Theses Doctoral

Essays In Heterogeneous Effects Of Monetary Policy

Mishra, Shruti

My dissertation within monetary macroeconomics focuses on uncovering the impact of micro level heterogeneity in household wealth portfolios and firm size on aggregate macroeconomic variables. Using household- and firm-level datasets, I study these outcomes in the context of exploring the effects of monetary policy shocks.

Most macroeconomic models use a representative agent framework to study the effects of monetary policy. In such models all consumers are assumed to be similar, therefore, it is only required to know the size of the monetary policy shock and its average impact to estimate the overall effect. But recent literature has emphasized the importance of agent heterogeneity for explaining observed aggregate dynamics and optimal policy design. Here, it matters which consumers get the extra income as people react differently to the shock. In a model with a realistically calibrated household balance sheet, monetary policy has redistribution effects because different agents have differential exposure to the interest rate and inflation risk born in their portfolios. For example, short-term or nominal borrowers will win from a sudden decrease in the interest rate and a sudden increase in inflation, while short-term lenders or nominal lenders will lose.

In the first chapter of the dissertation, co authored with Anastasia Burya, we study the effect of heterogeneity in consumers' portfolios on the unemployment response to monetary policy. We develop a search efforts model with heterogeneous agents and then decompose the effect of the monetary policy shock on aggregate unemployment. The direction and the magnitude of the wealth effect will determine whether people search for jobs more actively after a monetary contraction. For example, if unemployed consumers are indebted, they experience a negative wealth effect after a contraction, search for jobs more actively and increase their probability of finding a job, therefore, reducing unemployment. In this framework, the sign of the overall effect of monetary policy on unemployment will depend on whether unemployed consumers are indebted and the magnitude of their debt. We test the prediction of the model in both micro and aggregate data. To test the prediction of the model in the micro data using the PSID panel dataset, we estimate the coefficient of the interaction term between various mortgage measures and Romer \& Romer monetary policy shocks while looking at five main transition probabilities that indicate a higher increase in search efforts for indebted people after a monetary contraction: dynamic transition probability of moving from non employment to employment, moving from non participation in the labor force to employment, remaining a non participant in the labor force, remaining unemployed and taking up an extra job. In the aggregate data, we use a similar estimation approach with debt to income ratio. We also subject this to a variety of checks using age and Saiz instruments for increased robustness.

In the second chapter of the dissertation, co-authored with Anastasia Burya and Martsella Davitaya, we show that inflation expectations are anchored. If inflation expectations are anchored, then their sensitivity to monetary policy should be smaller than if they are de-anchored. When the Fed pursues inflation targeting, the market expectations of Fed's reaction should affect the response to current monetary policy shocks. We use daily bond yield data to show that the sensitivity of inflation expectations to monetary policy is lower if the Fed is more responsive to inflation during the previous CPI release. Intuitively, the Fed announcement leading to a rate change that is higher than expected from the CPI release indicates that the markets expect the Fed to react more aggressively in the future. Therefore, markets do not adjust inflation expectations as much (leading to anchored inflation expectations). The empirical strategy consists of two steps. First, we measure market expectations about the Fed's reaction to inflation by regressing the changes of different interest rates around the CPI release dates on the surprise change in CPI. Second, we estimate the sensitivity of inflation expectations' response to monetary policy based on the expectations about the Fed's reaction to inflation.

Product markets are characterized by the significant heterogeneity of demand elasticity between large and small firms. In many cases, the ability of larger firms to dictate prices is such that they are able to charge higher markups. In the third chapter, co-authored with Anastasia Burya, we develop a simple model of firms with heterogeneous market power. We connect the recent trend of increasing market power to the flattening of the Phillips Curve through the decreasing aggregate pass-through. We explore the sufficient statistic arising from this model and then proceed to estimate it in the data. Here, we consider heterogeneity in demand elasticity and superelasticity. In the recent literature as well, papers such as Baqaee, Farhi and Sangani (2021) and Wang and Werning (2020) have brought to attention that certain parameters of demand are important for various macroeconomic dynamics such as the flattening of the Phillips Curve. It was also shown that the degree of these effects depends on the demand parameters, such as elasticity and superelasticity.

We estimate these parameters in a novel format using an empirical procedure called Granular IV, which was first described in Gabaix and Koijen (2020) and makes use of the fact that in reality, unlike baseline macroeconomic models, some firms are big enough to impact the aggregates. For this estimation, we use firm-level price data from ACNielsen Retail Scanner database. Employing the novel empirical approach we estimate these relevant demand parameters. We estimate a demand elasticity of 3.23, in line with the literature. Our estimate for super elasticity is 3.74 which is in line with Marshall's second law of demand and for constant superelasticity parametrisation would signify the curvature of the demand curve between that of CES and linear demand.


  • thumnail for Mishra_columbia_0054D_17113.pdf Mishra_columbia_0054D_17113.pdf application/pdf 4.61 MB Download File

More About This Work

Academic Units
Thesis Advisors
Afrouzi Khosroshahi, Hassan
Ph.D., Columbia University
Published Here
April 13, 2022