مقاله سال ۲۰۱۶

In leading theories of monetary non-neutrality, the policy prescription for the optimal steady-state inflation rate varies between the negative of the real interest rate (the Friedman rule) and zero (price stability); see Schmitt-Grohé and Uribe (2010), for an overview. In this paper we explore a new channel where the interaction of nominal wage and labor market search and matching frictions affects the planner’s trade-off between the welfare costs and benefits of inflation. We show that the combination of such frictions can in fact generate a Ramsey optimal inflation rate that is significantly positive. Importantly, this is the case even in the presence of a monetary friction, which drives the optimal inflation rate towards the Friedman rule of deflation.

The mechanism we have in mind arises in a model with search frictions when nominal wages are not continuously rebargained and some newly hired workers enter into an existing wage structure.1 In this case, we show in a stylized model that inflation not only affects real-wage profiles over a contract spell, but also redistributes surplus between workers and firms, since incumbent workers impose an externality on new hires through the entry wage. Specifically, this affects the wage-bargaining outcome through the workers’ outside option and hence the expected present value of total labor costs for a match as well as firms’ incentives for vacancy creation. We derive a Hosios condition for the stylized model and show that the Ramsey planner has incentives to increase inflation if employment and vacancy creation are inefficiently low in order to push the economy towards the efficient allocation.

It is worth noting that this incentive vanishes at an efficient allocation, as in Thomas (2008) where the calibration is chosen so that the Hosios condition holds, and hence that search and wage setting externalities cancel out in steady-state (and the reverse occurs if employment is inefficiently high). Moreover, the model of Erceg et al. (2000) does not feature this mechanism either, since there is no extensive margin on the labor market in that model and hence the Ramsey planner has no leverage on job creation through the channel outlined above. However, and similar to our model, the Erceg et al. (2000) model features a markup in wage-setting where the actual markup can be different from the flexible price markup because of Calvo (1983)-style wage stickiness. The planner has incentives to tilt the real-wage profile in order to lower the actual markup and increase labor input. Note though, since the model lacks a leverage on job creation there is much less of a motive for the planner to use this channel and the effect of using inflation to affect the average markups in the economy is tiny.2 The mechanism also vanishes if search frictions vanish since the Ramsey planner loses any leverage over vacancy/job creation. Thus, models without an extensive margin on the labor market lack the mechanism described here. Furthermore, the Ramsey planner loses the ability to affect real-wage costs via inflation if all new workers get to rebargain their wage. In this case, the full effect of inflation on entry wages is internalized in the wage bargain, and firm and worker surpluses, as well as real wage costs, become neutral to inflation.