Getting Efficiency-Wage Theory Wrong

Print/Save PDF

Last week’s post elaborated on consequences produced by the mainstream macro theory’s absence of rational meaningful wage rigidity (MWR). Overall, the greatest damage has come from model-builders being deprived of crucial guidance in their attempts to construct stabilization-relevant macroeconomics. Surely we have by now learned that an accurate roadmap is a must. Useful macro analysis became much more complex after the industrial revolutions, diverting theorists into a multiple wrong turns.

This post looks at what might have been had the mainstream roadmap been more accurate. The analysis draws on the history of efficiency-wage theory, a promising approach to microfounding MWR that made great strides in the late 1970s and early 1980s before going badly astray.

Promising theory. Original efficiency wages (OEW), i.e. the morale-centric formulation that Michael Piore, Robert Solow and I pioneered, indicate how theorists could have gone about rationally disabling Keynes’s Second Classical Postulate. In the circumstances of costly, asymmetric workplace information, OEW moved labor pricing from the market to inside the profit-seeking firm. Building on that transfer, OEW theory informs a powerful framework for stabilization-relevant macro theory that is consistent with both optimization and equilibrium. After the original articles, Solow’s most complete EW statement was his Royer Lectures, published as The Labor Market as a Social Institution (1990); mine was The Price of Industrial Labor (1984). Piore moved on to other research interests. All three  notably failed to microfound MWR.

The development and impact of efficiency-wage theory were damaged by that failure. Today, OEW is badly misunderstood by economists. In their otherwise exemplary history of economic thought, Scepanti and Zamagni (2005, p.369), hereafter S&Z, provide an illustratively misleading description of useful efficiency wages: “The theory of efficiency wages is based on three principal ideas. The first is that the intensity of work effort of each employee, and therefore the marginal productivity of labour, increases with an increase in wages. The second is that the workers’ effort is also influenced by the level of unemployment, in that the fear of being dismissed for inefficiency increases with an increase in the probability of not immediately finding another job with the same pay – a probability that rises with the level of unemployment. The third hypothesis is that there is a type of asymmetric information, as firms are not able directly to ascertain the intensity of effort of hired worker or the ability of those to be hired. In these conditions, it is in the interest of firms to pay high wages to encourage workers’ effort.”

The first two “principal ideas” incorrectly describe OEW theory. The S&Z depiction is instead relevant to the less useful Shapiro-Stiglitz (1984) shirking variant of efficiency wages. Despite being best known of all the EWT branches, S&S badly represent its power. S&S fail to microfound MWR or provide a useful intra-firm framework for doing so.  It satisfies none of the grand objectives of efficiency-wage theory. Instead, S&S generates flexible labor pricing, confines job loss to dismissal for cause (a trivial category), doesn’t come close to explaining business cycles, and provides only dangerously wrong advice to stabilization policymakers. Keynes, whose overriding goal was to usefully analyze involuntary job loss, is simply pushed aside. For elaboration on the dominating, albeit curious, S&S efficiency-wage contribution, see the website’s e-book, chapter 9.

Available evidence strongly indicates that highly specialized bureaucratic employers believe that cooperative on-the-job behavior is damaged by cuts from existing labor pricing that is rationally rooted in employee reference standards that are established over time. The GEM Project easily demonstrates that rational employees must experience significant job downsizing, not temporary layoffs, to justify cooperative acceptance of wage reductions. Labor productivity that is simply increasing in labor pricing is a very different, nonintuitive process, badly complicating model stability and, outside of the limiting case of nutrition and health, supported by neither practitioner testimony nor available evidence. It cannot rise above being a silly assumption. In the GEM model class, the efficiency wage is rationally and powerfully independent of contemporaneous unemployment as well as market opportunity costs.

As already noted, S&Z mistook as general the subset of follow-up EWT models pioneered by Shapiro and Stiglitz. Reflecting the dominating mainstream preference for analyzing marketplace exchange, S&S and other late entrants to EW research attempt to model workplace exchange within the consensus general-market-equilibrium framework, identifying dismissal for cause as the forced job loss of interest and using fear generated by the high incidence of market unemployment to motivate cooperative labor input. Economists who have actually studied labor pricing probably recognize firing and fear as a familiar nineteenth-century practice that has had, for generations, little place in best-practices labor-management relations in bureaucratic firms. Reiterating for emphasis, S&S nominal labor pricing is downward flexible over the business cycle and cannot inform the channel through which nominal disturbances induce (temporary) layoffs and (permanent) job downsizing.  Such models cannot justify discretionary demand management. The substitution of a tiny-incidence category of joblessness (dismissal for cause) for the vastly more important involuntary layoffs and downsizing is indicative of the practical irrelevance of the S&S efficiency-wage model. That the two models, one (OEW) useful and the other (S&S) useless, share the same name has damaged progress toward a stabilization-relevant macro theory.

C-S Macroeconomics. The past two weeks have also focused on the second edition of the Carlin-Soskice textbook Macroeconomics (2015).  The C-S description of efficiency wage theory (EWT) begins well: “Efficiency wage setting is quite different [from other labor-price theories]; here it is the firm that sets wages. At first sight, it is counter-intuitive that an employer should voluntarily set a wage above the minimum at which it can hire labour in the market. The argument is that by setting a wage above the competitive one, the employer is able to retain a well-qualified and cooperative workforce. The term ‘efficiency’ wages arises from the notion that the firm sets a wage that allows it to efficiently solve its motivation, recruitment, and/or retention problems. These problems arise because it is generally not possible to specify fully what a worker does, i.e. the employer cannot observe accurately the worker’s effort. The employer must therefore use the wage to motivate the worker to perform well.” Then it falls apart: “As the labor market tightens, it becomes more difficult for the firm to solve these problems because workers can easily leave the firm and find work elsewhere. The result is that the employer sets a wage above the competitive wage and this ‘efficiency’ wage rises as unemployment falls. This is therefore a second way of understanding the wage setting curve.”

If C-S had taken seriously the core efficiency-wage idea of a workplace venue that rationally pays nonmarket wages, they could have fleshed out the fundamental importance of cooperation, motivation, and incomplete contracts instead of pushing them aside in their subsequent analysis that retreats to comfortable market analysis of recruitment and retention. Or they could have read the EW literature more diligently and discovered that the framework for optimizing employee behavior in the circumstances of information-challenged workplaces had already been deeply analyzed four decades ago and that work has more recently led to fully microfounded MWR.

Armed with rationally suppressed wage recontracting, C-S could have deduced the EK centrality in aggregate-demand modeling that is fully consistent with optimization and equilibrium a well as with the full range of important evidence. The microfounded EK model could be confidently presented as the clear winner over the RBC and NK (DSGME) models. C-S could then do their job, explaining why students should choose that model over its market-centric competitors in the effort to “to understand macroeconomic behaviour and policy issues in the real world.”

Blog Type: New Keynesians Chicago, Illinois



Write a Comment

Your email address will not be published.