrated that pure wage rent (PWR) is a chronic outcome of optimizing employer-employee interaction in information-challenged workplaces. The descriptor “pure” distinguishes the concept from more familiar, but less fundamental, versions of wage rent typically associated with labor unions, regulated firms, or market frictions studied in the labor search/match literature. PWR has long been the LSE literature’s elephant in the room.
PWR Theory in a Nutshell
Generalized-exchange modeling separates firms into two venues, capturing the size-related heterogeneity arising from compromised direct employee supervision in highly specialized workplaces. In the two-venue theory, labor is point-of-hire homogeneous, and Harris-Todaro transfer rationally governs inter-venue worker flows. Large-firm labor pricing is located in the workplace, where firms construct exchange mechanisms and pay the continuous-equilibrium efficiency wage (Wn) that equals rational employees’ reference wage (Wń). Optimizing workplace exchange microfounds keystone meaningful wage rigidity (MWR) and chronic PWR, as well as rational rational on-the-job behavior. (See Chapter 2 of the website’s e-book.) Meanwhile, small firms, capable of effective workplace oversight, optimize by paying the flexible market wage (Wm).
Macrodynamics are crucially informed by rational MWR and PWR, which combine to suppress wage recontracting, uniquely motivate involuntary job loss in response to adverse nominal demand disturbances, and push workers off their market labor-supply schedule. Keynes’s Second Classical Postulate and Wicksell-Wicksteed income distribution are scrapped in the modeling of information-challenged workplaces. Income becomes the primary determinant of consumption, and expectations of pure profit principally drive business investment. Interest rates play secondary roles in each.
Stationary spending disturbances are associated with temporary layoffs, while nonstationary demand shocks generate permanent job downsizing and eventually recalibrate worker reference standards, producing wage givebacks. Unemployment always follows a continuous-equilibrium macrodynamic path. Labor is employed in rationed rent-paying jobs or readily available market-wage jobs, involuntarily or voluntarily unemployed, or voluntarily out of the labor force. Job quits are procyclical and play no significant role in macro stabilization. Job-matching efficiency also does not much matter, insignificantly influencing employment fluctuations.
Rational exchange in information-challenged workplaces produces chronic wage rent and job rationing that constrain optimization in the marketplace, reconciling continuous decision-rule equilibrium and supply-demand disequilibrium. In the simplest two-venue version, all rational exchange, except between second-venue employers and employees, occurs in the marketplace and is largely governed by familiar textbook analysis. The crucial message, once the model pieces are assembled, is definitive. Macroeconomics that is both evidence-relevant and consistent with the neoclassical tenets of optimization and equilibrium is not feasible absent microfounded PWR and, consequently, the generalization of exchange.
Back to Labor-Supply Elasticity
As noted, there has been a great deal of empirical work and interpretation on labor-supply elasticity, almost all of it within the mainstream market-centric, general-equilibrium framework. The micro estimation strategy is to identify an evidence source, frequently panel data, and regress hours change on wage change. Major surveys of that LSE literature find generally consistent results. Labor supply elasticities are very small. That conclusion has not changed from the earliest econometric results.
Three examples of that pioneering work (MaCurdy (1981), Browning, Deaton, and Irish (1985), and Altonji (1986)) are illustrative. The estimates of labor-supply elasticities are respectively 0.15, 0.09, 0.31. The MaCurdy estimate implies a 10 percent wage increase raises labor supply by 0.8 percent. GEM analysis provides a robust alternative rational-behavior explanation for those results. It focuses on PWR and its effect on effective labor supply, especially the huge employment in the market-wage-paying venue that would voluntarily transfer to a rent-paying job if one became available. (That is, of course, the rational process captured by the Harris-Todaro labor-transfer model.) Labor supply is, as a result, functionally elastic to rent-paying firms at their going wage. Data on LEV labor hours change largely reflect shifts in sector labor demand; data on wage change, the movement of the efficiency wage. Contrary to the universal assumption of FGME theorists, wages and hours are not causally related.
The LSE literature identifies two broad classes of implications of the GEM analysis. In the first, understanding the efficiency cost of taxing labor income is not possible absent PWR modeling. In their review of the relevant evidence, Saez, Slemrof, and Giertz (2012) typically get it wrong: “… optimal progressivity of the tax-transfer system, as well as the optimal size of the public sector, depend on the … elasticity of labor supply.” The second is the macro issue concerning the degree to which intertemporal (work versus leisure) substitution can adequately explain unemployment macrodynamics. That Real-Business-Cycle story has always been bogus and PWR modeling just adds another nail to its coffin. Deep down, every thinking economist must know the substitution effect cannot explain cyclical unemployment. Substitution mechanics are limited to voluntary job loss, i.e., quits, while job loss in recession is heavily concentrated in involuntary layoffs.
Blog Type: Wonkish Saint Joseph, Michigan