Here are some unstructured points about underemployment equilibria.
Underemployment equilibria are real: we can agree (or assume, for the sake of discussion) that underemployment is real, that there are real welfare loses associated with depressed output, i.e. low capital utilization and unemployment. An economy that doesn’t fully employ its resources is passing on a “free lunch”, one that could be enjoyed in an alternative institutional arrangement, one that would coordinate a better use of the capital stock and available labor.
On RBC: Since a) most variation in hours worked comes from people working or not working (extensive margin) rather than the average number of hours per worker (intensive margin); and b) a vast majority of the workers work full-time standardized schedules, intertemporal substitution and a taste for leisure in general seem irrelevant for the question at hand, at least as a first order approximation. Also, because of the lack of strong endogenous propagation mechanisms, a RBC model will produce downturns only as persistent as the underlying shocks (mainly productivity). This is why baseline RBC models overpredict the speed and timing of recoveries.
Sticky Prices: Sticky prices (e.g. of the Calvo variety) are not the endogenous propagation mechanism we’re looking for because a) they don’t match well the prices microdata (e.g. scanners datasets); and b) they don’t match the spectral density of the aggregate price level either.
Keynes: In troubled times, Keynes (and for that matter Malthus) are hailed as visionaries. Yet, if my theory predicts something that does happen (underemployment), it doesn’t follow that my theory is correct. It seems to me, and this is purely a subjective judgment at this point, that Keynesian economics (of all types) doesn’t have a satisfactory, formal model exhibiting permanent or transitory underemployment, one that’s consistent with rational choice along what we would consider to be the relevant margins and one not featuring any expectations free lunches.
Full Employment: As long as we don’t have a good theory of underemployment we cannot say whether it is transitory or permanent. We see a 2% growth rate path of real GDP, well approximated by a log-linear trend and count that as proof of balanced growth path (steady state in per effective units) but Jones shows that the US real GDP time series is best understood as constant growth transitional dynamics rather than a balanced growth path because of significant trends in educational attainment and R&D expenditure.
Multisectoral models: Since in the data all sectors boom and depress together (as a first order approximation), stories of inter-sectoral reallocation (e.g. fluctuating degree of tastes-technology match) look irrelevant. Also, it is very hard to build models that match the data because, even in a two sector model, with different consumption and investment goods, the sectors move in different directions following shocks to productivity, rather than in the same direction.
Perfect competition: By assuming market clearing each period, in each market, we rule out the most obvious (trivial?) cause of underemployment, i.e. shortages and surpluses caused by “wrong” prices. On the other hand, there’s a sense in which perfect competition assumed far too much coordination: costless, immediate, intertemporal. It’s not that perfect competition always coordinates Pareto optimal outcomes (it doesn’t), but that perfect competition leads to a coherence of plans such that everyone’s goals are realized, subject to “market signals”.
Worker-Firm Surplus: The patters of matching and separation we observe in the data are potentially relevant: a) firms prefer layoffs to wage cuts; b) firms prefer to lay off junior and/or unskilled labor rater than senior and/or highly qualified labor; c) workers prefer to get fired than quitting (also) because of laws mandating compensation; d) Outside options, for both firms and workers are highly risky.
Monetary Policy: The potential benefits of a one-time, unexpected monetary expansions might be outweighed by the costs associated with the private sector loosing confidence in the Central Bank’s commitment to particular, commonly known, easy to forecast rules. The “rules of the game” are valuable not only because they facilitate interaction but also because they reduce uncertainty.
Liquidity: The phrase “liquidity preference” is highly misleading because it suggests that there are “tastes” for liquidity, analogous to tastes for risk or tastes for consumption goods. In fact, one’s optimal choice of liquidity, in terms of one’s portfolio, is endogenous, depending on the environment one faces.
Banking: The financial sector is supposed to match lenders with borrowers. A successful lender-borrower match will generate a surplus and the bank, depending on its bargaining power, can secure some of that surplus. A dysfunctional banking systems acts much like a shock to any other intermediate sector (e.g. an “oil shocks”), it shows up as depressed productivity in the aggregate production function and discourages present and future economic activity.
The Firm: It’s sometimes said that for most firms, demand is the binding constraint, not cost, the exact opposite of the Marshallian firm model studied in Principles and Intermediate classes. If firms set MC equal to P they would find themselves unable to sell all that they would want to produce. On the other hand, a firm cannot raise its price because the price elasticity of demand at the firm level is high and customers would just go to another quantity-rationed firm.