Research

The Effect of Monetary Policy on Unemployment and Income Distribution

This paper studies the monetary transmission effects on unemployment and income distribution using a Heterogeneous Agent New Keynesian (HANK) model. Unlike previous studies that assume unemployment risk as an exogenous part of the earnings process, the study endogenizes the dynamics of unemployment due to a monetary policy shock. The model when calibrated to the U.S. economy shows countercyclical left skewedness of earnings growth as evidenced in the empirical literature. Its prediction on the changes in skewness of earnings growth and unemployment during a contractionary phase accurately match the changes experienced by the economy during the 2001-2002 recession.

Working Paper

Heterogeneous Effects of Macroeconomic Factors on Income Inequality Dynamics

This paper studies the heterogenous effects of macroeconomic factors on income inequality dynamics for 56 countries using a dynamic linear model specification that allows for heterogeneous coefficients across cross-section and time. Heterogeneity in coefficients for different countries in the model specification is represented through latent group structures. Instead of assuming a predefined group structure based on economic development or geography, countries are assumed to group together according to common latent conditions. Group specific regression results point out that the underlying mechanisms that drive income inequality dynamics are not only dependent on the level of economic development but are also country specific.

Working Paper

Costly Interactions, Division of Labor and Non-Employment

This paper develops a model to study the conditions under which some workers may not be employed by a firm. Workers in the model are assumed to face time costs during the production process while operating with other workers on a task. That is, workers in the model forgo some proportion of their time endowment in preparation for a task or while interacting with other workers in the immediate vicinity. The presence of these costs could lead to non-employment of certain workers. For this situation, the optimal task assignment structure is analytically deduced and conditions for existence of non-employment are provided. The theory provides conditions for non-employment of low skilled and high skilled workers. In particular, workers with generic skills or those trained in specialized fields of knowledge operating in tasks requiring high level of personal interactions are likely to not be employed.

Working Paper

Work in Progress:

Estimation of Dynamic Panels with Latent Groups and Structural Breaks ( Joint with Dr.Pavel Cizek, Tilburg University)

We study the identification problem of latent group structures in a dynamic linear panel data model with fixed effects and fixed T assumption. We prove that latent group structures may not be identified because of the inconsistency of the Least Squares Dummy Variable (LSDV) estimator. In order to overcome the identification problem, we develop an estimation procedure that identifies latent group structures by eliminating individual-specific effects using data transformations that require subtracting several time averages computed from data sets with different time period sequences. Latent groups, slope coefficients and break points are then estimated from the transformed data using least squares approach. Asymptotic results show that the latent group structures are consistently estimated in a dynamic linear panel data model for the proposed estimator. Using Monte Carlo results, we demonstrate good finite sample properties of the proposed estimator.