Research

Working Papers

We propose factor-based out of sample forecasting models for US dollar real exchange rates. We estimate latent common factors employing an array of data dimensionality reduction approaches that include the Principal Component Analysis, Partial Least Squares, and the LASSO for a large panel of 125 monthly frequency US macroeconomic time series data. We augment two benchmark models, a stationary autoregressive model and the random walk model, with estimated common factors to formulate out-of-sample forecasts of the real exchange rate. Empirical findings demonstrate that our factor augmented models outperform the benchmark models at longer horizons when factors are extracted from real activity variables excluding financial sector variables. Factors obtained from financial market variables overall play a limited role in forecasting. Our data-driven models tend to perform better than models with international factors that are motivated by exchange rate determination theories.

In this paper, we propose a factor-based out of sample forecasting model for the real and nominal US $/Korean won exchange rates. We make use of a large panel of monthly frequency US and Korean macroeconomic time series data separately and then combine them together to gauge their predictability of the exchange rates. We use an array of data-dimensionality reduction methods that include the Principal Component Analysis, Partial Least Squares, and the LASSO. We focus on the level exchange rates as well as the returns. We augment the two benchmark models, a stationary autoregressive model and the nonstationary random walk model, with estimated common factors to formulate out-of-sample forecasts of the real and nominal exchange rates. We find that overall, American factors have superior predictive contents for US dollar/won real exchange rates. Adding Korean factors results in poorer performance of out-of-sample predictability of our models. Stationary models outperform difference stationary forecasting models, although the latter perform relatively better in the short run. Also, Financial factors provide useful information in the short-run, while real factors provide useful information in the long-run. We also show that factors that are estimated with data-driven approaches, such as ours are more useful than those motivated by economic propositions of exchange rates, such as purchasing power parity, uncovered interest parity, real interest parity, and the monetary model of exchange rates.

In this paper, we study the role of peoples’ attitudes on their labor market behavior. Focusing within a household, we estimate how one’s labor market decisions are dependent on their partner’s labor market outcomes, and how these decisions are driven by their culture component. Historically, man has been associated as the primary earner in a family. We argue that culture might play a role in determining a person’s labor market outcomes as it induces an aversion to the situation of when the wife earns more than the husband. We find that husbands increase their participation in the labor market if their wives earn more and this effect is even more prominent if they are from a country where people have the traditional view that man should be the primary bread- winner for the family. However, wives do not exhibit any such behavior. We argue that this irregularity is explained by the role that culture plays on forming labor market decisions. This result is important as it might contribute to the explanation of the slowdown in the convergence of the gender gap in the recent past.

Many papers in the past literature provide evidence on the impact of athletic performance on various school outcomes. This paper uses the weekly college football poll by the organization Associated Press (AP), to investigate the effect of a college team ranked in top 25 on various school outcomes such as revenues and expenses of school, coaches’ salary and enrollment. The college football poll also known as AP poll conducts weekly voting to assign the teams certain points based on which these team are ranked. First, by exploiting the discontinuity arising due to the points of 25th ranked versus 26th ranked team, we verify the visibility of a school using google trends. Secondly, our results provide evidence of the impact of this visibility of being in top 25 on positive school outcomes.

​Work in Progress

  • Behera S., Kim H., Wage Rigidity and Hysteresis

  • Behera S., Kim H., Unemployment: Hysteresis or not? - An Analysis using Threshold Model

  • ​Behera S., Sadana D., The Impact of FMLA on Childrens’ Future Outcomes

  • Behera S., Sadana D., The Effect of Contraceptive Pill on Crime