As promised, this week’s post looks at Olivier Blanchard’s treatment of efficiency-wage theory (EWT). It is a companion piece to last week’s analysis of his stylized bargaining theory (BT). Blanchard believes that a useful NK model of wage determination can be constructed on the two approaches. The GEM Project also relies on early EWT modeling but with very different results.
My objective here is to explain why Blanchard and the Project, despite drawing from the same literatures, produce such different outcomes. The latter is clearly more ambitious, deriving from axiomatic model primitives meaningful wage rigidity that rationally suppresses wage recontracting over the business cycle. MWR microfounds both the existence of involuntary job loss and powerful causality from nominal-demand disturbances to same-direction, recognizably-sized changes in employment, output, profits, wage income, investment, and consumption. Blanchard’s labor pricing induces none of those central macro outcomes. He instead gets stuck in the ubiquitous search-match-bargaining quagmire, trying and failing to squeeze some plausible cyclical joblessness out of rational market search that focuses on voluntary unemployment and routinized-job recruitment costs that are broadly understood by practitioners to be very small.
A particular model-building choice turns out to be crucial to wage-theory success or failure. Blanchard and almost everybody else choose to tackle modern labor pricing wholly within a market-centric analytic framework while the GEM Project, in large-establishment circumstances of costly, asymmetric information and routinized jobs, chooses to generalize rational employer-employee exchange from the marketplace to the workplace. The hard fact is that, beginning the late 19th century, simultaneously coherent, stabilization-relevant modeling requires rejection of the consensus restriction of rational, price-mediated exchange to the marketplace. That constraint is among the most egregiously non-descriptive assumptions in mainstream theory. Blanchard’s wage modeling is typically market-centric and therefore fails.
Efficiency-wage theory. From Blanchard’s textbook Macroeconomics (2009): “Regardless of workers’ bargaining power, firms may want to pay more than the reservation wage. They may want their workers to be productive, and a higher wage can help them achieve that goal. If, for example, it takes a while for workers to learn how to do a job correctly, firms will want their workers to stay for some time. But if workers are paid only their reservation wage, they will be indifferent between staying or leaving. In this case, many of them will quit, and the turnover rate will be high. Paying a wage above the reservation wage makes it financially attractive for workers to stay. It decreases turnover and increases productivity. Behind this example lies a more general position: Most firms want their workers to feel good about their jobs. Feeling good promotes good work, which leads to higher productivity. Paying a high wage is one instrument a firm can use to achieve these goals.” (p.143)
Blanchard’s market-centric analysis defaults, as it must, to preventing voluntary quits. There is no recovery from that misstep. Most problematic, rational strategies to discourage quits do not suppress wage recontracting. Nominal wages are downward flexible over the business cycle; forced layoffs, the signature feature of recessions, cannot exist. Despite the elimination of involuntary job loss, which henceforth is out of mind, Blanchard still assigns a starring role in wage determination to unemployment, which is now sort of deus ex machina.
In Blanchard’s market-centric approach, efficiency wages must be a black box. He intuits that the concept justifies paying wages greater than worker opportunity costs, but has no idea how to deal with obvious questions that result. How much labor rent is optimal? Does trend rent vary over time and why? Is the nominal efficiency wage and downward rigid over the business cycle? If so, how does it suppress rational labor-price recontracting? If not, how does it support involuntary job loss, especially in response to adverse nominal demand disturbances? If wage rents are rational, chronic and substantial, aren’t there implications for neoclassical micro modeling that, by broad application of the Lucas critique, need to be worked through before macroeconomics can be considered microfounded. Persistent, time-varying wage rents, for example, significantly alter how the rational SMB model, and its attendant bargaining, works.
It is clear that Blanchard has not carefully read the literature that he attempts to summarize. A central message of original efficiency-wage theory is the sharply diminished role of the labor market and market unemployment in rational large-establishment wage determination. If you do not know that, you cannot understand the efficiency-wage theory that attempted to incorporate actual, well-documented corporate labor-pricing practices into formal economic theory. Blanchard’s inadequately informed guesswork about how profit-seeking management seeks to foster unit-cost-minimizing worker behavior is badly at odds with the available evidence on efficiency wages, greatly damaging his macroeconomics. (Chapters 1, 10)
Opportunity. Blanchard has a casual familiarity with ideas – morale and commitment necessary for employees to respect employer’s objectives – that have evolved over time to motivate large-establishment human-resource management and its policies governing entry-wage levels and intra-firm structure. But he goes off the rails by insisting the workplace process, constructed to deal with workplace-specific constraints, be interpreted wholly through the familiar lens of marketplace exchange.
Most discouraging about Blanchard’s persistence in making the same mistake is that he is aware of the on-going research to generalize rational exchange from the marketplace to the large-establishment workplace. He must know that “feeling good” about one’s job and its relation to employee willingness to adopt management’s goals are nonmarket phenomena. Hopefully the GEM Project can help him get over the hump, transforming timid speculation into coherent, stabilization-relevant modeling. His mainstream gatekeeping status is such that macroeconomics would greatly benefit, especially when tasked to make sense out of crises such as the extreme instability of 2008-09.
Blog Type: Wonkish Chicago, Illinois