photo Harvard University - Economics Department
Home News About Us Faculty Staff Courses Graduate Undergraduate Visiting Scholars Journals Events Classrooms Links

Ian Dew-Becker's Papers

­­

Job-Market Paper:

­ ­ ­

A Model of Time-Varying Risk Premia with Habits and Production

Abstract:
I develop a new utility specification that incorporates Campbell–Cochrane–type habits into the Epstein–Zin class of preferences. In a simple calibration of a real business cycle model with EZ-habit preferences, the model generates a strongly countercyclical equity premium, substantial equity return predictability, and a stable riskless interest rate, as in the data. Moreover, conditional on the average level of risk aversion, time-variation in risk aversion increases the volatility and mean return of equities. On the real side, the model matches the short and long-term variances of output, consumption, and investment growth. As an additional empirical test, I measure implied risk aversion and find that it has an R˛ of over 50 percent for 5-year stock returns in post-war data. Variables that predict stock returns in the data also predict returns in the model with a similar degree of explanatory power.


Working Papers:

­ ­ ­

Investment and the Cost of Capital in the Cross-Section: The Term Spread Predicts the Duration of Investment Draft of 5.2011.

Abstract:
I study the determinants of investment in assets with different depreciation rates. When physical capital is priced like a bond with a similar duration, a high term spread is associated with low average duration for investment. Thirty percent of the cross-sectional variation in investment is explained by the term spread, implying an important role for the cost of capital in determining the composition of investment. The results are robust to including a variety of controls.

­ ­ ­

Bond Pricing with a Time-Varying Price of Risk in an Estimated Medium-Scale Bayesian DSGE Model Draft of 11.2011.

Abstract:
A New-Keynesian model in which households have Epstein–Zin preferences with time-varying risk aversion and the central bank has a time-varying inflation target can match the dynamics of nominal bond prices in the US economy well. The model generates a steady-state term spread of 152 basis points, compared to the sample average of 207 basis points. The fitting errors for individual bond yields are roughly as large as those obtained from a non-structural three-factor model, and two thirds smaller than in structural models with constant risk aversion or a constant inflation target. The term premium is estimated to have a standard deviation half as large as that of the term spread.
The model delivers rich variance decompositions for the pricing kernel and the real side of the economy. Shocks to risk aversion account for less than 5 percent of the variation in output, consumption, and investment growth at business-cycle frequencies, but 32 percent of the variation in the term spread. There is little connection between priced risk factors and output, consumption, investment, or hours worked in the short-run.

­ ­ ­

Essentially Affine Approximations for Economic Models Draft of 11.2011.

Replication files
Show/hide abstract:

This paper proposes a novel first-order approximation technique for standard economic models with stochastic volatility, risk aversion, or disaster risk. It is identical up to the first order to perturbation, but it includes terms that perturbation would treat as "higher-order" that follow from the use of an essentially affine stochastic discount factor in the Euler equations. I calculate Euler equation errors for the RBC model with time-varying risk aversion, volatility, and disaster risk and find that the essentially affine approximation has accuracy between that of second and third-order perturbations. The equilibrium dynamics take a fully linear state-space form, so models can be estimated with the Kalman filter, rather than a more computationally intensive nonlinear filter. The approximation encompasses a variety of well-known methods specialized for use in particular settings, including general equilibrium models, models of time-varying risk aversion, portfolio choice, and endowment-economy asset pricing.

­ ­ ­

Estimates of the volatility of the permanent component of consumption and their implications for asset pricing Draft of 12.2011 available on request.

Show/hide abstract:

The long-run variance of consumption (the variance of its permanent innovations) is the key moment in the endowment process that determines the price of risk in models where households have Epstein--Zin preferences with high values for the coefficient of relative risk aversion and the elasticity of intertemporal substitution. I study a broad range of methods for estimating the long-run variance. The methods give surprisingly consistent results. We can reject a random walk for consumption, but the long-run volatility is far smaller than the values used in calibrations in the long-run risks literature. Consumption-based asset pricing models should be calibrated so that the quarterly long-run standard deviation of consumption growth is no more than 2.0 percent; the preferred point estimate is 1.11 percent.


Publications:

­ ­

Unresolved Issues in the Rise of American Inequality (with R.J. Gordon). Brookings Papers on Economic Activity 38(2), Fall 2007.

Show/hide abstract:
This paper provides a comprehensive survey on six aspects of rising inequality: changes in labor's share, inequality at the bottom, inequality at the top, labor mobility, inequality in consumption as contrasted to inequality of income, and international differences in inequality, particularly at the top. Outside the scope of this paper are changes in the rate of return to higher education, the evolution of the college wage premium, and the mechanisms by which family human capital is transmitted to infants and children. We conclude that changes in labor's share play no role in rising inequality of labor income; by one measure labor's income share was almost the same in mid-2007 as in 1950. Within the bottom 90 percent as documented by CPS data, movements in the 50-10 ratio are consistent with a role of decreased union density for men and of a decrease in the real minimum wage for women, particularly in 1980-86. There is little evidence on the effects of imports, and an ambiguous literature on immigration which implies a small overall impact on the wages of the average native American, a significant downward effect on high-school dropouts, and potentially a large impact on previous immigrants working in occupations in which immigrants specialize.

The literature on skill-biased technical change (SBTC) has been valuably enriched by a finer grid of skills, switching from a two-dimension to a three- or five-dimensional breakdown of skills. We endorse the three-way "polarization" hypothesis that seems a plausible way of explaining differentials in wage changes and also in outsourcing.

To explain increased skewness at the top, we introduce a three-way distinction between market-driven superstars where audience magnification allows a performance to reach one or ten million people, a second market-driven segment consisting of occupations like lawyers and investment bankers, and a third segment consisting of top corporate officers. Our review of the CEO debate places equal emphasis on the market in showering capital gains through stock options and an arbitrary management power hypothesis based on numerous non-market aspects of executive pay.

The paper concludes that data on consumption inequality are too fragile to reach firm conclusions, and a perspective on international differences that blends institutional and market-driven explanations.

­ ­

Where Did the Productivity Go? Inflation Dynamics and the Distribution of Income (with R.J. Gordon). Brookings Papers on Economic Activity 36(2), 2005, pp. 67–127.

Show/hide abstract:
A basic tenet of economic science is that productivity growth is the source of growth in real income per capita. But our results raise doubts by creating a direct link between macro productivity growth and the micro evolution of the income distribution. We show that over the entire period 1966-2001, as well as over 1997-2001, only the top 10 percent of the income distribution enjoyed a growth rate of real wage and salary income equal to or above the average rate of economy-wide productivity growth. Growth in median real wage and salary income barely grew at all while average wage and salary income kept pace with productivity growth, because half of the income gains went to the top 10 percent of the income distribution, leaving little left over for the bottom 90 percent. Half of this inequality effect is attributable to gains of the 90th percentile over the 10th percentile; the other half is due to increased skewness within the top 10 percent.

In addition to its micro analysis, this paper also asks whether faster productivity growth reduces inflation, raises nominal wage growth, or raises profits. We find that an acceleration or deceleration of the productivity growth trend alters the inflation rate by at least one-for-one in the opposite direction. This paper revives research on wage adjustment and produces a dynamic interactive model of price and wage adjustment that explains movements of labor's share of income.

What caused rising income inequality? Economists have placed too much emphasis on "skill-biased technical change" and too little attention to the sources of increased skewness at the very top, within the top 1 percent of the income distribution. We distinguish two complementary explanations, the "economics of superstars," i.e., the pure rents earned by sports and entertainment stars, and the escalating compensation premia of CEOs and other top corporate officers. These sources of divergence at the top, combined with the role of deunionization, immigration, and free trade in pushing down incomes at the bottom, have led to the wide divergence between the growth rates of productivity, average compensation, and median compensation.


Older Papers:

­ ­ ­

An Analysis of the Superior Performance of Empirical Likelihood over GMM Draft of 12.2008.

  • Slides
  • Replication files to create the main results from the paper. Replication files for the remainder of the results are available on request, but they entail weeks to months of computing time.
    •
Show/hide abstract:

I examine the relative performance of the Generalized Method of Moments (GMM) and Empirical Likelihood (GEL) in an empirical example and in a large set of Monte Carlo simulations. Based on the example, I show that GMM provides estimates that vary widely depending on the set of moment conditions used and cannot be reconciled with one another (based on a Hausman test), while GEL estimates are statistically indistinguishable. I then calibrate simulations to match the empirical sample and show that GMM is substantially biased in this setting and EL is not. Moreover, the performance of GMM degrades under weak identification much more quickly than that of EL. Continuously updated GMM (CUE) is essentially identical to EL in the simulations, and would be a good replacement for GMM in many empirical settings. I investigate the power of expansions from Newey and Smith (2004) that give the asymptotic bias of these estimators and find that they are helpful in understanding the patterns of estimates from GMM and EL.

­ ­ ­

The Role of Labor Market Changes in the Post-1995 Slowdown in European Productivity Growth (with R.J. Gordon). Draft of 12.2007.

  • Slides
  • Vox EU column summarizing this paper
  • Titles of earlier drafts:
            "Why Did Europe's Productivity Catch-up Sputter Out? A Tale of Tigers and Tortoises"
            "The Slowdown in European Productivity Growth: A Tale of Tigers, Tortoises and Textbook Labor Economics"
    •
Show/hide abstract:

Throughout the postwar era until 1995 labor productivity grew faster in Europe than in the United States. But since 1995, productivity growth in the EU-15 has slowed while that in the United States has accelerated. But Europe’s productivity growth slowdown was largely offset by faster growth in employment per capita, leaving little difference in growth of output per capita between the EU and US going back to 1980. This paper is about the strong negative tradeoff between productivity and employment growth. We document this tradeoff in the raw data, in regressions that control for the two-way causation between productivity and employment growth, and we show that there is a robust negative correlation between productivity and employment growth not only across countries and time, but also across countries and industries.

We simplify the task of explaining intra-EU differences in the performance by reducing the dimensionality of the issue from the 15 EU countries to four EU country groups, chosen by geography. We provide a comprehensive analysis of the role of policy and institutional variables in causing changes in productivity and employment per capita growth across these country groups. Using both a calibrated theoretical model and several reduced-form regressions, we document the strong effects of European policies that raised labour costs, such as the tax wedge, employment and product market regulation, unemployment compensation, and union density, in causing employment to fall and productivity to rise before 1995, and for this process to be reversed after 1995.

Our paper concludes with policy implications, and we propose a new framework for thinking about EU policy reforms. The strong evidence that we find for a productivity-employment growth tradeoff, across countries, time, and industries, changes the questions that European policymakers should be asking. They should no longer ask how they should boost productivity growth or raise employment growth. Most policies will push productivity and employment in opposite directions, and we have shown that these offsetting effects make the effects of policies on growth in output per capita ambiguous. Our new policy framework suggests that policy changes be assessed as much on their effects on government budgets as on productivity or employment, since the productivity-employment tradeoff causes some policy changes to have a negligible effect on growth in output per capita.

­
Harvard © 2007 by the President and Fellows of Harvard College