Monetary policy in a Two–Agent New Keynesian (TANK) model with two sectors

Abstract

The objective of this study is to analyze the implications of consumer heterogeneity for monetary policy in a setting simpler than a general heterogeneous agent model. I present a Two–Agent New Keynesian (TANK) model with two sectors, extending the “Limited Asset Markets Participation” (LAMP) model of Bilbiie (2008) by including a second sector based on Woodford (2003) and Woodford (2010) and, following Cravino, Lan, and Levchenko (2018), the possibility that expenditure shares are different for each household type. The model includes three sources of heterogeneity that have been shown to be relevant to policy (Cravino et al. (2018), Clayton, Jaravel, and Schaab (2018)): Two household types (called ‘poor’ and ‘rich’) that differ in their ability to smooth consumption, two final good sectors with different degrees of price stickiness and subject to different technology shocks and, unequal expenditure shares of each type for each final good. I also present a method of aggregation following Cravino et al.’s (2018) construction that makes it possible to express the model equations in a manner similar to that of the baseline three–equation New Keynesian model. An analysis of impulse–responses under different parameter specifications reveal that the responses of variables when expenditure shares vary for each type are generally spread around the response when poor households’ share equals 0.5. I find that having a larger expenditure share for the relatively flexible price sector dampens the effects of monetary policy on both the poor and rich households’ consumption compared to the case where the shares are equal.