The concept of the menu cost has originally introduced by Eytan Sheshinski and Yoram Weiss (1977) in their paper looking at the effect of
inflation on the frequency of price changes. Sheshink and Weiss concluded that even fully anticipated
inflation results in an actual menu cost for the business. They suggested that
businesses will change prices in discrete jumps rather than continual changes when in an
inflationary environment. This justifies the fixed costs of changing prices when revenues are expected to increase. The idea of applying menu costs as an aspect of
Nominal Price Rigidity was simultaneously put forward by several
New Keynesian economists in 1985–1986. In 1985,
Gregory Mankiw concluded that even small menu costs create inefficient price adjustment and push equilibrium below the point which is socially optimal. He further suggested that the subsequent loss of welfare far exceeds the menu cost that causes it.
George Akerlof and
Janet Yellen put forward the idea that due to
bounded rationality firms will not want to change their price unless the benefit is more than a small amount. This
bounded rationality leads to inertia in nominal prices and wages which can lead to output fluctuating at constant nominal prices and wages. The menu cost idea was also extended to wages as well as prices by
Olivier Blanchard and
Nobuhiro Kiyotaki. The
new Keynesian explanation of price stickiness relied on introducing
imperfect competition with price (and wage) setting agents. This started a shift in
macroeconomics away from using the model of
perfect competition with price taking agents to use imperfectly competitive equilibria with price and wage setting agents (mostly adopting
monopolistic competition).
Huw Dixon and Claus Hansen showed that even if menu costs were applied to a small sector of the economy, this would influence the rest of the economy and lead to prices in the rest of the economy becoming less responsive to changes in demand. In 2007, Mikhail Golosov and
Robert Lucas found that the size of the menu cost needed to match the micro-data of price adjustment inside an otherwise standard business cycle model is implausibly large to justify the menu-cost argument. The reason is that such models lack "real rigidity". This is a property that markups do not get squeezed by large adjustment in factor prices (such as wages) that could occur in response to the monetary shock. Modern New Keynesian models address this issue by assuming that the labor market is segmented, so that the expansion in employment by a given firm does not lead to lower profits for the other firms. == Magnitude of menu costs ==