From the introduction to a new paper by Paul Beaudry and Franck Portier:

Changes in measured macroeconomic activity are mainly driven by changes in the volume of trade between individuals. There are two types of causes to changes in volume of trade between individuals. On the one hand, it may be that agents’ capacity to exploit existing gains from trade changes over time due to changes in the importance of frictions or policy interventions. On the other hand, it may be that the actual scope in gains from trade between individuals changes. It would therefore seem natural that macro-economic models have at their core agents which gain from trading with each other in goods markets. However, many modern macro-models do not have such a feature as they either adopt a representative agent setup or introduce heterogeneity among individuals that leads only to trades in assets, not intra-temporal trade in goods. The aim of this paper is to illustrate how an explicit treatment of intra-temporal gains from trade in a macro setting provides new insights regarding the mechanisms behind fluctuations and the effects of policy. The main assumption we introduce in the model to generate trade among individuals is the notion that not all agents are perfect substitutes in the production of goods from different sectors.

What follows is a whole lot of fancy math. Eventually, you get to this:

this set of boom bust cycles in hours happens with very little change in inflation (although inflation is still slightly procyclical over the period, it is barely moving). Using traditional logic, it would appear doubtful that these cycles can be caused by changes in aggregate demand as inflation would be expected to rise as a natural byproduct of a demand driven cycle. At the same time, it is interesting to note that labor productivity and employment have been essentially uncorrelated over the period. This suggests that these cycles were not driven by cyclically high level of productivity. Finally, we can see from the investment to output ratio series that cycles are essentially large booms and bust in investment. The question then arises what has caused these cycles if they are neither driven by neither technological supply shocks nor demand surprises. One easy interpretation of these cycles is that they represent cycles driven primarily by the expectation that current investment will have a high return. The analysis of this paper provides a simple mechanism which can explain the resulting boom-bust cycle while simultaneously being consistent with stable inflation. For the same reasons that long and protracted periods of aggregate booms can be non-inflationary, in our framework recessions are not necessarily creating strong deflationary pressures which is also a feature of the period.

The notion of “cycles driven primarily by the expectation that current investment will have a high return” has connections to Austrians as well as to Keynes, but the authors do not make such connections. The empirical point, that we see large changes in employment combined with relative small changes to inflation, poses a problem for a number of macro model of the AS-AD sort. The AS-AD monetary contraction story would be:

1. Monetary contraction lowers nominal GDP relative to trend.
2. Lower inflation leads to higher real wages, reducing labor demand.
3. Real output is reduced.

Scott Sumner tends to replace step (2) with hand-waving, at least for describing the current cycle. See my earlier post on Tyler Cowen’s concerns.