This paper carries the analysis one step further. It argues that because different prices (including prices of labor and capital) are determined in different ways, shocks lead to marked changes in relative prices, and those disturbances in relative prices greatly exacerbate economic fluctuations.[6] The determination of asset prices is often best described by the auction markets emphasized in traditional economics texts. These prices adjust quickly; and because the value of assets today depends on expectations of future values, current prices can be highly volatile, as those expectations change.[7] On the other hand, prices of most products, wages, and bank interest rates are set by firms (and banks), albeit in the context of markets. Firms face downward-sloping demand curves for their products in the short run and upward-sloping supply curves for factors. These slopes can be explained in part by imperfections in competition due to product differentiation or to the small number of firms in the market and in part by information imperfections.
The importance of downward-sloping product demand (at least in the short run) has been brought home forcefully by the recent crisis in East Asia. Consider the example of Thailand, which, with less than 1 percent of world GDP, is a small player in the world economy.[8] Under standard theory, it would face a horizontal demand curve for its products; certainly a 30 percent real devaluation should lead to huge increases in the demand for its products. Exports would quickly fill up any loss in domestic demand. Yet in fact the value of Thai exports actually declined in the aftermath of the crisis.[9]
For price- and wage-setting firms, the consequences of changes in prices and wages are uncertain, not only because prices and wages have to be set before demand and supply curves are fully known, but also because the reactions of other agents in the market are uncertain. The uncertainty arises for two reasons: first, because changes in a firm’s circumstances (including the signals they receive) are only imperfectly known by other firms, and second, because firms receiving the same information or the same shocks will react differently, depending on their circumstances and characteristics (which are not common knowledge). Both because of agency problems and imperfections of capital markets, which lead to firms acting in a risk-averse manner[10], firms are sensitive to the risks associated with different decisions. The problem of price- and wage-setting thus should be approached within a standard dynamic portfolio model, one that takes into account the risks associated with each decision, the non-reversibilities, as well as the adjustment costs associated with both prices and quantities.[11] (See Figure 1.) In our earlier papers, we argued that (a) the risks associated with wage and price adjustments may well be larger than those associated with output adjustments, at least for goods that could be stored; and (b) there were fixed costs associated with hiring and firing, and those costs were often asymmetric. These asymmetries helped explain the pattern of hours and employment over the cycle, with increases in hours (entailing high overtime payments) typically preceding increases in employment, and with labor-shedding typically lagging downturns (giving rise to the phenomenon described as labor-hoarding.)
In the following subsections, we look more closely at the price-setting process, providing further insights into why prices exhibit rigidities and why different prices may adjust at different rates. We then explore the consequences of asymmetric price responses.