Endogenously heterogeneous inflation expectations and monetary policy

Abstract

This study adopts insights from Heterogeneous Agent Models (HAM) in order to formalize Post Keynesian insights and further develop heterodox macroeconomics. We formulate a macrodynamic model consisting of a standard Post Keynesian investment function and conflict inflation, resulting from workers and capitalists making conflicting claims over their shares in output. The central bank targets the inflation rate by employing the nominal rate of interest as an instrument. However, in a departure from standard Post Keynesian literature, expectations about inflation are endogenously heterogeneous. An endogenously determined fraction of the agents expects the inflation target of the central bank to prevail, while the rest of the agents are trend-followers. The proportion of trend followers depends directly on the proportionate deviation of the actual rate of inflation from the target rate. We examine the stability of the model and discuss some of its dynamic properties. We find that even along the steady-state neither class is able to achieve its desired share. Further, due to the endogenously heterogeneous nature of inflation expectations, any large shock to the system might destabilize it. We reexamine the effectiveness of a monetary policy that relies solely on inflation targeting in this context.

Keywords:

• Heterogeneous expectations
• inflation targeting
• Post Keynesian

JEL CLASSIFICATIONS:

• B1
• B3
• B5

Acknowledgments

An early version of this article was presented in 2nd International Post-Keynesian and Institutionalist Conference at Grenoble, France in December 2017. The authors are grateful to Louis-Philippe Rochon and various other participants of the conference for their comments. The authors also acknowledge the useful comments of the two referees on earlier versions of this article. The usual disclaimer applies.

Notes

1 See, for instance, a summary of this literature in Lavoie (2014, chapter 8).

2 We should also note that, in addition to theoretical and empirical critiques of Taylor-type interest rate rules, a critical literature has also emerged in search for alternatives to Taylor rules. (see, for instance, Rochon 2007; Rochon and Setterfield 2007; Wray 2007; Lavoie 2014; Nishi 2015)

3 Patnaik (1997) has also argued that fiscal policy will not be successful in reconciling the class conflict over income distribution and a requirement for a high level of activity. An increase in direct tax rates on wages, profits, etc. will reduce the level of activity while it will intensify the degree of conflict over income distribution since all classes will bargain over their post-tax share in income. In other words changes in tax rates may have non-monotonic effects on the variables of the capitalist economy and therefore may be less amenable to control by the government. Further even if the direct tax rates on different classes and the ratio of government expenditure to the capital stock is successfully varied by the government to reconcile the conflict over income distribution among classes then this reconciliation may happen at a level of activity that is incompatible with social stability.

4 An earlier study by Dalziel (1990) extended this idea using a game-theoretic framework.

5 The ratio of full capacity output to the capital stock is assumed to be unity for simplicity.

6 We should point out here that we do not directly consider the dynamics of the employment rate. We assume that the supply of labor adjusts to its demand. Further work in this direction would involve a consideration of the dynamics of the employment rate and the recognition that the target share of profits in output that workers are willing to concede to capitalists is a decreasing function of the rate of employment.

7 Rowthorn (1977) pointed out that the bargaining power of capitalists rises when the degree of capacity utilization rises. The cost of higher interest payments is passed on through price increases by capitalist entrepreneurs (see, for instance, Patnaik 1997, Lima and Meirelles 2003). However we leave out these concerns in the present study for the sake of simplicity.

8 In reality, the central bank would attempt to learn about the steady state real rate of interest, $\overline{r}$ from available information. This might be put into practice in real world by the central bank by adopting an “assumed real rate of interest” as suggested by Taylor (1993), and then periodically reviewing it in light of current data. We leave the analysis of conditions under which this learning process converges to the true value of $\overline{r}$ as an exercise for future research. For the moment, we consider the naïve static expectation, that is ${\overline{r}}_{cb}^e=r$ as a first approximation.

9 In fact, results of our numerical exercise in Section “Inflation targeting with exogenously determined heterogeneous expectations” and Section “Inflation targeting with endogenously determined heterogeneous expectations” are contrary to the New Keynesian agreement on the need for active interest rate rules.

10 We endogenize υ in Section “Inflation targeting with endogenously determined heterogeneous expectations.”

11 This exogenous component has sometimes been described in the literature on investment function as “animal spirits”; see, for instance, Duménil and Lévy (1999, 686) and Dutt (2006, 326).

12 Steady state values calculated in this section using Maxima version 5.37.2.

13 We are grateful to the referee for pointing this out.

14 All numerical analyses in this study are performed using Matlab version R2017b.

15 We should note that the other steady state is economically not meaningful, as the steady state share of profits, m in this case is equal to 1 for all relevant parameter configurations.

16 A detailed discussion on active monetary policy might be found in Taylor (1993), Benhabib, Schmitt-Grohé, and Uribe (2001b), and Benhabib, Schmitt-Grohé, and Uribe (2002).

17 See, for instance, (Bernanke et al. 1999; Johnson 2002; Cornand and M’baye 2018; Gürkaynak, Levin, and Swanson 2010).

18 We should, however, note here that the extent to which the central bank targets can anchor inflation expectations might be more complex than the one discussed in this paper. One possibility, for instance, could be when the fundamentalists anchor their expectations around a level different from the central bank target. The central bank might, in such cases, attempt to persuade the fundamentalists to adopt the targets set by them. However, we leave the complications arising from these issues as possible avenues for future research.

Additional information

Notes on contributors

C. Saratchand

C. Saratchand is an Assistant Professor in Department of Economics, Satyawati College, University of Delhi, India.

Soumya Datta

Soumya Datta is Senior Assistant Professor in Faculty of Economics, South Asian University, New Delhi, India.