Aggregate demand, sunk costs and discontinuous adjustments in an amended new consensus model

ABSTRACT

In standard new consensus macroeconomics models, the impact of shocks disappears until the economy reaches a time-independent steady-state equilibrium. Introducing sunk costs and capital indivisibilities in capacity adjustment decisions implies the rejection of asymptotic stability and a reconsideration of the relevance and usefulness of traditional steady-state analysis based on a fixed and exogenous ‘center of gravity’. Moreover, effective demand and Keynesian discretionary policies regain a central role in economic policy by determining the transient equilibriums that emerge endogenously.

KEYWORDS:

• Path-dependence
• hysteresis
• potential output
• monetary policy

JEL CODES:

• C62
• E12
• E17
• E47
• E58

Acknowledgement

This manuscript is part of my PhD dissertation at the University of Rome “La Sapienza” and the Centre d’Economie de Paris Nord (CEPN) of the University “Paris 13, Sorbonne Paris Cité”. I wish therefore to acknowledge my supervisors Dany Lang and Luca Zamparelli for their important contribution. Early versions of this paper were presented at the 26th annual conference of the European Association for Evolutionary Political Economy (EAEPE); at the Séminaire d’Economie Politique of the CEPN; at the Doctoral Students Seminar of the Doctoral School of Economics of “La Sapienza”. I wish to acknowledge all the participants in these conferences and workshops for their useful feedbacks, particularly Paolo Piacentini, Andrea Roventini, Carlo D’Ippoliti, Cédric Durand and Claudio Sardoni. Finally, I am grateful to the anonymous referees for their useful comments. The usual disclaimers apply.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 See Elster (1976) and Cross (1993, 1998) regarding the epistemological importance of hysteresis.

2 Palacio-Vera (2009) shows that, by introducing a non-linearity in the price equation, temporary demand shocks may have permanent effects, despite the potential output dynamics modeled via Equation 1. Nevertheless, in this article we consider NCM models with a linear accelerationist Phillips curve and a hysteretic potential output equation.

3 might represent either a stochastic demand shock or a fiscal policy shock. In the remainder of this article we will treat it as a fiscal policy shock.

4 This characterization of the accelerationist Phillips curve implies simple backward-looking expectations. For an overview of different inflation rules with forward-looking expectations, refer to Kapadia (2005)

5 The values of the parameters used during the simulations are provided in Appendix 1.

6 It can be argued that standard new consensus models do not aim at explaining growth, but rather business cycle fluctuations, and assume potential output stability only for simplicity. According to Taylor (2000b), potential output is obtained through a Solow model augmented with endogenous technological change. However, the decision to clearly distinguish long-run growth behavior and short-run cyclical behavior must rely on the hypothesis of substantial independence of the growth path from business-cycle fluctuations. Clarida, Gali, and Gertler (2000) specify potential output as an AR(1) process: potential output fluctuations are explained through stationary and stochastic exogenous shocks, independent of business-cycle fluctuations. In other words, even though potential output is assumed to be theoretically endogenous, it is actually treated as if it were exogenous.

7 The specific random distribution does not affect the results and properties of the model (Piscitelli et al. 2000).

8 During simulations, random numbers are actually generated according to a ‘pseudo-random’ data-generating algorithm, and the seed represents the starting point of the sequence of ‘pseudo-random’ numbers. Hence, if one runs simulations using the same seed, the sequence of ‘pseudo-random’ numbers will always be the same. Running simulations with different seeds, on the other hand, implies assuming different sequences of ‘pseudo-random’ numbers, which is tantamount to assuming different initial conditions. In our model, for instance, the time zero of simulations is set after a finite sequence of random shocks. Therefore, assuming the same random seed implies assuming the same sequence of random shocks and, consequently, the same structural conditions of the system in the time zero of simulations; assuming different random seeds implies, on the other hand, assuming different sequences of random shocks and, consequently, a different initial structure of the economy.

9 We show in figure 4 the results of the simulations with the same random seed. Indeed, results are consistent and robust whatever the random seed chosen.

10 In new consensus models, the central bank neutralizes demand shocks through the natural rate of interest. Since the exogenous and stochastic shock is, by assumption and by construction, a non-anticipated (fiscal policy) shock, it does not affect the natural rate of interest. A permanent exogenous shock to can be interpreted either as a non-anticipated, long-lasting shock or as a correctly anticipated but not neutralized shock. For the same reasons, a temporary shock to must be interpreted either as a non-anticipated temporary shock or as an anticipated structural shock that is neutralized through the natural rate of interest (in this case, the central bank will identify the permanent stochastic shock as an increase in the structural disturbance .