Economics and Accounting
Ph.D. Economics, Harvard University
Fields: Macroeconomics, Finance, Behavioral Economics, Expectations, and Learning
Office Phone: 508 793 3873
Office: Stein Hall 505
This paper examines key facts about the U.S. housing market. The price to rent ratio is highly volatile and significantly autocorrelated. Returns on housing are positively autocorrelated. The price to rent ratio is negatively correlated with future returns on housing and future rent growth. Finally, housing returns exhibit significant time varying volatility. I show that a benchmark rational expectations general equilibrium asset pricing model is inconsistent with these facts. I modify the model in two ways to improve its fit with the data. First, I allow for pricing frictions so prices adjust slowly to their fundamental value. Second, I assume the agent does not know if housing fundamentals, captured by rental flows, are stationary or non-stationary and has changing beliefs depending on how well each model fits the current data. I find that these modifications allow the model to increase the volatility of the price to rent ratio and to match the autocorrelation of housing returns. The price to rent ratio then negatively forecasts returns and rent growth. Finally the model generates time varying volatility consistent with the data.
I consider a real-business cycle, DSGE model where consumption is a function of the present discounted value of wage and capital income. The agent is uncertain if these income variables are stationary or non-stationary and puts positive probability on both representations. The agent uses Bayesian learning to update his probability weights on each model and these weights vary over time according to how well each model fits the data. The model exhibits an improved fit to the data relative to the rational expectations benchmark. The model requires half the level of exogenous shocks to match the volatility of output and still matches the relative volatilities of key business cycle variables. The model lowers the contemporaneous correlation of consumption and wages with output and generates positive autocorrelation in model growth rates. Impulse responses exhibit persistent responses and consistent with survey evidence forecast errors are positively serially correlated. Finally, in contrast to the existing literature, the model endogenously generates observed time varying volatility and long run predictability of business cycle variables, especially for investment.
Equity Return Predictability, Time Varying Volatility and Learning About the Permanence of Shocks (Presentation Slides) Forthcoming, Journal of Economic Behavior and Organization
I consider a consumption based asset pricing model where the consumer does not know if shocks to dividends are stationary (temporary) or non-stationary (permanent). The agent uses a Bayesian learning algorithm with a bias towards recent observations to assign probability to each process. While the true process is stationary, the consumer's beliefs change as he misinterprets a drift in dividends from their steady state value as an increased likelihood that the dividend process is non-stationary. Belief changes result in large swings in asset prices which are subsequently reversed. The model then is consistent with a broad array of asset pricing puzzles. It predicts the negative correlation between current returns and future returns and the PE ratio and future returns. Consistent with the data, I also find that consumption growth negatively correlates with future returns and the PE ratio and consumption growth forecast future consumption growth. The model amplifies return volatility over the benchmark rational expectations case and exactly matches the standard deviation of consumption. Finally, the model generates time varying volatility consistent with the data on quarterly equity returns.
“Restructuring China’s Research Institutes: Impact on China’s Research Orientation and Productivity” Economics of Transition, Vol. 24, Pages 163–208, January 2016 (joint w/ Gary Jefferson and Renai Jiang) Working Paper Version
Over the past two decades, China has exhibited a dramatic surge in its R&D intensity and patenting output. The institutional arrangements underlying this surge are somewhat of a mystery. One notable event was the Chinese government’s restructuring program for the country’s approximately 5,000 research institutes, begun in 1999. By investigating the impact of the restructuring, this paper examines the institutional side of China’s science and technology takeoff. Using aggregate data, the paper first reviews the evolution of China’s research institute sector over the period 1995 to 2010. This establishes a context for a detailed analysis of a large sample of the research institutes spanning 1998 to 2005. A central challenge of the paper is to control for the selection bias that is likely to be associated with the selective conversion process. Applying OLS and fixed effects methods, while also conducting an event study and using propensity score analysis, we find that the restructuring program appears to be moving toward its goals. While the converted S&T enterprises shifted toward a more commercial mission, the institutes that were converted to non-profit research institutes have focused on a more research-oriented mission, involving the use of government grants and becoming more engaged with basic research.
"Credit Constraints, Learning and Aggregate Consumption Volatility", Macroeconomic Dynamics, Vol. 18, Pages 338–368, March 2014. Working Paper Version
This paper documents three empirical facts. First, the volatility of consumption growth relative to income growth rose from 1947-1960 and then fell dramatically by 50 percent from the 1960s to the 1990s. Second, the correlation between consumption growth and personal income growth fell by about 50 percent over the same time period. Finally, the absolute deviation of consumption growth from its mean exhibits one break in U.S. data, and the mean of the absolute deviations has fallen by about 30 percent. First, I find that a standard dynamic, stochastic general equilibrium model is unable to explain these facts. Then, I examine the ability of two hypotheses: a fall in credit constraints and changing beliefs about the permanence of income shocks to account for these facts. I find evidence for both explanations and the beliefs explanation is more consistent with the data. Importantly, I find that estimated changes in beliefs about the permanence of income shocks have significant explanatory power for consumption changes.
“Endogenous Separation, Wage Rigidity and the Dynamics of Unemployment” Journal of Macroeconomics, Vol. 38, Part B, Pages 179-191, Dec. 2013. Working Paper Version
Previous attempts to generate sufficient unemployment volatility in the Mortensen-Pissarides model rely on either endogenous separation or wage rigidity. In this paper I simulate a version of the Mortensen-Pissarides (MP) model with both wage rigidity and endogenous separation. I find the model generates sufficient volatility in unemployment, the separation rate and the finding rate, 75% of the observed volatility in vacancies, and 70% of the Beveridge curve (the negative correlation between unemployment and vacancies). The model matches the volatility of the average wage and does not generate counterfactually low responses of the wage of new hires to productivity and unemployment. I then simulate the model while restricting the separation rate to be constant and show that the model predicts only 70% of the variance of unemployment though the model is more consistent with the volatility of vacancies and the Beveridge curve. I conclude that finding rate fluctuations explain 70% of unemployment fluctuations halfway in between the most prominent estimates in the literature.
I compare unemployment expectations from the Michigan Survey of Consumers to VAR forecastable movements in unemployment. I document three key facts: First, one-half to one-third of the population expects unemployment to rise when it is falling at the end of a recession, even though the VAR predicts the fall in unemployment. Second, more people expect unemployment to rise when it is falling at the end of a recession than expect it to rise when it is rising at the beginning of a recession even though the VAR predicts these changes. Finally, the lag change in unemployment is almost as important as the VAR forecast in predicting the fraction of the population that expects unemployment to rise. Professional forecasters do not exhibit these discrepancies. Least squares learning or real time expectations do little to help explain these facts. However, delayed updating of expectations can explain some of these facts, and extrapolative expectations explains these facts best. Individuals with higher income or education are only slightly less likely to have expectations which differ from the VAR, and those whose expect more unemployment when the VAR predicts otherwise are 8-10 percent more likely to believe it is a bad time to make a major purchase.
"Objective and self-report work performance measures: a comparative analysis" (Joint with: Glenn Pransky, Stan Finkelstein, Ernst Berndt, Margaret Kyle, Joan Mackell) International Journal of Productivity and Performance Management, 2006, Vol. 55, Issue 5, pp. 390 - 399
Purpose – The purpose of this paper is to assess the feasibility and comparability of daily self-report and objective measures of work performance in complex office tasks, and factors affecting the correlation between these measures. Design/methodology/approach – Medical bill auditors provided daily information for 12 weeks through interactive voice response (IVR) on their speed, concentration and accuracy at work, compared to their best job performance. Findings – The paper found that 124 of 142 recruited subjects (87 percent) completed > 50 percent of daily IVR reports. Concentration, speed and accuracy were highly inter-correlated (
“Measuring health impacts on work performance: Comparing subjective and objective reports,” (Join with with Glenn Pransky, Ernst Berndt, Stan Finkelstein, Maragret Kyle and Joan Mackell), Value in Health 5(6), Pages 448-449, 2002.
Econ 256 -- Macroeconomics [Syllabus]
In this course we first discuss how to measure inflation, national income, and unemployment. Then we develop models of how policy affects these variables. We study both the short-term and long-run effects of policy on economic well-being as well as the most influential theories of long-run economic growth.