October 20, 2015 Dear Recruiting Chair: We are pleased to provide below the curriculum vitae and dissertation abstracts of the Penn Economics Ph.D. students who seek employment in this year's job market. Also find in the below table a summary indicating fields of interest and advisors' names. Full dissertation abstracts and research papers will be supplied directly from the candidates as they apply for positions. Each candidate is also responsible for having confidential letters of recommendation sent upon request. We encourage you to contact the faculty members who are most familiar with the students’ work (each vita contains a list of faculty references). Also, please feel free to contact either of the placement officers. If you or a member of your institution will be in the Philadelphia area and would like to meet with some of our students, Kelly Quinn, our Graduate Group Coordinator, would be pleased to arrange such interviews. She can be reached by phone 215-898-5691 or email at kquinn@econ.upenn.edu. If we can help in any way regarding the placement of this year's University of Pennsylvania students, please call or e-mail us. Sincerely, Iourii Manovskii Graduate Placement Officer manovski@econ.upenn.edu (215) 898-6880 IM-AP/kaq Andrew Postlewaite Graduate Placement Officer apostlew@econ.upenn.edu (215) 898-7350 UNIVERSITY OF PENNSYLVANIA Department of Economics 3718 Locust Walk Philadelphia, PA 19104-6297 Placement Officers: Iourii Manovskii Associate Professor of Economics manovski@econ.upenn.edu 215-898-6880 Placement Coordinator Kelly A. Quinn Direct Line: 215-898-5691 Main Department: 215-898-7701 Fax: 215-573-2057 Andrew Postlewaite Harry P. Kamen Professor of Economics apostlew@econ.upenn.edu 215-898-7350 SUMMARY LISTING OF DOCTORAL STUDENTS SEEKING EMPLOYMENT, 2015/2016 Candidate Name Murat Celik celik@sas.upenn.edu (215) 285-7848 Research Interest Macroeconomics, Economic Growth Hongseok Choi aitch.choi@gmail.com +82 10 5115 2037 Selman Erol erols@sas.upenn.edu (215) 667-5036 Rong Hai ronghai@uchicago.edu (267)254-8866 Asset Pricing, Portfolio Choice, Behavioral Finance (Ambiguity) Network Formation, Economic Theory Ju Hu Juhu1@sas.upenn.edu (267) 909-1250 Junwen (Caroline) Liu junwenl@sas.upenn.edu (215) 688-1211 Microeconomic Theory, Game Theory, Dynamic Games Labor Economics, Public Economics, Education, Consumer Finance Ekim Muyan muyan@sas.upenn.edu (267) 994-6081 Political Economics, Development Economics, Public Economics Daniel Neuhann neuhann@sas.upenn.edu (646) 238-3915 Macroeconomics, Financial Economics, Financial Institutions Francisco Silva fsilva@sas.upenn.edu (215) 421-4450 Vesa-Heikki Soini soini@sas.upenn.edu (267) 252-4408 Applied Microeconomic Theory, Public Economics, Law and Economics Macroeconomics, Financial Economics, Labor Economics Health Economics, Public Economics, Labor Economics, Household Finance Job Market Paper Does the Cream Always Rise to the Top? The Misallocation of Talent in Innovation Learning under Ambiguous Reversion Faculty Advisor, Email Ufuk Akcigit uakcigit@uchicago.edu Jeremy Greenwood Network Hazard and Bailouts Rakesh Vohra rvohra@sas.upenn.edu A Dynamic Model of Health, Education and Wealth with Credit Constraints and Rational Addiction (with James J. Heckman) Biased Learning and Permanent Reputation Kenneth Wolpin Kenneth.I.Wolpin@rice.edu The Effects of Tuition and Student Loan Policies on College Outcomes and Lifetime Earnings Public Investment and Preference Aggregation under Alternative Decentralization Institutions Macroeconomic Effects of Secondary Market Trading Hanming Fang hanming.fang@econ.upenn.edu The Optimal Design of a Criminal Justice System Andrew Postlewaite apostlew@econ.upenn.edu Academic Specialization and Misallocation of Skills in the Labor Guillermo Ordonez ordonez@econ.upenn.edu Domenico Cuoco cuoco@wharton.upenn.edu George Mailath gmailath@econ.upenn.edu Antonio Merlo amerlo@rice.edu Harold Cole coleh@sas.upenn.edu Yu Wang wangyu5@sas.upenn.edu (215) 919-1358 Labor, Public, Political Economy, Education Chunzan Wu chunzan@sas.upenn.edu (530) 219-7231 Macroeconomics, Public Finance, Labor Economics, Computational Economics YinYin Yu yinyiny@sas.upenn.edu (410) 979-6345 Education, Industrial Organizations, Applied Microeconomics Market The Impact of Student Debt on the Education, Career, and Marriage Choices of Female Lawyers More Unequal Income but Less Progressive Taxation: Economics or Politics? The Intended and Unintended Consequences of Regulating For-Profit Colleges: A Model of Enrollment and Retention Holger Sieg holgers@econ.upenn.edu Dirk Krueger dkrueger@econ.upenn.edu Hanming Fang hanming.fang@econ.upenn.edu MURAT A CELIK https://economics.sas.upenn.edu/graduate-program/candidates/murat-celik celik@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA MANOVSKI@ ECON. UPENN.EDU Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn APOSTLEW@ECON.UPENN.EDU KQUINN @ ECON. UPENN.EDU 215-898-6880 215-898-7350 215-898-5691 Office Contact Information 3718 Locust Walk, 160 McNeil Building Philadelphia, PA, 19104 Phone: +1 (215) 285-7848 Personal Information: September 26, 1987, Male, Turkish, F-1 Visa. Undergraduate Studies: Bachelor of Arts, Economics (Minor in Mathematics), Sabanci University, Valedictorian, 2010 Graduate Studies: University of Pennsylvania, 2010 to present Thesis Title: “Essays on Building Growth from Ideas” Expected Completion Date: June 2016 Thesis Committee and References: Professor Ufuk Akcigit (advisor) Department of Economics University of Chicago 1126 E 59th Street Chicago, IL, 60637 +1 (773) 702-0433 uakcigit@uchicago.edu Professor Daron Acemoglu Department of Economics Massachusetts Institute of Technology 77 Massachusetts Avenue Building E18, Room 269D Cambridge, MA, 02142 +1 (617) 253-1927 daron@mit.edu Professor Jeremy Greenwood (advisor) Department of Economics University of Pennsylvania 3718 Locust Walk, 160 McNeil Building Philadelphia, PA, 19104 +1 (215) 898-1505 recommendations@jeremygreenwood.net Teaching and Research Fields: Macroeconomics, Economic Growth Teaching Experience: Spring, 2014 Fall, 2013 Spring, 2012-2013 Fall, 2012 Fall, 2012 Fall, 2011 Money and Banking, Teaching Assistant for Prof. Harold Cole Introduction to Microeconomics, Instructor Graduate Macroeconomic Theory II, Recitation Instructor for Prof. Harold Cole and Prof. Jeremy Greenwood Economics of Family, Teaching Assistant for Prof. Jeremy Greenwood Decision Making, Teaching Assistant for Prof. Anqi Li Intermediate Macroeconomics, Recitation Instructor for Prof. Jesus Fernandez-Villaverde Professional Activities: Presentations: Macroeconomics Across Time and Space (Federal Reserve Bank of Philadelphia and NBER) (scheduled - 2015) Globalisation and Economic Policy (U of Nottingham) (2015) Annual Meeting of the Society for Economic Dynamics (2014) Bilkent University Macroeconomics Jamboree (2014) Koc University Winter Workshop (2013) Sabanci University Seminar (2013) University of Pennsylvania Macro Club (2013, 2014, 2015) Referee for: International Economic Review, Review of Economic Dynamics Honors, Fellowships and Awards: 2015-2016 SAS Dissertation Completion Fellowship (University of Pennsylvania) 2014-2015 Maloof Family Dissertation Fellowship in Economics 2010-2015 Lawrence R. Klein Fellowship 2010-2011 Pew Presidential Fellow 2010 Sakip Sabanci Award (Sabanci University Valedictorian) Publications in Journals: “Buy, Keep, or Sell: Economic Growth and the Market For Ideas” (with Ufuk Akcigit and Jeremy Greenwood, Econometrica, forthcoming) This paper develops an operational concept of technological propinquity between new ideas which are the seeds of economic growth, and the firms which transform these ideas into consumer products and processes. The empirical analysis, conducted by combining firm balance sheet data of U.S. public firms with patent grant and sale micro data from the United States Patent and Trademark Office, reveals three new stylized facts: (1) Patents that are (technologically) closer to the innovating firm contribute more to its market value. (2) Patents that are more distant are more likely to be sold to other firms. (3) When a patent is transferred between two firms, it is closer to the buyer firm than the seller firm on average. An endogenous growth model is developed in accordance with these facts, where each period firms invest in researching and developing new ideas. An idea increases a firm's productivity. By how much depends on the technological propinquity between an idea and the firm's line of business. Ideas can be bought and sold on a market for patents. A firm can sell an idea that is not relevant to its business or buy one if it fails to innovate. The developed model is matched up via indirect inference with the obtained stylized facts about the market for patents in the United States. The calibrated model is then used to gauge how efficiency in the patent market affects economic growth. Research Papers: “Does the Cream Always Rise to the Top? The Misallocation of Talent in Innovation” (Job Market Paper) The misallocation of talent between routine production versus innovation activities is found to have a first-order impact on the welfare and growth prospects of an economy. Surname level empirical analysis employing micro-data on patents and inventors in the United States between 1975-2008 combined with census data from 1930 reveals new stylized facts: (1) People with from richer backgrounds have a higher probability of becoming an inventor. (2) People from more educated backgrounds become more prolific inventors. Motivated by this discrepancy, a heterogeneous agents model with production and innovation sectors is developed, where individuals can become inventors even if they are of mediocre talent by excessive spending on credentialing. This is individually rational but socially inefficient. The model is calibrated to match the new stylized facts and data moments from the U.S. economy, and is then used to measure the magnitude of the misallocation of talent in innovation. A thought experiment in which the credentialing spending channel is shut down reveals that the aggregate growth rate of the economy can be increased by 10% of its value through a reduction of misallocation. Optimal progressive bequest taxes that alleviate the misallocation are calculated. This serves to increase social welfare by 6.20% in consumption equivalent terms. “Young, Restless, and Creative: Openness to Disruption and Creative Innovations” (with Daron Acemoglu and Ufuk Akcigit, Submitted) Openness to new, unconventional and disruptive ideas is argued to have a first-order impact on creative innovations that break new ground in terms of knowledge creation. A motivating model focusing on the choice between incremental and radical innovation is developed. This choice has implications for how managers of different ages and human capital are sorted across firms that differ in terms of openness to disruption. Empirical analysis presents firm-level, patent-level and cross-country evidence consistent with the predictions of the model. Creativity of innovations as embodied in U.S. patents is measured using different metrics such as the average number of citations per patent, the fraction of superstar inventors, the likelihood of being at the tail of the citation distribution and generality in terms of the technology classes of citing patents. Based on the idea that only companies or societies open to disruption will allow the young to rise up within the hierarchy, the main proxy chosen for this unobserved characteristic is manager age. Using the manager age proxy at the firm, patent and country levels, we present robust evidence that once the effect of the sorting of young managers to firms that are more open to disruption is factored in, the (causal) impact of manager age on creative innovations is small. Hence, it is concluded that the firms (and countries) which hire younger managers are more creative because of their unobserved openness to disruption characteristic, as opposed to the direct effect of hiring young managers instead of older ones. “Patents as Collateral and Directed Technological Change” (Research in progress with Ufuk Akcigit, Olga Itenberg and Guillermo Ordonez) In recent years, and in spite of their intangible nature, patents have been increasingly used as collateral. We show that this financial innovation has disproportionately drawn firm entry into more crowded innovative industries, those where patents are easier to trade. We study the effects of the use of patents as collateral in a multi-sector endogenous growth framework with expanding input varieties, where the intermediate sectors differ in market size. Our model predicts that the use of patents as collateral directs future technological progress towards sectors with more patenting firms, which are not necessarily the most productive sectors. Even though the use of patents as collateral relaxes financial frictions, they also have the potential to direct technological innovation inefficiently, which we explore quantitatively. Languages: English (fluent), German (advanced), Turkish (native). Computational Skills: Matlab, Fortran, C++, Stata. Dissertation Abstract Murat A. Celik – University of Pennsylvania, Department of Economics In my dissertation, I study the interplay between innovation and economic growth with an emphasis on the misallocation of ideas and talent. The first chapter analyzes the link between the misallocation of talent and innovation, asking the question whether the talented people in the society are given the chance to create new innovations that enhance social welfare. The second chapter focuses on the allocation of the innovations themselves, considering the optimal allocation of ideas across firms, and how it can be achieved via a market for ideas. The last chapter investigates why some organizations and societies are more prolific than others in coming up with radical, path-breaking innovations, and marshals evidence showing that the cause might be the “openness to disruption” in their culture. Taken together, I document an array of new empirical facts from micro-data on innovation, propose new ways to better understand the link between innovation and economic growth, and offer government policies that can help in increasing long-run growth and social welfare. Chapter 1: Does the Cream Always Rise to the Top? The Misallocation of Talent and Innovation (Job Market Paper) The misallocation of talent between routine production versus innovation activities in the United States is investigated, and found to have a first-order impact on the welfare and growth prospects of the economy. Empirical analysis combines micro data on patents granted in the United States and their inventors from 19752008 with U.S. census data from 1930 by using surnames as a linking proxy. The analysis reveals two new stylized facts: (1) People from richer backgrounds have a higher probability of becoming an inventor; but this is not the case for those from more educated backgrounds. (2) People from more educated backgrounds become much more prolific inventors; but those from richer backgrounds exhibit no such aptitude. Motivated by this discrepancy, a new model which can accommodate the observed correlation patterns is developed. Firms undertake routine production using unskilled labor, and generate productivity-improving innovations (featuring positive intertemporal spillovers between firms) via research and development conducted by hired inventors. The households are heterogeneous in wealth, education, and unobserved innate ability that is persistent across generations. Parents invest in the education of their offspring and leave bequests. The training necessary to become inventors is scarce; hence individuals compete against each other in a tournament setting to receive it. Factors that improve inventor productivity such as innate ability and education increase the probability of receiving this training; but so does private credentialing spending which is unproductive by itself. Thus, individuals who inherit generous bequests can become inventors even if they are of mediocre talent through excessive spending on credentialing, preventing more talented individuals from poorer backgrounds from becoming one. This is individually rational but socially inefficient; reducing the quality of the inventor pool used in generating productivity-improving innovations that drive economic growth. The model is calibrated to match the new stylized facts and data moments from the U.S. economy where an exercise in indirect inference pins down the influence of the new credentialing spending channel by replicating the two regressions from the empirical analysis using model-generated data. The calibrated model is then used to measure the economic importance of the misallocation of talent in innovation. A thought experiment in which the credentialing spending channel is shut down reveals that the aggregate growth rate of the economy can be increased by 10% of its value by assigning more talented and better educated individuals as inventors, while improving welfare by 5.96%. Seeking to alleviate the effects of misallocation in a decentralized economy, optimal progressive bequest taxes are calculated, which are found to increase social welfare by 6.20% in consumption equivalent terms. Chapter 2: Buy, Keep or Sell: Economic Growth and the Market for Ideas (with Ufuk Akcigit and Jeremy Greenwood) – Econometrica, forthcoming This paper develops an operational concept of technological propinquity between new ideas which are the seeds of economic growth, and the firms which transform these ideas into consumer products and processes. The empirical analysis, conducted by combining firm balance sheet data of U.S. public firms with patent grant and sale micro data from the United States Patent and Trademark Office, reveals three new stylized facts: (1) Patents that are (technologically) closer to the innovating firm contribute more to its market value. (2) Patents that are more distant are more likely to be sold to other firms. (3) When a patent is transferred between two firms, it is closer to the buyer firm than the seller firm on average. An endogenous growth model is developed in accordance with these facts, where each period firms invest in researching and developing new ideas. An idea increases a firm's productivity. By how much depends on the technological propinquity between an idea and the firm's line of business. Ideas can be bought and sold on a market for patents. A firm can sell an idea that is not relevant to its business or buy one if it fails to innovate. The developed model is matched up via indirect inference with the obtained stylized facts about the market for patents in the United States. The calibrated model is then used to gauge how efficiency in the patent market affects economic growth. Chapter 3: Young, Restless and Creative: Openness to Disruption and Creative Innovations (with Daron Acemoglu and Ufuk Akcigit) – Submitted Openness to new, unconventional and disruptive ideas is argued to have a first-order impact on creative innovations that break new ground in terms of knowledge creation. A motivating model focusing on the choice between incremental and radical innovation is developed. This choice has implications for how managers of different ages and human capital are sorted across firms that differ in terms of openness to disruption. Empirical analysis presents firm-level, patent-level and cross-country evidence consistent with the predictions of the model. Creativity of innovations as embodied in U.S. patents is measured using different metrics such as the average number of citations per patent, the fraction of superstar inventors, the likelihood of being at the tail of the citation distribution and generality in terms of the technology classes of citing patents. Based on the idea that only companies or societies open to disruption will allow the young to rise up within the hierarchy, the main proxy chosen for this unobserved characteristic is manager age. Using the manager age proxy at the firm, patent and country levels, we present robust evidence that once the effect of the sorting of young managers to firms that are more open to disruption is factored in, the (causal) impact of manager age on creative innovations is small. Hence, it is concluded that the firms (and countries) which hire younger managers are more creative because of their unobserved openness to disruption characteristic, as opposed to the direct effect of hiring young managers instead of older ones. HONGSEOK CHOI aitch.choi@gmail.com +82 10 5115 2037 https://economics.sas.upenn.edu/graduate-program/candidates/hongseok-choi UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN@ECON.UPENN.EDU (215) 898-6880 (215) 898-7350 (215) 898-5691 Office Contact Information 432 Danjae Building Republic of Korea Air Force Academy 635 Danjae-ro, Sangdang-gu Cheongju, 28187, Republic of Korea +82 43 290 6516 Personal Information Citizen of the Republic of Korea Undergraduate Studies B.Sc. in Physics and B.A. in Economics, Seoul National University, Summa cum Laude, 2002-2006 Graduate Studies Ph.D. in Economics, University of Pennsylvania, 2006-2012 Dissertation Title: “Essays on Learning under Ambiguity” Dissertation Committee and References Professor Domenico Cuoco Department of Finance University of Pennsylvania 3620 Locust Walk Philadelphia, PA 19104 cuoco@wharton.upenn.edu (215) 898-8290 Professor Philipp Karl Illeditsch Department of Finance University of Pennsylvania 3620 Locust Walk Philadelphia, PA 19104 pille@wharton.upenn.edu (215) 898-3477 Professor Dirk Krueger Department of Economics University of Pennsylvania 3718 Locust Walk Philadelphia, PA 19104 dkrueger@econ.upenn.edu (215) 573-1424 Professor Andrew Postlewaite Department of Economics University of Pennsylvania 3718 Locust Walk Philadelphia, PA 19104 apostlew@econ.upenn.edu (215) 898-7350 Teaching and Research Fields Primary Fields: Asset Pricing, Portfolio Choice, Behavioral Finance (Ambiguity) Secondary Field: Microeconomic Theory Research Experience and Other Employment Feb. 2013 to Nov. 2015 Assistant Professor of Economics, Republic of Korea Air Force Academy1 Summer 2008 Research Assistant for Professor Larry Selden at Columbia Business School Teaching Experience Republic of Korea Air Force Academy, 2013-2015 Introduction to Economics (Spring 2013 to Spring 2015) Microeconomics (Fall 2013 and Fall 2015) Macroeconomics (Spring 2013 and Spring 2014) Game Theory (Spring 2014 and Spring 2015) Public Economics (Fall 2014 and Fall 2015) University of Pennsylvania, 2007-2010 Spring 2010 Statistics for Economists, Recitation Instructor for Professor XunTang Fall 2009 Statistics for Economists, R.I. for Professor Suleyman Ozmucur Spring 2009 Introduction to Econometrics, R.I. for Professor Kyungchul (Kevin) Song Fall 2008 Introduction to Microeconomics, R.I. for Professor Uriel Spiegel Spring 2008 International Economics, Teaching Assistant for Professor Wilfred Ethier Fall 2007 Introduction to Economics for Business, R.I. for Professor Rebecca Stein and Professor Gwen Eudey Presentations 2014 2015 Seoul National University Korea University Finance Department, 11th Annual Conference of the Asia-Pacific Association of Derivatives, Korea University Mathematics Department, Korea Institute of Finance, Yonsei University Honors and Scholarships 2003-2006 Scholarship, Korea Foundation for Advanced Studies 2000 Bronze Medal, Republic of Korea National Physics Olympiad Programming Languages Mathematica, MATLAB, Fortran 1 As part of compulsory military service (September 2012 to November 2015, 39 months). Job Market Paper “Learning under Ambiguous Reversion” The question of whether returns and growth rates are predictable is yet to be resolved. This paper considers learning under such ambiguous circumstances, or, more specifically, when mean reversion in the state variables is plausible but not certain. The basic idea is that the agent as a result entertains multiple theories close to those of mean reversion and uses only sufficiently plausible ones in predicting the outcomes of the immediate future. A novel measure of market uncertainty is identified and its dynamics explicitly characterized up to a system of differential equations that generalizes the KalmanBucy filter in the presence of model uncertainty. Asset pricing applications show that (i) learning under ambiguity induces a declining trend in the equity premium, (ii) an improvement in the quality of public information can increase the equity premium, and (iii) the component premium for bearing ambiguity exhibits hysteresis with respect to and furthermore can comove negatively with the conditional variance of returns, thereby obfuscating the relationship of the latter with the total equity premium. Research Paper in Progress “Ambiguous Returns, Learning, and Portfolio Choice” This paper considers the portfolio choice of a log investor who learns under model uncertainty. Specifically, the investor entertains a “neighborhood” of her benchmark theory that expected returns are mean-reverting and uses the candidate theories that are sufficiently plausible vis-à-vis observed returns, in predicting future returns. After calibrating the model to U.S. stock market data (Barberis, 2000), I numerically compute the optimal Markov policy and observe that ignoring learning under ambiguity can result in a significant underestimation of hedging demand. Most notably, even with a moderate degree of ambiguity (1% point in annual terms) in the equity premium, the unconfident log investor sells short an amount of the risky asset worth half of her wealth when the point estimate of the equity premium is zero. I also observe that the hedging demand reflects the agent’s desire to eliminate the effect of misspecification on continuation utility, and that the related model of learning under ambiguity by Miao (2009) is a limiting case of the present model. Learning under Ambiguity and Its Financial Implications: When Mean-Reversion in the Latent Variables Is Plausible but Not Certain Hongseok Choi University of Pennsylvania Learning under Ambiguous Reversion (Job Market Paper) In the more realistic class of economic models, agents only partially observe the economy—variables of interest such as growth rates are noisy observations of some “state variables.” In these models, how agents estimate the state of the economy, what the risk involved in the estimation is, and how the estimate and the estimation risk evolve over time, are central questions leading to distinct economic implications. However, these models are still demanding too much knowledge and confidence from agents. Consider, for example, the growth of dividends. While some authors find it unpredictable (Cochrane, 2008), others reject the null hypothesis of random walk (Binsbergen and Koijen, 2010); and accordingly, expected dividend growth rates have been assumed at some instances to be constant and at others to be time-varying following an AR(1), a regime-switching, or some other process. This divergence of opinion among their modelers notwithstanding, agents in economic models behave as if they knew the governing dynamics of the economy. In this context, this paper relaxes the assumption of incomplete information a step further to model uncertainty and investigates how agents estimate the generating process of the state of the economy, what the ambiguity involved in the estimation is, how the estimate and the estimation ambiguity evolve over time, and ultimately, how this learning under ambiguity affects asset returns. A key aspect of this work is that it incorporates an intuitive and familiar way to learn under ambiguity into an axiomatically founded framework of preferences, recursive multiple-priors utility (Chen and Epstein, 2002). As its name suggests, recursive multiple-priors utility differs from the traditional recursive utility in that in the recursive representation, the expectation against a unique one-step-ahead conditional is replaced by the minimum of expectations over a set of one-step-head conditionals. The size of the last set, in particular, is interpreted to represent that of the conditional ambiguity associated with one-step-ahead uncertainties. Importantly, the theory of recursive multiplepriors utility is silent as to where the one-step-ahead conditionals come from, just as Savage’s is regarding the unique prior, and it is thus up to the modeler to specify them. In this paper, then, I propose that they come from multiple a priori theories and successive hypothesis tests. The main result of the paper is an explicit characterization of the dynamics of learning up to a system of differential equations that generalizes the classical filtering equations of a conditionally Gaussian process in the presence of ambiguity (the Kalman-Bucy filter refers to the filtering equations of an ordinary Gaussian process, which is trivially conditionally Gaussian). The new filter tracks, among others, the evolution of conditional ambiguity; and these endogenous dynamics of conditional ambiguity are the novelty of this paper. There are various ways to model decision makers under ambiguity, and one of the most popular is Gilboa and Schmeidler’s (1989) multiple-priors utility; recursive multiple-priors utility refers to its dynamic version. Multiple-priors models, static or dynamic, have been successful in explaining puzzling phenomena in financial markets such as stock market nonparticipation, excess volatility, and excess equity premium. Most applications of recursive multiple-priors utility, however, have disregarded learning and specified the time variation, or a lack thereof, in ambiguity exogenously. The one paper in the existing literature, to the best of my knowledge, that derives time variation in the set of one-step-ahead conditionals from a model of learning is Epstein and Schneider (2007). In that paper, too, the predictive set is constructed by a statistical test over multiple theories. The main differences are twofold. First, whereas Epstein and Schneider consider memoryless data-generating 1 mechanisms, I consider those with serial dependence. The latter consideration clearly complements the former; for example, with the present model, we can study the effects of learning under ambiguity when growth rates are (ambiguously) predictable. Second, whereas Epstein and Schneider set their model in discrete time, I set mine in continuous time; and continuous-time modeling is known to facilitate analysis with its analytical tractability. But I also note that the continuous-time counterpart of Epstein and Schneider’s portfolio choice example results in no learning because the likelihood function degenerates to infinity everywhere; and consequently, their discrete-time finding that learning resolves ambiguity does not immediately carry over to continuous time. To investigate its asset pricing implications, I apply the model of learning to simple Lucas economies populated by log agents. In these economies, agents suspect that dividend growth rates are predictable, containing a persistent component, but perceive a degree of ambiguity in the law of motion for the component. The resulting equilibrium equity premium consists of the risk premium, namely the conditional variance of growth rates, and an ambiguity premium proportional to the conditional ambiguity in the expected growth rate. The first observation we can make from the decomposition of the equity premium is that the relationship between the equity premium and the conditional variance of returns is unclear. Motivated by standard asset pricing models such as the I-CAPM of Merton (1973), numerous empirical papers have investigated the relationship between the two, but the findings are mixed. In the present model, the equity premium is given by the conditional variance of returns plus the ambiguity premium, and the latter (market ambiguity) exhibits hysteresis with respect to, and furthermore can comove negatively with, the former (market risk). In the long run, the equity premium exhibits a downward trend. This is because the ambiguity premium does: it is a reward for bearing ambiguity in the expected growth rate, and the ambiguity is to exhibit a downward trend, because the component ambiguity in the time-invariant factors of the data-generating mechanism turns out to resolve completely as time goes to infinity. The equity premium has indeed been observed to have declined over the postwar period, and part of it could be due to learning and the accompanying resolution of ambiguity. I also demonstrate that an improvement in the quality of public information can increase the asymptotic level of the equity premium. In a Bayesian framework, Veronesi (2000) made a similar observation that higher precision of signals tends to increase the risk premium. What distinguishes the present observation from Veronesi’s is that his result relies on agents’ being sufficiently risk-averse, more so than log agents; in his model, lower precision of signals decreases the equity premium because the agents’ hedging demand tones down the covariation between returns and consumption growth. In contrast, the present paper shows that the equity premium can exhibit such counterintuitive behavior even under unit risk aversion, which is conventionally associated with myopia. Ambiguous Returns, Learning, and Portfolio Choice This paper applies the model of learning developed in the foregoing paper to the portfolio choice problem of a log investor. Specifically, the investor entertains a “neighborhood” of her benchmark theory that expected returns are meanreverting and uses the candidate theories that are sufficiently plausible vis-à-vis observed returns, in predicting future returns. After calibrating the model to U.S. stock market data (Barberis, 2000), I numerically compute the optimal Markov policy and observe that ignoring learning under ambiguity can result in a significant underestimation of hedging demand. Most notably, even with a moderate degree of ambiguity (1% point in annual terms) in the equity premium, the unconfident log investor sells short an amount of the risky asset worth half of her wealth when the point estimate of the equity premium is zero. I also observe that the hedging demand reflects the agent’s desire to eliminate the effect of misspecification on continuation utility, and that the related model of learning under ambiguity by Miao (2009) is a limiting case of the present model. 2 SELMAN EROL http://erols.me/ erols@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN@ECON.UPENN.EDU +1-215-898-6880 +1-215-898-7350 +1-215-898-5691 Office Contact Information 3718 Locust Walk, 160 McNeil Building Philadelphia, PA, 19104 +1-215-667-5036 Undergraduate Studies: B.Sc., Mathematics, Bilkent University, 2007 Masters Level Work: M.A., Economics, Bilkent University, 2009 Graduate Studies: University of Pennsylvania Thesis Title: “Network Formation and Its Impact on Systemic Risk” Expected Completion Date: June 2016 Thesis Committee and References: Professor Rakesh Vohra (Advisor) 3718 Locust Walk, 451 McNeil Building Philadelphia, PA, 19104 rvohra@seas.upenn.edu +1-215-898-6777 Professor Andrew Postlewaite 3718 Locust Walk, 458 McNeil Building Philadelphia, PA, 19104 apostlew@econ.upenn.edu +1-215-898-7350 Professor Guillermo Ordonez 3718 Locust Walk, 428 McNeil Bldg. Philadelphia, PA, 19104 ordonez@econ.upenn.edu +1-215-898-1875 Professor Camilo Garcia-Jimeno 3718 Locust Walk, 528 McNeil Bldg. Philadelphia, PA, 19104 gcamilo@sas.upenn.edu +1-215-898-8206 Professor Steven Matthews (Teaching reference) 3718 Locust Walk, 521 McNeil Building Philadelphia, PA, 19104 stevenma@econ.upenn.edu +1-215-898-7749 Teaching and Research Fields: Network Formation, Economic Theory Teaching Experience: University of Pennsylvania, Economics Department Spring, 2015 212, Game Theory, T.A. for Prof. Yuichi Yamamoto Spring, 2014 212, Game Theory, T.A. for Prof. Yuichi Yamamoto Fall, 2013 212, Game Theory, T.A. for Prof. Steven Matthews Fall, 2013 234, Law and Economics, T.A. for Prof. Camilo Garcia-Jimeno Summer, 2012 897, (Graduate) Summer Math Camp, Instructor Spring, 2012 702, (Graduate) Macroeconomic Theory, T.A. for Prof. Dirk Krueger, Prof. Jesus Fernandez-Villaverde Fall, 2011 681, (Graduate) Microeconomic Theory, T.A., for Prof. Mallesh M. Pai Spring, 2011 702, (Graduate) Macroeconomic Theory, T.A. for Prof. Dirk Krueger Fall, 2010 102, Intermediate Macroeconomics, T.A. for Prof. Jesus Fernandez-Villaverde Bilkent University, Economics Department and Mathematics Department (Graduate) Microeconomics, T.A. for Prof. Tarik Kara (Graduate) Math Review Course, T.A. for Prof. H. Cagri Saglam Math for Econ T.A. for Prof. H. Cagri Saglam Discrete Math (R.I.), Calculus (R.I.), Advanced Calculus (R.I.) The Scientific and Academic Research Council of Turkey Instructor of Geometry, Instructor of Discrete Mathematics Research Experience and Other Employment: Fall, 2014 Fall, 2010 University of Pennsylvania, R.A. for Prof. Rakesh Vohra “Growth Through Heterogeneous Innovations”, N.B.E.R. Working Paper #16443, R.A. for Prof. Ufuk Akcigit and Prof. William R. Kerr Professional Activities Service May, 2015 2014, 2015 Organizing committee: Networks and Systemic Risk Workshop, The Warren Center, University of Pennsylvania Referee: SAGT 2014, Games and Economic Behavior Selected presentations Dec, 2015 (Scheduled) Philadelphia FED Nov, 2015 (Scheduled) Carnegie-Mellon University, Tepper School of Business, Seminar Oct, 2015 Pennsylvania State University, Midwest Economics Theory Meetings Aug, 2015 11th World Congress of the Econometric Society, Montreal Aug, 2015 3rd Summer School of the Econometric Society, University of Tokyo July, 2015 Conference on Economic Design, Society for Economic Design, Bilgi University May, 2015 Information Transmission in Networks Conference, NSF, Harvard University Dec, 2014 NYCE, Microsoft Research, New York (Poster) Awards: 2010 Lawrence Robbins Prize (best first year performance in Economics Ph.D. program, University of Pennsylvania, shared with Zehao Hu) 2003 Silver Medal, 44th International Math Olympiads, Japan Publications: “Negative Certainty Independence without Betweenness”, Economics Letters, 2013 (with David Dillenberger) “Existence, Optimality and Dynamics of Equilibria with Endogenous Time Preference”, Journal of Mathematical Economics, 2011 (with Cuong LeVan and Cagri Saglam) Research Papers: “Network Hazard and Bailouts” (Job Market Paper) I introduce a model of contagion with endogenous network formation and strategic default, in which a government intervenes to stop contagion. The anticipation of government bailouts introduces a novel channel for moral hazard via its effect on network architecture. In the absence of bailouts, the network formed consists of small clusters that are sparsely connected. When bailouts are anticipated, firms in my model do not make riskier individual choices. Instead, they form networks that are more interconnected, exhibiting a core-periphery structure (wherein many firms are connected to a smaller number of central firms). Interconnectedness within the periphery increases spillovers. Core firms serve as a buffer when solvent and an amplifier when insolvent. Thus, in my model, ex-post time-consistent intervention by the government improves ex-ante welfare but it increases systemic risk and volatility through its effect on network formation. This paper can be seen as a first attempt at introducing a theory of mechanism design with endogenous network externalities. “Network Formation and Systemic Risk” (with Rakesh Vohra) This paper introduces a model of endogenous network formation and systemic risk in OTC markets. A link is a trading opportunity that yields benefits only if the counterparty does not subsequently default. After links are formed, they are subjected to exogenous shocks that are either good or bad. Bad shocks reduce returns from links and incentivize default. Good shocks the reverse. Defaults triggered by bad shocks might propagate via links. The model yields three insights. First, higher probability of good shocks generates higher probability of system wide default. Increased interconnectedness in the network offsets the effect of better fundamentals. Second, the network formed critically depends on the correlation of shocks to the system. As a consequence, an outside observer who is mistaken only about the correlation structure of shocks, upon observing a highly interconnected network, will underestimate the probability of system wide default. Third, when the risk of contagion is high, the networks formed in the model are utilitarian efficient. “A Network Theory of Civil Liberties and Social Structure” (with Camilo Garcia-Jimeno) An important dimension of civil liberties is related to the state’s ability to exercise coercion -detentions, searches, seizures, torture- to collect information regarding individuals. Because the density and depth of social ties shape how information is distributed across society, the extent of civil liberties can determine both how coercion is exercised and the social structure itself. We propose a model in which individuals are better informed about socially closer individuals, whom they care more about. A government agent searches for a threat by arresting individuals and demanding information about members of the threat employing coercion. We characterize the optimal search protocol, which features targeting for longest “paths” of social acquaintance. We then endogenize the network structure to study how the social fabric is affected by the extent of civil liberties via the search protocol. Under low civil liberties, society is segregated and peripheral individuals are targeted. For intermediate civil liberties, society gradually becomes integrated and central nodes are targeted. Nevertheless, the relation between civil liberties and social integration exhibits discontinuities. We also illustrate how the framework can be used to explore questions of organizational design such as corruption fighting. “Dominant Strategy Implementation for Interdependent Ordinal Preferences” (with Talat Senocak) “A Simple Proof of Mueller-Satterthwaite Theorem”, mimeo. at Bilkent University (with Ali Adali, Semih Koray, Nizameddin Ordulu) Last updated: November 11, 2015 Selman Erol (University of Pennsylvania) Dissertation Abstract “Network Formation and Its Impact on Systemic Risk” 1 In the aftermath of the financial crisis of 2008, many policy makers and researchers pointed to the interconnectedness of the financial system as one of the fundamental contributors to systemic risk. The argument is that the linkages between financial institutions served as an amplification mechanism: shocks to smaller parts of the system propagate through the system and result in broad damage to the financial system. In my dissertation, I explore the formation of networks when agents take into account systemic risk. The dissertation consists of three complementary papers on this topic. The first paper titled “Network Formation and Systemic Risk”, joint with Professor Rakesh Vohra. We set out the framework and construct a model of endogenous network formation and systemic risk. We find that fundamentally ‘safer’ economies with higher probability of getting good shocks generate higher interconnectedness, which leads to higher systemic risk. This provides network foundations for “the volatility paradox” arguing that better fundamentals lead to worse outcomes due to excessive risk taking. Second, the network formed crucially depends on the correlation of shocks to the system. As a consequence, an observer, such as a regulator, facing an interconnected network who is mistaken about the correlation structure of shocks will underestimate the probability of system wide failure. This result relates to the “dominoes vs. popcorn” discussion by Edward Lazear. He comments that a fundamental mistake in addressing the crisis was to think that it was “dominoes” so that saving one firm would save many others in the line. He continues to argue that it was “popcorn in a pan”: all firms were exposed to same correlated risks and saving one would not save many others. We complement his discussion by arguing that the same mistake might have been the reason behind why sufficient regulatory precaution was not taken prior to the crisis. The third result is that the networks formed in the model are utilitarian efficient because the risk of contagion is high. This causes firms to minimize contagion by forming dense but isolated clusters that serve as firebreaks. This finding is suggestive that, the worse the contagion, the more the market takes care of it. In the second paper, titled “Network Hazard and Bailouts”, which is my job market paper, I ask how the anticipation of ex-post government bailouts affects network formation. I deploy a significant generalization of the model in the first paper and allow for time-consistent government transfers. I find that the presence of government bailouts introduces a novel channel for moral hazard via its effect on network architecture, which I call “network hazard”. In the absence of bailouts, firms form sparsely connected small clusters in order to eliminate second-order counterparty risk: expected losses due to defaulting counterparties that default because of their own defaulting counterparties. Bailouts, however, eliminate second-order counterparty risk already. Accordingly, when bailouts are anticipated, the networks formed become more interconnected, and exhibit a core-periphery structure (many firms connected to a smaller number of central firms, which is observed in practice). Interconnectedness within the periphery leads to higher extent of contagion with respect to the networks formed in the absence of intervention. Moreover, solvent core firms serve as a buffer against contagion by increasing the resilience of the many peripheral firms that are connected to the core. However, insolvent core firms serve as an amplifier of contagion since they make peripheral firms less resilient. This implies that in my model, ex-post time-consistent intervention by the government, while ex-ante welfare improving, increases systemic risk and volatility, solely through its effect on the network. A remark is that firms, in my model, do not make riskier individual choices regarding neither their choice of investment risk, nor the number of their counterparties they have. In this sense, network hazard is a genuine form of moral hazard solely through the formation of the detailed network. On another note, the model can also be viewed as a first attempt towards developing a theory of mechanism design with endogenously formed network externalities which might be useful in various other scenarios such as provision of local public goods. In the final paper, titled “Regulation of Systemic Risk under Network Formation”, I consider the role of liquidity and capital requirements to alleviate network hazard and systemic risk. In the model, financial firms set up credit lines with each other in order to meet their funding needs on demand. Accordingly, higher liquidity requirements induce firms to form higher interconnectedness in order to be able to find deposits as needed. At Selman Erol (University of Pennsylvania) Dissertation Abstract “Network Formation and Its Impact on Systemic Risk” 2 a tipping point of liquidity requirements, the network discontinuously jumps in its interconnectedness, which contributes discontinuously to systemic risk. On the other hand, the reaction to capital requirements is smooth. Capital requirements indirectly work as an upper bound in the interconnectedness firms would form. This way, interconnectedness can be effectively reduced to a desired level via capital requirements. Yet capital requirements cannot be used to induce higher interconnectedness. Thusly, in times of credit freeze, capital requirements may not help promote circulation of credit. A conjunction of both liquidity and capital requirements is more effective in promoting desired circulation while reducing systemic risk. The work in this dissertation suggests that endogenous network architecture is an essential component of the study of financial markets. In particular, network hazard is a genuine form of moral hazard that will be overlooked unless network formation is taken into account, while it has implications about systemic risk. Moreover, this work illustrates how the reaction of networked financial markets to both fundamentals of the economy and to the policy can be non-trivial, featuring non-monotonicity and discontinuity. Keywords: network formation, contagion, counterparty risk, systemic risk, bailouts, network hazard, moral hazard, regulation, capital requirements, reserve requirements, phase transition, mechanism design RONG HAI https://sites.google.com/site/ronghaiecon/ ronghai@uchicago.edu UNIVERSITY OF PENNSYLVANIA MANOVSKI@ECON.UPENN.EDU Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn APOSTLEW@ECON.UPENN.EDU KQUINN@ECON.UPENN.EDU Office Contact Information 5740 S. Woodlawn Avenue, Room 172 Chicago, IL, 60637 Mobile: (+1) 267-254-8866 Personal Information: Sex: Female 215-898-6880 215-898-7350 215-898-5691 Home Contact Information 5514 S. Blackstone Ave., Apt H Chicago, IL, 60637 Citizenship: China US Visa Status: H1B. Academic Position: Postdoctoral Scholar, 05/2013-Present Becker Friedman Institute for Research in Economics and The Center for the Economics of Human Development, University of Chicago Research Fields: Health Economics, Public Finance, Labor Economics, Household Finance Education: Ph.D., Economics, University of Pennsylvania, 2013 M.A., Economics, Peking University, 2007 B.A., Economics, Fudan University, 2004 References: Professor James J. Heckman Department of Economics University of Chicago Professor Dirk Krueger Department of Economics University of Pennsylvania 1126 East 59th Street Chicago, Illinois 60637 (773) 702-0634 jheckman@uchicago.edu 3718 Locust Walk, Philadelphia, PA 19104 (215) 573-1424 dkrueger@econ.upenn.edu Professor Andrew Postlewaite Professor Petra Todd Department of Economics University of Pennsylvania Department of Economics University of Pennsylvania 3718 Locust Walk, Philadelphia, PA 19104 (215) 898-7350 apostlew@econ.upenn.edu Professor Kenneth Wolpin Department of Economics Rice University P.O. Box 1892, Houston, TX (713) 348-2522 Kenneth.I.Wolpin@rice.edu 3718 Locust Walk, Philadelphia, PA 19104 (215) 898-4084 ptodd@econ.upenn.edu Teaching Fields: Econometrics, Public Finance, Health Economics, Labor Economics Teaching Experience: Summer 2010 Econometrics, Instructor Spring 2012 Public Finance, Teaching Assistant for Dr. Francesca Acacia Fall 2011 Introductory Economics for Business Students, Recitation Instructor for Dr. Uriel Spiegel Spring 2011 International Finance, Teaching Assistant for Prof. Wilfred Ethier Fall 2010 Labor Economics, Teaching Assistant for Prof. Kenneth Burdett Fall 2009-10 Advanced Microeconometrics, Teaching Assistant for Prof. Petra Todd Spring 2009-10 Time Series Analysis (Ph.D.), Recitation Instructor for Prof. Francis Diebold Fall 2008 Statistics for Economists, Review Session Instructor for Prof. Aureo de Paula, and Dr. Suleyman Ozmucur Referee: Journal of Political Economy; Review of Economic Studies; International Economic Review; European Economic Review; LABOUR: Review of Labour Economics and Industrial Relations Presentations: 2015 2014 2013 2012 University of Chicago Economics; University of Chicago Harris; University of Miami Economics; University of Miami Finance; Chinese University of Hong Kong; Florida International University; Peking University; Annual Conference of the Royal Economic Society (x2); Fourth SOLE/EALE World Meeting; 11th World Congress of Econometric Society (x2); National Tax Association's Annual Meeting University of Cornell; University of Miami; Annual Health Econometrics Workshop at UToronto; Cowles Summer Conference: Structural Microeconomics at Yale; Conference on the Affordable Care Act and the Labor Market at Chicago Fed; Fifth Biennial Meeting of the American Society of Health Economists at USC; Launch Event of the Center for the Economics of Human Development; Southern Economic Association Annual Meeting Ohio State University; Syracuse University; University of Chicago; University of South Florida; North American Summer Meetings of the Econometric Society; Southern Economic Association Annual Meeting Federal Reserve Bank of Chicago; University of Pennsylvania Other Professional Activities: 2015 NBER Summer Institute Labor Studies, Health Economics, Health Care, Aging; NBER Summer Institute Macro Public Finance (2014-2015); NBER Summer Institute Economic Fluctuations and Growth Working Group on Aggregate Implications of Microeconomic Consumption Behavior 2014 Fifth Biennial Meeting of the American Society of Health Economists (Discussant); Southern Economic Association Annual Meeting (Discussant); Southern Economic Association Annual Meeting (Session Chair); NBER Summer Institute Macro Public Finance 2009-2012 Coordinator at Graduate Student Housing Community Services at University of Pennsylvania 2009-2010 Co-organizer of Econometrics Lunch at University of Pennsylvania Research Experience and Other Employment: Summer 2012 Federal Reserve Bank of Chicago, Summer Fellow Winter 2008 The World Bank, Office of VP for Chief Economist, Short Term Consultant Honors, Scholarships, and Fellowships: 2012 CSWEP Dissertation Fellowship, Federal Reserve Bank of Chicago 2007 - 12 International Economic Review Fellowship, University of Pennsylvania 2011 - 12 Xinmei Zhang Fellowship, University of Pennsylvania 20010 Price Theory Summer Camp, University of Chicago Booth School of Business 2009 - 10 Judith Rodin Fellowship, University of Pennsylvania Research Papers: “A Dynamic Model of Health, Education and Wealth with Credit Constraints and Rational Addiction”, with James J. Heckman (University of Chicago) (Job Market Paper) This paper develops and structurally estimates a life-cycle model where health, education, and wealth are endogenous accumulated processes depending on the history of an individual's optimal behaviors, on parental factors, and on cognitive and noncognitive abilities. The model investigates many different pathways between education, health, and wealth by introducing endogenous health capital production and addictive preferences of unhealthy behavior in the presence of credit constraints. Using data from National Longitudinal Survey of Youth 97, we estimate the model using a two-step estimation procedure based on factor analysis and simulated method of moments. The estimated model decomposes the causal effects of education on health into the direct benefits of improving health production efficiency and the indirect benefits of reducing unhealthy behavior and raising earnings. We show that rational addiction has important quantitative implication on predicted patterns of unhealthy behavior by socioeconomic status and over the life-cycle. We find sizable impacts of both health and parental transfers on individuals' college decisions. We find that relaxing credit constraints improves education while its effects on healthy behavior and health evolves nonlinearly over the life cycle. Finally, we show that both cognitive and noncognitive factors are important determinants of health, education, and wealth. “The Determinants of Rising Inequality in Health Insurance and Wages”, July 2015 Under Review Over the last 30 years in the U.S., less educated workers have experienced a sharp decline in health insurance coverage rate and stagnant wage growth. In contrast, more educated workers’ health insurance coverage rate has stayed relatively stable and their wages have rapidly grown. This paper investigates the determinants of the increase in inequality in health insurance coverage and wages by estimating an overlapping generations equilibrium model of labor and health insurance markets’ demand and supply. The estimated model is used to quantify the effects of changes in aggregate factors (including rising cost of medical care services, Medicaid eligibility expansion, skill-biased technological changes in the labor market, and changes in the labor force composition) on the inequality of health insurance coverage and wages. I find that the interaction between the rising cost of medical services and labor market technological change is the most important determinant of the widening gap of health insurance coverage. “On the Welfare Cost of Consumption Fluctuations in the Presence of Memorable Goods”, with Dirk Krueger (University of Pennsylvania) and Andrew Postlewaite (University of Pennsylvania), Jan 2015 We propose a new category of consumption goods, memorable goods, that generate a flow of utility after consumption. We analyze an otherwise standard consumption model that distinguishes memorable goods from other nondurable goods. Consumers optimally choose lumpy consumption of memorable goods. We empirically document differences between levels and volatilities of memorable and other goods expenditures. Memorable goods expenditures are about twice durable goods expenditures and half the volatility. The welfare cost of consumption fluctuations driven by income shocks is overstated if memorable goods are not accounted for and estimates of excess sensitivity of consumption might be due to memorable goods. Income Inequality, Tax Policy, and Economic Growth, with Siddhartha Biswas (UNC) and Indraneel Chakraborty (University of Miami), July 2015 Under Review We investigate how reduction of income inequality through tax policy affects economic growth. Taxation at different points of the income distribution has heterogeneous impacts on households’ incentives to invest, work, and consume. Using U.S. state-level data and micro-level household tax returns over the last three decades, we find that reduction of income inequality between low and median income households improves economic growth. However, reduction of income inequality through taxation between median and high income households reduces economic growth. These economic growth effects are attributable both to supply-side factors (changes in small business activity and labor supply) and to consumption demand. “The Impact of an Early Career Recession on Health and Health-Related Behaviors”, with Naijia Guo (Chines University of Hong Kong), June 2015 Under Review Using National Longitudinal Survey of Youth 1997, we estimate the long-term impact of an early career recession on various health outcomes and health-related behaviors. The early career recession is measured by the state-specific unemployment rate in the year of labor market entry. Using an instrumental variable approach, we address the potential selection issues on the timing of labor market entry. We find that first, an early career recession has an adverse impact on health status, depression, smoking, heavy drinking, illicit drug use, exercise, and fruit intake; and second, these adverse effects are especially pronounced among loweducated individuals. Research Papers in Progress: “Equality of Educational Opportunity in China”, with James J. Heckman (University of Chicago), Sha Tao (Beijing Normal University), and Yike Wang (University of Chicago) Using a new and nationally representative large sample of children in China aged six to fifteen, we study socioeconomic determinants of children's cognitive and noncognitive ability development. Building upon the recent development on factor analysis and interactive fixed effect model, we estimate the causal effects of various school inputs and family inputs (as well as their interactions) on children's human capital development. Our estimator is consistent and allows for flexible correlations between individual level unobservables and observed inputs. Using estimated human capital production function, we then decompose the inequality of children's skill development into variations in the family inputs and school inputs. “Real Effects of Financial (Dis)integration: A Spatial Equilibrium Analysis of Europe”, with Indraneel Chakraborty (University of Miami), Hans Holter (University of Oslo), and Serhiy Stepanchuk (Ecole Polytechnique Federale de Lausanne). Using data from 15 Euro area economies, we quantify the real effects of supply side frictions in bank lending due to financial separation of European countries since the crisis. We develop a spatial equilibrium model with heterogeneous countries. Banks allocate capital endogenously across countries, which in turn determines firms' cost of capital and nations' wealth. Costs of financial segmentation include reduction in access to capital for firms and resulting output reduction; and reduction in households' wages and consumption. We show that financial segmentation can explain between a quarter and a third of the reduction in firm output since the crisis. We also estimate the benefits of further financial integration to the European economy. Programming Skills: Extensive experience with C++ (including OpenMP & MPI), Fortran, Matlab, SAS, Stata, R, Latex Completed Research Papers Rong Hai A Dynamic Model of Health, Education and Wealth with Credit Constraints and Rational Addiction, with James J. Heckman (UChicago) This paper develops and structurally estimates a life-cycle model where health, education, and wealth are endogenous accumulated processes depending on the history of an individual's optimal behaviors, on parental factors, and on cognitive and noncognitive abilities. The model investigates many different pathways between education, health, and wealth by introducing endogenous health capital production and addictive preferences of unhealthy behavior in the presence of credit constraints. Using data from National Longitudinal Survey of Youth 97, we estimate the model using a two-step estimation procedure based on factor analysis and simulated method of moments. The estimated model decomposes the causal effects of education on health into the direct benefits of improving health production efficiency and the indirect benefits of reducing unhealthy behavior and raising earnings. We show that rational addiction has important quantitative implication on predicted patterns of unhealthy behavior by socioeconomic status and over the life-cycle. We find sizable impacts of both health and parental transfers on individuals' college decisions. We find that relaxing credit constraints improves education while its effects on healthy behavior and health evolves nonlinearly over the life cycle. Finally, we show that both cognitive and noncognitive factors are important determinants of health, education, and wealth. The Determinants of Rising Inequality in Health Insurance and Wages, July 2015 Under Review Over the last 30 years in the U.S., less educated workers have experienced a sharp decline in health insurance coverage rate and stagnant wage growth. In contrast, more educated workers’ health insurance coverage rate has stayed relatively stable and their wages have rapidly grown. This paper investigates the determinants of the increase in inequality in health insurance coverage and wages by estimating an overlapping generations equilibrium model of labor and health insurance markets’ demand and supply. The estimated model is used to quantify the effects of changes in aggregate factors (including rising cost of medical care services, Medicaid eligibility expansion, skill-biased technological changes in the labor market, and changes in the labor force composition) on the inequality of health insurance coverage and wages. I find that the interaction between the rising cost of medical services and labor market technological change is the most important determinant of the widening gap of health insurance coverage. On the Welfare Cost of Consumption Fluctuations in the Presence of Memorable Goods, with Dirk Krueger (UPenn) and Andrew Postlewaite (UPenn), Jan 2015 We propose a new category of consumption goods, memorable goods, that generate a flow of utility after consumption. We analyze an otherwise standard consumption model that distinguishes memorable goods from other nondurable goods. Consumers optimally choose lumpy consumption of memorable goods. We empirically document differences between levels and volatilities of memorable and other goods expenditures. Memorable goods expenditures are about twice durable goods expenditures and half the volatility. The welfare cost of consumption fluctuations driven by income shocks is overstated if memorable goods are not accounted for and estimates of excess sensitivity of consumption might be due to memorable goods. Income Inequality, Tax Policy, and Economic Growth, with Siddhartha Biswas (UNC) and Indraneel Chakraborty (UMiami), July 2015 Under Review We investigate how reduction of income inequality through tax policy affects economic growth. Taxation at different points of the income distribution has heterogeneous impacts on households’ incentives to invest, work, and consume. Using U.S. state-level data and micro-level household tax returns over the last three decades, we find that reduction of income inequality between low and median income households improves economic growth. However, reduction of income inequality through taxation between median and high income households reduces economic growth. These economic growth effects are attributable both to supply-side factors (changes in small business activity and labor supply) and to consumption demand. The Impact of an Early Career Recession on Health and Health-Related Behaviors, with Naijia Guo (CUHK), June 2015 Under Review Using National Longitudinal Survey of Youth 1997, we estimate the long-term impact of an early career recession on various health outcomes and health-related behaviors. The early career recession is measured by the state-specific unemployment rate in the year of labor market entry. Using an instrumental variable approach, we address the potential selection issues on the timing of labor market entry. We find that first, an early career recession has an adverse impact on health status, depression, smoking, heavy drinking, illicit drug use, exercise, and fruit intake; and second, these adverse effects are especially pronounced among low-educated individuals. Working in Progress Equality of Educational Opportunity in China, with James J. Heckman (UChicago), Sha Tao (BNU), and Yike Wang (UChicago) Using a new and nationally representative large sample of children in China aged six to fifteen, we study socioeconomic determinants of children's cognitive and noncognitive ability development. Building upon the recent development on factor analysis and interactive fixed effect model, we estimate the causal effects of various school inputs and family inputs (as well as their interactions) on children's human capital development. Our estimator is consistent and allows for flexible correlations between individual level unobservables and observed inputs. Using estimated human capital production function, we then decompose the inequality of children's skill development into variations in the family inputs and school inputs. Real Effects of Financial (Dis)integration: A Spatial Equilibrium Analysis of Europe, with Indraneel Chakraborty (UMiami), Hans Holter (Oslo), and Serhiy Stepanchuk (EPFL). Using data from 15 Euro area economies, we quantify the real effects of supply side frictions in bank lending due to financial separation of European countries since the crisis. We develop a spatial equilibrium model with heterogeneous countries. Banks allocate capital endogenously across countries, which in turn determines firms' cost of capital and nations' wealth. Costs of financial segmentation include reduction in access to capital for firms and resulting output reduction; and reduction in households' wages and consumption. We show that financial segmentation can explain between a quarter and a third of the reduction in firm output since the crisis. We also estimate the benefits of further financial integration to the European economy. JU HU <https://economics.sas.upenn.edu/graduate-program/candidates/ju-hu> <juhu1@sas.upenn.edu> UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN @ ECON.UPENN.EDU Office Contact Information 328 McNeil Building 3718 Locust Walk Philadelphia, PA 19104 Cell Phone: 267-909-1250 Personal Information: Citizenship: China Date of Birth: April 02, 1985 Gender: Male Undergraduate Studies: B.A., Finance, Fudan University, Shanghai, China, 2007 Masters Level Work: M.A., Finance, Fudan University, Shanghai, China, 2010 Graduate Studies: University of Pennsylvania, 2010 to present Thesis Title: “Essays on Reputations and Dynamic Games” Expected Completion Date: May 2016 Thesis Committee and References: Professor George J. Mailath (Advisor) 432 McNeil Building 3718 Locust Walk Philadelphia, PA 19104 gmailath@econ.upenn.edu 215-898-7908 Professor Yuichi Yamamoto 457 McNeil Building 3718 Locust Walk Philadelphia, PA 19104 yyam@sas.upenn.edu 215-898-8761 Teaching and Research Fields: Microeconomic Theory, Game Theory Professor Rakesh V. Vohra 451 McNeil Building 3718 Locust Walk Philadelphia, PA 19104 rvohra@seas.upenn.edu 215-898-6777 215-898-6880 215-898-7350 215-898-5691 Teaching Experience: Fall, 2011 Fall, 2012 Spring, 2012, 2013 Spring, 2015 Summer, 2012, 2013, 2014 Econ 701 Microeconomic Theory I (graduate level), University of Pennsylvania, Teaching Assistant for Prof. Andrew Postlewaite and Steven Matthews Econ 701 Microeconomic Theory I (graduate level), University of Pennsylvania, Teaching Assistant for Prof. Andrew Postlewaite and Mallesh Pai Econ 212 Game Theory, University of Pennsylvania, Teaching Assistant for Prof. Tymofiy Mylovanov Econ 212 Game Theory, University of Pennsylvania, Teaching Assistant for Prof. Yuichi Yamamoto Econ 897 Math Camp (Session II), University of Pennsylvania, Instructor Professional Activities: Presentations: 2012-2015 Penn Micro Theory Lunch Club, University of Pennsylvania 2015 Micro Theory Seminar, University of Pennsylvania Referee Activities: Theoretic Economics, International Economic Review Other: 2013-2014 Co-organizer, University of Pennsylvania Micro Theory Lunch Club Honors, Scholarships, and Fellowships: 2011-2013 Xinmei Zhang Fellowship, School of Arts and Science, University of Pennsylvania 2013-2014 Sidney Weintraub Memorial Fellowship, University of Pennsylvania Publications: “Reputation in the Presence of Noisy Exogenous Learning”, Journal of Economic Theory, 2014, Volume 153, 64-73. Research Paper: “Biased Learning and Permanent Reputation” (Job Market Paper) Abstract: This paper studies reputation effects between a long-lived seller and different short-lived buyers where the short-lived buyers do not know how long the seller has been in business. Departing from standard assumptions in repeated games, this paper assumes that buyers enter the market at random times and only observe a coarse public signal upon entry. The signal measures the difference between the number of good and bad outcomes in a biased way: a good outcome is more likely to increase the signal than a bad outcome to decrease it. The seller has a short-run incentive to exert low effort, but makes high profits if it were possible to commit to high effort. First, we show if the bias is large, in the complete information game, always exerting low effort is the unique equilibrium. We then introduce incomplete information. There are two types of the seller. One is a commitment type who always exerts high effort and the other is a normal type who behaves strategically to maximize long-run payoff. If there is small but positive chance that the seller is a commitment type, in any equilibrium, the normal seller must exert high effort at some signals to build up his reputation. Moreover, the seller builds up his reputation only to milk it. In any equilibrium, once the seller builds up reputation through reaching a high enough signal, the seller then exploits by exerting low effort. Because all buyers only have limited information, they are unable to distinguish the two types of the seller no matter how long the game has been played. Consequently, the incentives of building reputation never disappear in the long-run. “Reputation in the Presence of Noisy Exogenous Learning” Abstract: This paper studies the reputation effect in which a long-lived player faces a sequence of uninformed short-lived players and the uninformed players receive informative but noisy exogenous signals about the type of the long-lived player. We provide an explicit lower bound on all Nash equilibria payoffs of the long-lived player. The lower bound shows when the exogenous signals are sufficiently noisy and the long-lived player is patient, he can be assured of a payoff strictly higher than his minmax payoff. Research Papers in Progress “Social Learning and Market Experimentation” (Manuscript in preparation) Abstract: This paper studies optimal dynamic monopoly pricing when a monopolist sells a product with unknown quality to a sequence of short-lived buyers who have private information about the quality. Because past prices and buyers’ purchase behavior convey information about private signals, they jointly determine the public belief about the quality of the monopolist’s product. The monopolist’s is essentially doing experimentation in the market because every price charged generates not only current period profit but also additional information about the quality. We focus on information structures with a continuum of signals. Under a mild regularity condition on information structures, we show in equilibrium, the optimal price is an increasing function of the public beliefs. In addition, we fully characterize information cascade sets in terms of information structure. We find the standard characterization in terms of boundedness of information structure in the social learning literature no longer holds in the presence of a monopoly. In fact, whether herding occurs or not depends more on the values of the conditional densities of the signals at the lowest signal. Reputations in Dynamic Games Ju Hu October 14, 2015 In the paper “Biased Learning and Permanent Reputation”, I study reputation effects between a long-lived seller and different short-lived buyers where the short-lived buyers do not know how long the seller has been in business. Departing from standard assumptions in repeated games, this paper assumes that buyers enter the market at random times and only observe a coarse public signal upon entry. In particular, they do not know when the game started. The signal measures the difference between the number of good and bad outcomes in a biased way: a good outcome is more likely to increase the signal than a bad outcome to decrease it. This information setting models online rating systems such as eBay feedback score and the bias reflects the fact recently found in the empirical literature that satisfied buyers are more likely to give a positive feedback than unsatisfied buyers are to give a negative feedback. The long-lived seller faces a moral hazard problem: he has a short-term incentive to exert low effort, but makes higher profits if it were possible to commit to high effort. Buyers are unsure of the characteristics of the seller. There are two types of the seller. One is a commitment type who always exerts high effort. The other is a normal type who behaves strategically to maximize long-run payoff. This information structure has two conflicting effects on reputations. On one hand, buyers only have very coarse information about the past and face a complicated inference problem about the type of the seller since they do not know how long the seller has been in business. This discourages the seller’s reputation building because his ability to manipulate buyers’ beliefs through past behavior is limited. On the other hand, because bad outcomes are less likely to change the public signal, the seller has a larger incentive to milk his reputation which encourages him to build reputation first. 1 In the absence of commitment type, as a benchmark, I show that high effort can not be supported as equilibrium outcome. In particular, in the unique equilibrium, the seller always exerts low effort. In contrast, if buyers’ prior belief about the commitment type is small but positive, in every equilibrium, the normal seller must exert high effort at some signals. The normal seller builds up his reputation by imitating the commitment type and exerting high effort in order to increase the chance of getting a higher signal. Moreover, I show that the seller is not willing to do so at all signals. In fact, the seller builds up his reputation only to milk it. In every equilibrium, once the seller builds up reputation through reaching a high enough signal, the uninformed buyers are convinced that they are facing the commitment type with large probability. The normal seller then exploits by exerting low effort. As a result of low effort, the public signal gradually decreases with larger probability. Once it becomes small, the seller again has incentive to build it up. Because all buyers only have limited information about the past, they are unable to distinguish the commitment and normal sellers no matter how long the game has been played. Consequently, the incentives of building reputation never disappear in the long-run. In another paper, “Reputation in the Presence of Noisy Exogenous Learning”, I explores reputation effects in which a long-lived player faces a sequence of uninformed short-lived players and the uninformed players receive informative but noisy exogenous signals about the type of the long-lived player. Because informative exogenous signals will eventually reveal the true type of the long-lived player to the short-lived players, typically there is no guarantee that the short-lived players play a best response to a commitment action in all but finitely many periods even if this commitment action is always played, as is the case in the standard reputation model. Nonetheless, this paper characterizes the reputation effects by providing an explicit lower bound on all Nash equilibria payoffs of the long-lived player. The lower bound is characterized by prior belief over types, discount factor and most importantly the speed of exogenous learning. Because of exogenous learning, in general this lower bound is lower than that in the standard reputation models without exogenous signals, and these two bounds coincide if the signals are completely uninformative. In particular, this lower bounds shows when the exogenous signals are sufficiently noisy, hence slow speed of learning, and the long-lived player is patient, he can be assured of a payoff strictly higher than his minmax payoff. 2 JUNWEN CAROLINE LIU https://economics.sas.upenn.edu/graduate-program/candidates/junwen-caroline-liu junwenl@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN @ ECON.UPENN.EDU 215-898-6880 215-898-7350 215-898-5691 Office Contact Information Department of Economics University of Pennsylvania 160 McNeil Building Philadelphia, PA 19104 Cell: (+1)215-688-1211 Undergraduate Studies: BSc in Economics and Finance (First Class Honors) Hong Kong University of Science and Technology, 2010 Graduate Studies: University of Pennsylvania, 2010-Present Thesis Title: “Essays in Empirical Microeconomics” Expected Completion Date: May 2016 Thesis Committee and References: Professor Hanming Fang (Advisor) University of Pennsylvania Department of Economics Philadelphia, PA 19104 215-898-7767 hanming.fang@econ.upenn.edu Professor Kenneth I. Wolpin Rice University Department of Economics Houston, TX 713-348-2522 kenneth.i.wolpin@rice.edu Professor Petra E. Todd University of Pennsylvania Department of Economics Philadelphia, PA 19104 215-898-4084 ptodd@econ.upenn.edu Teaching and Research Fields: Research Fields: Labor Economics, Public Economics, Education, Consumer Finance Teaching Fields: Labor Economics, Public Economics, International Economics Teaching Experience (University of Pennsylvania): Spring, 2015 Intermediate Microeconomics, Teaching Assistant for Prof. Kenneth Burdett Fall, 2014 International Economics, Teaching Assistant for Prof. Cecilia Fieler Spring, 2014 International Economics, Instructor Spring, 2012/13 International Economics, Teaching Assistant for Prof. Iourii Manovskii Summer, 2012 Introduction to Macroeconomics, Instructor Fall, 2011 Introduction to Economics for Business, Teaching Assistant for Prof. Uriel Spiegel Research Experience and Other Employment: 2012-2014 Research Assistant for Prof. Hanming Fang, University of Pennsylvania Project on the elderly: Summarized the risks facing the elderly by surveying extensively various strands of literature. Collected information from multiple data sources for the OECD countries and analyzed the data using advanced econometric techniques, assisted in the writing of a Handbook chapter. Project on the Chinese labor market: Analyzed a large administrative data set from a Chinese employment website with 1 million observations for each of the 38 variables. Identified interesting data patterns and assisted in developing quantitative models to capture key data features. Research Papers: “The Effects of Tuition and Student Loan Policies on College Outcomes and Lifetime Earnings” (Job Market Paper) To increase college access and reduce the burden of student loan debt, the US government has recently developed several new tuition and student loan policies. These include the newly proposed free community college plan and the recently enacted Pay As You Earn plan that makes student loan repayments contingent on earnings. In this paper, I develop and estimate a dynamic life-cycle model of the decisions individuals make with regard to schooling, work, savings and student loan borrowing. The model is estimated with micro-level US data and is used to evaluate the effects of these educational policies on education outcomes, lifetime earnings and welfare. My results show that the free community college plan benefits individuals from lower-income families the most, increasing their community college enrollment rate by 17 percentage points from 41 percent to 58 percent. However, it reduces the population proportion of individuals who achieve a bachelor's degree by 9 percent. The Pay As You Earn plan reduces labor supply in college, lowers the time it takes to complete a bachelor's degree, and enables individuals to attend higher-quality colleges. The overall education level is improved with the percent of individuals holding a bachelor's degree increasing from 31 to 33 percent. I also evaluate the effects of a hypothetical loan forgiveness plan for college dropouts, which is found to increase college enrollment but reduce college completion. Of the three policies, the Pay As You Earn plan achieves the highest welfare gain and reduces lifetime earnings inequality. “Student Loan Debt and Risk-Return Trade-Off in Occupation Choices” (Work In Progress) The total outstanding student loan debt almost quadrupled between 2004 and 2015. This rapid growth in student loan debt has raised concerns about the effects of student loan debt on individuals' post-graduation labor market outcomes. In this paper, I study how student loan debt affects the risk-return trade-off in the occupation choices after college. Using data on a recent US cohort, I document that college graduates with student loan debt on average have lower post-graduation wage levels and wage dispersion. I capture the riskreturn aspect of occupations by categorizing jobs into performance-pay and non-performance-pay jobs. The data show that individuals with student loan debt are less likely to choose performance-pay jobs. Motivated by these data patterns, I build a dynamic life-cycle model in which college graduates make occupation and savings decisions and occupations differ in wage returns and risk levels. Simulation results show that higher student loan debt levels make individuals more risk averse and push them toward choosing the less risky and lower-paying occupation while they repay the debt. As individuals accumulate occupation-specific human capital, it becomes more costly to move away from the lower-paying occupation even after their student loan debt is repaid, which leads to a lower wage level later in life. Skills Computation: Fortran, Latex, Matlab, Microsoft Office, Stata; Language: Chinese (native), English (fluent) Junwen Caroline Liu University of Pennsylvania The Effects of Tuition and Student Loan Policies on College Outcomes and Lifetime Earnings How can the government increase college access and reduce the student loan debt burden? This question has become a major focus of U.S. educational policy. To tackle the problem, the Obama administration has developed many initiatives. For example, in his 2015 State of the Union Address, President Obama proposed a free community college plan that would make two years of community college education free for all students. To improve federal student loan offerings, the government has implemented a new repayment plan called Pay As You Earn (PAYE) in 2012. These policies are expected to be costly, and the outcomes are not immediately available for assessment. However, it is possible to evaluate the effects of these policies ex ante using a behavioral model. In this paper, I develop and estimate a dynamic life-cycle model of the decisions that individuals make with regard to college choice, years of schooling, work, savings and student loan borrowing. I use the estimated model to answer two questions. First, how effective are the new tuition and student loan policies in improving education outcomes and lifetime earnings? Second, what are the fiscal costs of the policies and are they welfare-improving after accounting for fiscal costs? A number of distinct factors motivated these recent higher education policy initiatives. First, despite existing need-based student grant and loan programs, there remains a wide college attainment gap with respect to family income levels. Second, students from lower-income families are more likely to attend colleges with a quality level that is below what their SAT score qualifies them for. Third, the time it takes to complete a bachelor's degree has risen over time. Fourth, the majority of students rely on student loan debt to finance their college education. The debt burden for many students is high and it has potential repercussions for savings and labor market decisions. To evaluate the new tuition and student loan reforms in a way that captures these considerations, I incorporate three novel features into my model. First, the colleges considered in the model are of different qualities.1 Specifically, I allow for five types of colleges: community colleges and four types of four-year colleges varying by college quality. Having different college quality levels captures the trade-off between college quality and cost. Second, my model incorporates both student loan debt and standard savings decisions. Third, my model endogenizes labor supply decisions while in college, which affects loan and savings choices and also possibly influences college completion. Individuals are heterogeneous in family income, SAT scores and unobserved heterogeneity. They are risk averse and face uncertainty over both schooling accumulation and labor income. These risks imply that the student loan debt repayment structure potentially affects their schooling and subsequent labor supply decisions. I estimate the model using data from the National Longitudinal Survey of Youth 1997 (NLSY97) and the Integrated Postsecondary Education Data System (IPEDS). The NLSY97 follows a nationally representative sample of individuals who were born between 1980 and 1984 in the US. They have been interviewed on an annual basis since 1997. The data set contains rich information on college choices, employment, asset and student loan debt, which allows identification of model parameters. IPEDS includes a range of information on colleges in the US, such as institutional characteristics, 1 Following a similar approach as Black and Smith (2005), I construct the quality index using average SAT score and average faculty salary. I categorize four-year colleges into four quality levels based on the index. admission and test scores, and student charges. I employ this information to calculate the net cost of attendance for each college enrollee and to construct a college quality index. Estimation is carried out using the Simulated Method of Moments. Key data moments are matched along various dimensions, including enrollment patterns, college outcomes, the labor supply and average wages series, and student loan debt level. Using the estimated model, three policy experiments are conducted. First, the impact of the proposed free community college plan is examined. Second, the newly implemented Pay As You Earn repayment plan is considered. Lastly, a hypothetical loan forgiveness program for college dropouts is assessed. This plan is motivated by the low college completion rate, and the sizable student loan debt college dropouts have on average upon exiting college. Under the free community college plan, I find that the community college enrollment increases significantly for individuals from lower-income families. Among individuals with family income below the median, community college enrollment rate increases by 17.30 percentage points from 40.00% to 57.30%. The associate degree completion rate also rises considerably. The fraction of individuals in the population with an associate degree increases from 9.20% to 16.35%. However, it reduces the population proportion of individuals who achieve a bachelor's degree by 9.00 percent from 31.10% to 28.30%. There is a decrease in earnings inequality among lower-earning individuals. The Pay As You Earn plan increases the rate of enrollment in both the community and four-year colleges primarily for individuals from lower-income families. The plan also enables students to attend higher quality colleges. For students whose family income is lower than the median, the fraction of students enrolled in the top two college quality levels goes up by 2.40%, from 41.30% to 42.30%. Under this plan, the average years worked by four-year college graduation falls by 5.60%, from 2.33 years to 2.20 years. The time it takes to receive a bachelor’s degree declines slightly. The overall education level in the population improves as the fraction of individuals who have a bachelor’s degree increasing from 31.10% to 32.78%. There is a reduction in earnings inequality in the upper half of the distribution. The loan forgiveness plan for college dropouts increases college enrollment but discourages students from completing the degree. The fraction of individuals who earn an associate degree falls by 28.30%, from 9.20% to 6.60%, while the fraction who earn a bachelor's degree falls by 8.68%, from 31.10% to 28.40%. The reduction in the number of individuals who hold a college degree reduces average lifetime earnings. In terms of fiscal costs, the Pay As You Earn plan costs the least while the loan forgiveness plan is the most expensive. All three polices are welfare improving for a utilitarian government that maintains a neutral budget.2 The Pay As you Earn plan leads to the highest welfare gain, and the free community college plan ranks second. Of the three policies, the Pay As You Earn plan produces the largest improvement in the quality of the college attended and in the four-year college completion rate. The free community college plan is the most effective in terms of improving college outcomes for lower-income students. The two policies are complementary. 2 To make welfare comparisons, I adjust the labor tax rate to maintain a neutral government budget by re-solving the model and simulating it under the new tax rate. Note that the effects of all polices on college outcomes and lifetime earnings presented here derive from simulations that do not impose the cost of the policy on individuals. These results are qualitatively and largely quantitatively unaffected when the costs are imposed through tax changes. EKIM CEM MUYAN https://economics.sas.upenn.edu/graduate-program/candidates/ekim-cem-muyan muyan@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN @ ECON.UPENN.EDU 215-898-6880 215-898-7350 215-898-5691 Office Contact Information Department of Economics 3718 Locust Walk Rm 160 Philadelphia, PA 19104 Phone: +1 (267) 994-6081 Citizenship: Turkish Undergraduate Studies: BA, Economics with a minor in Mathematics, Sabanci University, Dean’s Highest Honors, Istanbul, Turkey, 2008 Masters Level Work: Laurea Magistralis, Economics, Università di Bologna, Summa Cum Laude, Bologna, Italy, 2010 Graduate Studies: University of Pennsylvania, 2010 to present Thesis Title: “Essays in Political Economics” Expected Completion Date: June 2016 Thesis Committee and References: Professor Antonio Merlo Rice University Department of Economics 6100 Main Street Houston, TX 77005 +1 (713) 348-3699 amerlo@rice.edu Professor Kenneth I. Wolpin Rice University Department of Economics 6100 Main Street Houston, TX 77005 +1 (713) 348-2522 kenneth.i.wolpin@rice.edu Professor Camilo Garcia-Jimeno University of Pennsylvania Department of Economics 3718 Locust Walk Philadelphia, PA 19104 +1 (215) 898-8206 gcamilo@sas.upenn.edu Research Fields: Political Economics, Development Economics, Public Economics Teaching Experience: Fall 14 Mathematics for Economist (Graduate), Rice University, Instructor Spring 12/14 Political Economics, University of Pennsylvania, T.A. for Professor Antonio Merlo Fall 12/13 Honors Game Theory, University of Pennsylvania, T.A. for Professor Andrew Postlewaite Summer 13 Statistics for Economists, University of Pennsylvania, Instructor Fall 13 Public Economics (Graduate), University of Pennsylvania, T.A. for Professor Andrew Postlewaite Fall 11/12 Game Theory, University of Pennsylvania, T.A. for Professor Steve Matthews Fall 09 Econometrics (Masters), University of Bologna, T.A. for Andrea Ichino Research Experience and Other Employment: 2009 University of Bologna, Research Assistant for Professor Andrea Ichino Other Professional Activities: 2014 Teaching Mentor, Economics Department, University of Pennsylvania Honors, Scholarships, and Fellowships: 2010-2015 University Fellowship, University of Pennsylvania 2009-2010 Fellow of Collegio Superiore, University of Bologna 2005-2008 Sakip Sabanci Full Scholarship, Sabanci University Research Papers: “Public Investment and Preference Aggregation under Alternative Decentralization Institutions” (Job Market Paper) Decentralization institutions differ from each other in terms of the degree of power they grant to local authorities in decision-making. Successfully designing decentralization institutions depends on understanding the local and central authorities' preferences over the types of public investments and how alternative decentralization institutions aggregate them. Focusing on these key components, I build and estimate a dynamic committee decision-making model to study how public investment choices vary with the degree of power granted to local governments. I characterize alternative decentralization institutions as voting mechanisms the committee can employ. I implement my model using a novel dataset from a unique institution in Colombia. I find that the local governments are more likely to invest on targeted transfers than is the central government. Counterfactual exercises show that a complete decentralized system would significantly increase the number and size of the targeted transfer spending. “How many votes does a dollar buy: Campaign Spending and Strategy in the U.S. Congressional Elections” (with Devin J. Reilly) A central yet unresolved question in political economy is the degree to which money affects election outcomes. We argue that answering this question requires one to take the productivity of different campaign strategies into account. To this end, we construct a model of political campaigns in which candidates allocate their budgets between positive and negative campaigning. This allows us to estimate the marginal productivities of such spending through the equilibrium of a political campaign game. Elections vary according to politician and districtspecific characteristics, as well as the unobservable (to the econometrician) measure of voter types. We structurally estimate our model using a wide array of data, including the Wisconsin Political Advertisement project, FEC campaign finance records, election results, and politician characteristics. Our estimates show that campaign spending overall is not particularly effective at increasing votes -- a 10% increase in the average Democratic candidate's budget, corresponding to about a $230,000 increase, would raise his or her expected vote share by 0.4 percentage points. This finding is consistent with the previous literature. Research Papers in Progress: “The Network Economics of Foreign Aid” (with Camilo Garcia-Jimeno) In this paper we study a novel aspect of the strategic nature of foreign aid. Although it has been suggested that aid is used as a tool to influence policy, we argue that the underlying ties between recipient countries lead to influence spillovers. We argue that these spillovers are of first order importance from the point of view of donor countries, which implies that bilateral aid flows cannot be treated in isolation. Thus, we develop and estimate a network model of foreign aid where donor countries provide aid to recipient countries to influence their polices. Recipient countries are linked to each other in a way that a unit of influence over a recipient also indirectly influences its neighbors. Both direct and indirect influence considerations shape donor’s aid incentives. Because donors compete for policy influence with each other, our model predicts that the strategic nature (strategic complementarity versus substitutability) of the competition game between donors is endogenous and depends on the relative policy positions of donors and recipients. We implement our model with data we construct from UN votes and international aid flows. We estimate the best response dynamics of the model and analyze the entrance of China to the foreign aid market. “Why do PACs donate: Ideology or Favors” (with Devin J. Reilly) Political Action Committees (PACs) donate large sums of money to political campaigns. In 2000, around 25% of the candidate expenditures was financed by around 3,000 active PACs. We investigate the motivations behind these donations. A PAC has two potential motivations to donate to a campaign. The first is that, by contributing, the PAC may increase its probability to receive favors or have candidate access ("favor benefit"). The second is, if the PAC is ideologically similar to a candidate, it may contribute simply to help the candidate win ("inherent benefit"), independent of any specific "favors" the candidate may provide conditional on contributing. We build a model of campaign contributions where these donations affect the probability a candidate wins and the "favors" the PAC may receive, but not the ideological similarity between a single PAC and candidate. Thus, this model captures both potential motivations for political contributions. We use estimates from Muyan and Reilly (2015) to approximate the marginal impact of a PAC contribution to the probability of winning. We argue that the identification of the model parameters is possible due to the variations in donations a candidate receives from all PACs and variations in donations of a single PAC to candidates. With data from the 2000, 2004 and 2008 Congressional elections, as well as detailed PAC and candidate characteristics, we implement our model using energy PACs grouped into various ideological camps. RESEARCH STATEMENT Ekim Cem Muyan https://tinyurl.com/ekimcmuyan muyan@sas.upenn.edu My research spans the intersection of political institutions and economics, with a particular focus on development. The questions I ask go in both directions of this relationship: how do political institutions impact economic outcomes and how do economic relationships shape political institutions? During my time as a graduate student at the University of Pennsylvania, I initiated projects that answered questions in both directions. The goal of my research agenda is to provide theoretical and empirical insights that shape and guide policy debates governing institutional design and development, and to understand the effects of institutions in rigorous quantitative detail. In what follows, first, I briefly summarize the three projects I am currently working on. My job market paper project is in the first direction: how do political institutions shape economic outcomes? The other two, which amount to a total of three papers, are about the impact of economic relations on political outcomes. For each of these projects, I discuss future avenues of research. Finally, I end with two other projects, currently in their initial stages, that I hope to develop further in the first years as an assistant professor. In my job market paper, titled Public Investment and Preference Aggregation Under Alternative Decentralization Institutions, I take the heterogeneity of decentralization institutions around the globe seriously and ask how the degree of power granted to local governments impacts public investments. This is distinct from the previous empirical and theoretical literature that viewed decentralization as a binary process. Answering this question is a challenging task that requires several layers of analysis. First, I build a model that can flexibly represent different degrees of decentralization. To do this, I treat and model decentralization institutions as voting mechanisms that aggregate the preferences of local and central governments. A voting mechanism can be mapped into a particular decentralization institution depending on the weight assigned to the preferences of each layer of government. The model, therefore, establishes a systematic method to analyze the effects of decentralization institutions on public investment choices. Second, I provide an empirical method to identify and estimate the preferences of different layers of government exploiting a unique institution in Colombia. My model, together with empirically uncovering the preferences of different layers of government, enables me to predict the outcome of any possible decentralization attempt in Colombia. I find that the local governments are more (less) prone to invest on targeted transfers (public goods) compared to the central government. This leads to a striking result in my counterfactuals that complete decentralization would significantly increase the number and size of the targeted transfer projects and decrease the number and size of the public goods projects. The share of expenditure that is directed toward public goods would decrease by about 16 percentage points. The next step in this line of research is to investigate the reasons behind the divergence of the preferences between the local and central governments. The framework that studies local elite capture and political agency problems at local and central levels is a promising start toward this direction. In another paper I am working on with Camilo Garcia-Jimeno at the University of Pennsylvania, titled The Network Economics of Foreign Aid, we study the strategic nature of foreign aid. Although it has been suggested in the literature that aid is used as a tool to influence policy, we argue that the underlying ties between recipient countries lead to influence spillovers. We argue that these spillovers are of first order importance from the point of view of donor countries, which implies that bilateral aid flows cannot be treated in isolation. Thus, we develop and estimate a network model of foreign aid where donor countries provide aid to recipient countries to influence their polices. Recipient countries are linked to each other in a way that a unit of influence over a recipient also indirectly influences its neighbors. Both direct and indirect RESEARCH STATEMENT Ekim Cem Muyan https://tinyurl.com/ekimcmuyan muyan@sas.upenn.edu influence considerations shape donor’s aid incentives. Because donors compete for policy influence with each other, our model predicts that the strategic nature (strategic complementarity versus substitutability) of the competition game between donors is endogenous and depends on the relative policy positions of donors and recipients. Moreover, the recipient countries that are strategically placed in the recipient network (the geopolitically important countries) receive more aid than countries that could benefit from it more. We implement our model with data we construct from UN votes and international aid flows. We estimate the best response dynamics of the model and analyze the entrance of China to the foreign aid market. Currently, this paper is at its empirical implementation stage and we expect it to be ready in a couple of months. The future avenue in this research is to ask the following simple but very important question: do countries with geopolitical importance substitute aid for building state capacity? If this channel is important, it could be possible that aid actually hurts some countries in the long run, instead of helping them. Another working project, with Devin J. Reilly at the University of Pennsylvania, is on the influence of campaign contributions on elections and the motivations of Political Action Committees (PACs) in contributing to political campaigns. We have collected a very large dataset that contains campaign strategies of each congressional candidate, their characteristics as well as dyadic records of campaign contributions and detailed PAC characteristics for the period 2000-2008. This project thus far produced one working paper and another paper is still in progress. The first paper, titled How Many Votes Does a Dollar Buy: Campaign Spending and Strategy in the U.S. Congressional Elections, investigates a central yet unresolved question in political economy: how does money influence election outcomes? To do this, we develop a model of strategic campaigning where the candidates allocate their budgets between negative and positive campaigning. We then estimate this model and find that a 10% increase in the average Democratic candidate’s budget would raise her expected vote share by about 0.97%. In the other paper, titled Why Do PACs Donate: Ideology or Favors, we investigate the motivation behind PAC donations to congressional candidates. The main channel of identification is the fact that the favors depend on the magnitude of contributions whereas the ideological proximity does not. As we recover the impact of contributions on the probability of winning in the previous paper, our methodology outlines a novel way to identify these motivations. This paper is in the empirical implementation phase and I expect the working paper to be out in about a year. In the future, I would like to build on my agenda to understand the connection between political institutions and economic development. A particular interest is to analyze when good policies are not good politics. Voters often reward bad policies more than the good ones. This is especially true in developing economies where broad public programs with proven development records find little electoral support from the voters. Hence, these countries fail to provide basic public goods and services. I have developed a framework where I can identify and estimate voter responsiveness to alternative development policies using the set of Randomized Control Trials (RCTs) in a country. I have begun to collect village level local election results in India. The next step is to collect village level RCT data, which is proprietary, and requires a long bureaucratic process. More information about this project is available upon request. Another project I am working on is about how governments in developing countries use public procurements for their entrenchment. I have collected primary data on all public procurements in Turkey in the years 2002-2010 as well as inter-company network data to initialize this project. DANIEL NEUHANN https://economics.sas.upenn.edu/graduate-program/candidates/daniel-neuhann neuhann@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii MANOVSKI@ ECON.UPENN.EDU 215-898-6880 APOSTLEW@ECON.UPENN.EDU 215-898-7350 @ ECON.UPENN.EDU 215-898-5691 Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn KQUINN Office Contact Information 3718 Locust Walk, 160 McNeil Building Philadelphia, PA 19104 C: +1 (646) 238 - 3915 Citizenship: Germany Undergraduate Studies: B.Sc., Economics, Humboldt-Universität zu Berlin, 2009 Graduate Studies: University of Pennsylvania, 2010 to present Thesis Title: “Essays on the Macroeconomics of Financial Markets” Expected Completion Date: June 2016 Thesis Committee and References: Professor Harold L. Cole Department of Economics University of Pennsylvania Philadelphia, PA 19104 Phone: +1 (215) 898-7788 Email: colehl@sas.upenn.edu Professor Guillermo Ordonez Department of Economics University of Pennsylvania Philadelphia, PA 19104 Phone: +1 (215) 898-1875 Email: ordonez@econ.upenn.edu Professor Itay Goldstein Finance Department - The Wharton School University of Pennsylvania Philadelphia, PA 19104 Phone: +1 (215) 746 - 0499 Email: itayg@wharton.upenn.edu Professor Dirk Krueger Department of Economics University of Pennsylvania Philadelphia, PA 19104 Phone: +1 (215) 573-1424 Email: dkrueger@econ.upenn.edu Pre-Ph.D. Graduate Studies: M.Sc., Economics, Tilburg University, 2010 Teaching and Research Fields: Primary fields: Macroeconomics, Financial Economics Secondary fields: Applied Microeconomic Theory, Financial Institutions, Banking Teaching Experience at University of Pennsylvania: Fall 2011 Introduction to Macroeconomics, Recitation Instructor for Prof. Luca Bossi Spring 2012 Introduction to Macroeconomics, Recitation Instructor for Prof. Luca Bossi Fall 2012 Intermediate Microeconomics, Lecturer Spring 2013 Intermediate Microeconomics, Recitation Instructor for Prof. Tymofiy Mylovanov Summer 2013 Intermediate Microeconomics, Lecturer Fall 2013 Public Finance, Teaching Assistant to Prof. Hanming Fang Spring 2014 Macroeconomic Modeling, Teaching Assistant to Prof. Dirk Krueger Research Experience and Other Employment: July 2014-May National Bureau of Economic Research, Research Assistant 2015 July 2011-May University of Pennsylvania, Research Assistant to Prof. Iourii Manovskii (2011), 2013 Research Assistant to Prof. Harold L. Cole (2012-2013), Research Assistant to Prof. John Yiran Zhu (2012) July 2012Bonn University (Institute for Macroeconomics and Econometrics, Visiting August 2012 Researcher August 2009 - Humboldt-Universität zu Berlin December 2009 Professional Activities: Presentations Jackson Hole Finance Conference, Wharton Finance, Penn Money Macro, Penn Macro Club, Philadelphia Fed (scheduled) Refereeing Journal of Monetary Economics, International Economic Review Honors, Scholarships, and Fellowships: 2015 Robert Summers Dissertation Fellowship in Economics (University of Pennsylvania) 2015 Lamfalussy Research Fellowship of the European Central Bank 2014 President Gutmann Leadership Award (University of Pennsylvania) Research Papers: “Macroeconomic Effects of Secondary Market Trading” (Job Market Paper) Starting around 1990, financial intermediaries in the United States increasingly began to sell, rather than hold to maturity, many of the loans that they provided to households and firms. This paper presents a theory in which the endogenous growth of such secondary market trading generates a macroeconomic credit cycle. Growing secondary markets initially boost credit volumes but gradually lead credit to flow to excessively risky investments. Aggregate risk exposure builds as asset quality falls. Ultimately, a negative shock leads to a simultaneous collapse of secondary markets and credit volumes -- as in the financial crisis of 2008. Booms are triggered by periods of low interest rates, and longer booms lead to sharper crises. Saving gluts and expansionary monetary policy thus lead to financial fragility over time. Procyclical regulation of secondary market traders, such as asset managers or hedge funds, can improve welfare even when such traders are not levered. “Debt Crises: For Whom the Bell Tolls” (with Harold L. Cole and Guillermo Ordonez) What a country has done in the past, and what other countries are doing in the present can feedback for good or for ill. We develop a simple model that can address hysteresis and contagion in sovereign debt markets. When a country’s fundamentals change, those changes affect information acquisition about that country and also affect the allocation of investment funds worldwide, inducing changes in the dynamics of sovereign spreads in seemingly unrelated countries. “Are Universal Banks Better Intermediaries?” (with Farzad Saidi) Revise and Resubmit, Journal of Financial Economics Are banks of wide scope better intermediaries? Using the variation in bank scope generated by the stepwise repeal of the Glass-Steagall Act in the U.S. and the sub- sequent rise of universal banking, we provide evidence that economies of scope in concurrent lending and underwriting improve the access to finance for risky ventures of publicly traded companies. Exploiting a bank-level deregulatory shock, as well as detailed data on bank-firm interactions, we identify increases in sales-growth, stock- return, and option-implied volatilities for universal-bank-financed firms. These firms also exhibit lasting increases of 3 to 4% in total factor productivity, 6 to 7% in capital expenditure, and 5 to 9% in market capitalization. Our findings suggest that the facilitation of cross-selling of loans and non-loan products may have led to an increase in the supply of credit for firms making risky, productivity-increasing investments. “Does Bank Scope Improve Monitoring Incentives in Syndicated Lending?” (with Farzad Saidi) We propose a model to study the provision of monitoring incentives in loan syndicates when banks differ in scope. Because bank scope increases a bank’s total exposure to firm performance beyond its loan share, banks of wide scope have incentives to monitor the firm even when they receive small loan shares. As such, they are more likely to be chosen as lead arrangers, yet receive comparatively small lead shares. We confirm these predictions empirically by exploiting the repeal of the Glass-Steagall Act. Our findings suggest that the observed increases in syndicated- loan volumes and simultaneous decreases in lead shares over the last two decades are not associated with losses in monitoring efficiency. Daniel Neuhann1 Dissertation Abstract I study the macroeconomics of financial markets, with a particular focus on financial institutions and settings with asymmetric information. I do so using theoretical and empirical tools. During my Ph.D., my research has focused on evaluating the macroeconomic consequences of two large structural changes in the U.S. financial system: the growing prevalence of securitization and secondary markets for financial assets, and the rise of universal banking after the repeal of the Glass-Steagall Act. I have also studied the international contagion of sovereign debt crises. In my job market paper, “Macroeconomic Effects of Secondary Market Trading,” I study the aggregate implications of a major structural change in the U.S. financial system over the last 25 years. From 1990 onwards, financial intermediaries in the U.S. increasingly began to sell, rather than hold to maturity, many of the loans that they originated. This development is significant in that intermediaries had held the majority of their loans on balance sheet for most of the postwar period. I present a theory in which the endogenous emergence of such secondary market trading generates a macroeconomic credit cycle in line with the build-up to the 2008 financial crisis. Growing secondary markets initially boost credit volumes but gradually lead credit to flow to excessively risky investments. Aggregate risk exposure builds as asset quality falls. Ultimately, a negative shock leads to a simultaneous collapse of secondary markets and credit volumes as in the financial crisis of 2008. In the theory, booms are triggered by periods of low interest rates, and longer booms lead to sharper crises. In the U.S., a dramatic inflow of capital from abroad and expansionary monetary policy combined to reduce interest rates in the early 2000s. My model therefore provides an argument for why secondary markets grow so sharply from then on, and how this grow led to a credit boom that ended in crisis. I then use the model to study policy. I find that pro-cyclical regulation of secondary market traders, such as asset managers or hedge funds, can improve welfare even when such traders are not levered. Moreover, capital requirements on bankers may lead asset quality to fall than in their absence. The second chapter of my dissertation is “Debt Crises: For Whom the Bell Tolls,” co-authored with Harold Cole and Guillermo Ordoñez. We provide a new approach to understanding the contagion of sovereign debt crises across countries. We do so in a model in which investors can privately acquire information about a country’s fundamentals, such as its growth prospects or political willingness to default. We show that small shocks to a country’s fundamentals can trigger information acquisition and lead to increased bond price volatility. We then show that increased volatility in one country may trigger information acquisition in other countries. The reason is that risk-averse investors are strongly averse to holding overvalued bonds in multiple countries at once, and are thus willing to pay more for the privilege of knowing the true fundamentals. Small shocks in a single country can therefore trigger increased volatility in other countries and generate contagious debt crises. Because information acquisition is socially inefficient, switching to such “informed” equilibria generates welfare losses. We argue that this mechanism is of relevance for the recent Eurozone crisis. 1 Department of Economics, University of Pennsylvania. neuhann@sas.upenn.edu 1 The last two chapters of my dissertation are co-authored with Farzad Saidi and analyze the impact of the repeal of the Glass-Steagall Act on the performance of bank-dependent firms. The repeal gave rise to universal banks, which are large financial institutions that offer both commercial and investment banking services under one roof. As a result, banks and firms were able to form lending relationships across a wider variety of financial products. In “Are Universal Banks Better Intermediaries?,” we provide empirical evidence that universal-bank financed firms exhibited lasting increases of 3 to 4% in total factor productivity (TFP), 6 to 7% in investment, and 5 to 9% in market capitalization, a proxy of firm value, after the repeal of Glass-Steagall. Universal banking also leads to increases of up to 14% in a benchmark measure of firm-level risk, namely the volatility of firm sales. This this risk is not reflected in higher default rates, however. We thus contribute to the macroeconomic literature that studies the run-up of idiosyncratic firm-level risk over the last thirty years. This paper is currently being revised for resubmission to the Journal of Financial Economics. In “Does Bank Scope Improve Monitoring Incentives in Syndicated Lending?,” we explore the mechanism by which bank scope matters for the performance of bank-dependent firms. We provide a model to argue that universal banks may be particularly efficient at monitoring firms because there are economies of scope across multiple financial products. To evaluate this hypothesis empirically, we study the market for syndicated loans, in which multiple banks jointly lend to a single firm. Syndicates typically feature a single lead arranger tasked with monitoring the firm. These lead arrangers tend to retain a large share of the loan to ensure that they have incentives to actually monitor the firm. Our hypothesis is that universal banks are more likely to be lead arrangers, and that they require lower loan shares to maintain monitoring incentives. We confirm both predictions empirically. In contrast to the existing literature, our findings suggest that the observed increases in syndicated-loan volumes and simultaneous decreases in lead shares over the last two decades are not necessarily associated with losses in monitoring efficiency. By emphasizing scope over size, these two papers on universal banking therefore present evidence of firm-level benefits of universal banking. 2 FRANCISCO SILVA https://economics.sas.upenn.edu/graduate-program/candidates/francisco-silva fsilva@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN @ ECON.UPENN.EDU 215-898-6880 215-898-7350 215-898-5691 Office Contact Information Department of Economics 160 McNeil Building, 3718 Locust Walk Philadelphia, PA 19104 Cell Phone: +1 2154214450 Undergraduate Studies: Bsc (Licenciatura), Economics, University of Porto, 2008 Masters’ Level Work: Msc, Economics, Catholic University of Portugal, 2010 Graduate Studies: University of Pennsylvania, 2010 to present Thesis Title: “Essays in Applied Microeconomic Theory” Expected Completion Date: May 2016 Thesis Committee and References: Professor Andrew Postlewaite (Advisor) Department of Economics, 3718 Locust Walk Philadelphia, PA 19104 215-898-7350, apostlew@econ.upenn.edu Professor Steven Matthews Department of Economics, 3718 Locust Walk Philadelphia, PA 19104 215-898-7749, stevenma@econ.upenn.edu Professor Rakesh Vohra Department of Economics, 3718 Locust Walk Philadelphia, PA 19104 215-898-6777, rvohra@seas.upenn.edu Teaching and Research Fields: Primary fields: Applied Microeconomic Theory, Public Economics, Law and Economics Secondary fields: Behavioral Economics Teaching Experience: Fall, 2013-2014 Summer, 2013-2014 Spring, 2012-2013 and Fall, 2001 Fall, 2002 Introduction to Micro and Macro Economics and its applications, UPenn, Head Recitation Instructor for Professor Gizem Saka Economic Analysis of the Public Sector (MPA course), UPenn, Teaching Assistant for Professor David Crawford Introduction to Macroeconomics, UPenn, Recitation Instructor for Professor Luca Bossi Introduction to Micro and Macro Economics and its applications, UPenn, Recitation Instructor for Professor Gizem Saka Summer, 2002 Spring, 2009-2010 Fall, 2008-2009 Intermediate Microeconomics, UPenn, Instructor Intermediate Microeconomics, Catholic University of Portugal, Teaching Assistant for Professor Fernando Branco Intermediate Microeconomics, Catholic University of Portugal, Teaching Assistant for Professor Fernando Machado Research Experience and Other Employment: 2009-2010 Catholic University of Portugal, Research Assistant for Professors Leonor Modesto and Teresa Lloyd Braga 2009 Ministry of Economics and Innovation (Portugal), Economics Studies Office, Research Intern Honors, Scholarships, and Fellowships: 2014 Edward Mansfield Prize for best performance of a recitation instructor in Intro Economics, Honorable Mention, UPenn 2010-2015 University Fellowship, UPenn Research Papers “The Optimal Design of a Criminal Justice System” (Job Market Paper) I consider the problem a social planner faces in constructing a criminal justice system which addresses two needs: to protect the innocent and to punish the guilty. I characterize the socially optimal criminal justice system under various assumptions with respect to the social planner's ability to commit. In the optimal system, before a criminal investigation is initiated, all members of the community are given the opportunity to confess to having committed the crime in exchange for a smaller than socially optimal punishment which is independent of any future evidence that might be discovered. Agents who choose not to confess might be punished once the investigation is completed if the evidence gathered is sufficiently incriminatory. In this paper's framework, leniency for confessing agents is efficient not because it saves resources or reduces risk, but because there are informational externalities to each confession. When an agent credibly confesses to be guilty he indirectly provides the social planner additional information about the other agents: the fact that they are likely to be innocent. “Inducing Overconfidence” (R&R at Economic Inquiry) In this paper, I show how the intervention of others (parents, bosses or coworkers) may generate overconfidence. I compare two explanations. The first is that people take their success at face value and disregard the influence that outside help may have. Even though this assumption may sometimes be reasonable, I argue that it is not when such help is received systematically as is the case, for example, of parenting. Hence, I provide a more novel explanation for induced overconfidence that is based on the idea that receiving help makes the information gathered less precise. Overconfident people, who tend to receive more help, will remain overconfident as the future information they will gather has very little precision. On the contrary, underconfident people, who tend to receive less help, will learn their true ability faster. Using data on teenagers and young adults I compare these two alternative explanations. “Should the government provide public goods if it cannot commit?” (R&R at Journal of Public Economic Theory) I compare two different systems of provision of discrete public goods: a centralized system, ruled by a benevolent dictator who has no commitment power; and an anarchic system, based on voluntary contributions, where there is no ruler. If the public good is binary, then the public good provision problem is merely an informational one. In this environment, I show that the anarchic system can always replicate any outcome of the centralized system. However, as one increases the number of alternatives available, the classical free riding problem described in Samuelson (1954) emerges. As the classical free riding problem becomes more important relative to the informational free riding problem, the centralized system becomes the preferred system of the two. “Euthanasia: the fear of becoming a burden” It has been widely documented in the medical literature that terminal patients often choose to end their lives out of the fear of becoming a burden to their family. But the real size of the burden, in particular its emotional component, is only known by the patient's family. I analyze the impact of legalizing euthanasia on the ability of the family to communicate with the patient through a cheap talk model. I argue that, if euthanasia is legalized, either the patient makes his decision (of whether or not to commit suicide) uninformed of the size of the burden; or chooses what his family would have chosen if it had the power, and not necessarily what he would have preferred. I also consider the role of the physician and argue that, if the physician anticipates the family's influence in the patient's decision, providing incentives for the physician's interests to be aligned with the patient's might not be on the patient's best interest. Languages: Portuguese (native), English (fluent), Spanish (intermediate), French (basic) Dissertation Abstract Essays in Applied Microeconomic Theory Francisco Silva My dissertation is composed of four papers. Paper 1, my job market paper, studies the design of efficient criminal justice systems using some of the techniques of mechanism design. Paper 2 proposes a new explanation for the phenomenon of overconfidence. Paper 3 re-examines the question of whether the government should provide public goods while assuming it cannot commit. Paper 4 investigates the consequences of the legalization of euthanasia in the communication between patient, patient's family and physician using a cheap talk model. "The Optimal Design of a Criminal Justice System" (Job Market Paper) In this paper, I investigate the desirable properties of a criminal justice system. I approach this problem by analyzing a simple scenario that I believe illustrates the main challenge a criminal justice system faces. Consider a community of N agents and a principal. Imagine that the suspicion that a crime has been committed arises and it is the principal's responsibility to select punishments to be inflicted upon the agents. Think of the principal as a sort of social planner or benevolent decision maker who wants the best for the community. In a perfect world, she would like to punish only agents who are guilty of committing the crime. Of course, the problem is that the principal does not know who is guilty and who is innocent. And, knowing that the principal is interested in punishing those agents that are guilty makes them reluctant to announce their guilt. I study the principal's problem of creating a mechanism that, to the extent that is possible, punishes those who are guilty while protecting the rights of the innocent. There are two particular ways to solve this problem - two systems - which are important for my analysis. The first is what I call a "trial system". In a trial system, if the principal suspects the crime has been committed she initiates a police investigation aimed at obtaining evidence. Based on the evidence, the principal forms beliefs about the guilt of each agent and chooses punishments accordingly. Only agents whose evidence strongly indicates guilt are punished - agents are punished if they are found to be guilty beyond "reasonable doubt". The second system is what I call a "confession inducing system" (CIS). A CIS has two stages. In the first stage, before the investigation begins, all agents are given the opportunity to confess the crime, in exchange for a punishment independent of any evidence which might be gathered in the future. In the second stage, if necessary, the principal conducts a police investigation, and, based on the information gathered, chooses the punishments, if any, to apply to agents who chose not to confess in the first stage. It essentially is a trial system preceded by a confession stage. An example of these type of systems is “self-reporting” in environmental law where firms which infringe environmental regulations are able to contact the corresponding law enforcement authority and self-report this infringement in exchange for a reduced punishment. The main result of the paper is that, no matter what the principal’s commitment power may be, it is possible to construct a CIS that is preferred to any other system. The appeal of the optimal CIS is that it explores the Dissertation Abstract Essays in Applied Microeconomic Theory Francisco Silva correlation between the agents' innocence in that, when an agent credibly confesses to be guilty, he is also indirectly informing the principal that everyone else is likely to be innocent. "Inducing Overconfidence" (R&R at Economic Inquiry) In this paper, I show how the intervention of others (parents, bosses or coworkers) may generate overconfidence. I compare two explanations. The first is that people take their success at face value and disregard the influence that outside help may have. Even though this assumption may sometimes be reasonable, I argue that it is not when such help is received systematically as is the case, for example, of parenting. Hence, I provide a more novel explanation for induced overconfidence that is based on the idea that receiving help makes the information gathered less precise. Overconfident people, who tend to receive more help, will remain overconfident as the future information they will gather has very little precision. On the contrary, underconfident people, who tend to receive less help, will learn their true ability faster. Using data on teenagers and young adults I compare these two alternative explanations. "Should the government provide public goods if it cannot commit?" I compare two different systems of provision of discrete public goods: a centralized system, ruled by a benevolent dictator who has no commitment power; and an anarchic system, based on voluntary contributions, where there is no ruler. If the public good is binary, then the public good provision problem is merely an informational one. In this environment, I show that the anarchic system can always replicate any outcome of the centralized system. However, as one increases the number of alternatives available, the classical free riding problem described in Samuelson (1954) emerges. As the classical free riding problem becomes more important relative to the informational free riding problem, the centralized system becomes the preferred system of the two. "Euthanasia: the fear of becoming a burden" It has been widely documented in the medical literature that terminal patients often choose to end their lives out of the fear of becoming a burden to their family. But the real size of the burden, in particular its emotional component, is only known by the patient's family. I analyze the impact of legalizing euthanasia on the ability of the family to communicate with the patient through a cheap talk model. I argue that, if euthanasia is legalized, either the patient makes his decision (of whether or not to commit suicide) uninformed of the size of the burden; or chooses what his family would have chosen if it had the power, and not necessarily what he would have preferred. I also consider the role of the physician and argue that, if the physician anticipates the family's influence in the patient's decision, providing incentives for the physician's interests to be aligned with the patient's might not be on the patient's best interest. VESA-HEIKKI SOINI http://economics.sas.upenn.edu/graduate-program/candidates/vesa-heikki-soini soini@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN Office Contact Information 417 McNeil Building, 3718 Locust Walk Philadelphia, PA 19143 Cell phone number: 267-252-4408 @ ECON.UPENN.EDU 215-898-6880 215-898-7350 215-898-5691 Home Contact Information 1315 S Chadwick St Philadelphia, PA 19146 Home phone number: 267-252-4408 Personal Information: Male, Citizen of Finland Undergraduate Studies: B.Sc & M.Sc, Economics, Helsinki School of Economics, 2006 Masters Level Work: M.Phil, Economics, University of Oxford, 2008 Graduate Studies: University of Pennsylvania, 2010 to present Thesis Title: “Academic Specialization and Misallocation of Skills in the Labor Market” Expected Completion Date: May 2016 Thesis Committee and References: Professor Guillermo Ordonez (Primary Advisor) Room 428 McNeil Building, 3718 Locust Walk 215-898-1875 ordonez@econ.upenn.edu Professor Harold Cole Room 436 McNeil Building, 3718 Locust Walk 215-898-7788 colehl@econ.upenn.edu Professor Dirk Krueger Room 428 McNeil Building, 3718 Locust Walk 215-573-1424 dkrueger@econ.upenn.edu Teaching and Research Fields: Primary field: Macroeconomics Secondary fields: Financial Economics, Labor Economics Teaching Experience: Spring, 2015 Money Credit and Banking, University of Pennsylvania, teaching assistant for Prof. Harold Cole Spring, 2014 Intermediate Macroeconomics, University of Pennsylvania, teaching assistant for Prof. Ufuk Akcigit Fall, 2013 Introduction to Microeconomics, University of Pennsylvania, grader for Prof. Rebecca Stein Summer, 2013 Introduction to Macroeconomics, University of Pennsylvania, instructor Spring, 2013 Macro Modelling, University of Pennsylvania, teaching assistant for Prof. Guillermo Ordonez Spring, 2013 Law and Economics, University of Pennsylvania, teaching assistant for Prof. Camilo GarciaJimeno Fall, 2012 Introduction to Economics, Wharton Business School, recitation instructor for Prof. Gizem Saka Spring, 2012 Fall, 2011 Intermediate Macroeconomics, University of Pennsylvania, teaching assistant for Prof. Ufuk Akcigit Introduction to Economics, Wharton Business School, recitation instructor for Prof. Uri Spiegel Research Experience and Other Employment: 2007-2008 Said Business School, University of Oxford, Research Assistant for Prof. Tarun Ramadorai 2006-2007 Nokia Corporation, Financial Analyst 2006 Nokia Networks, Business Analyst 2005-2006 Nokia Networks, Business Analyst Trainee Professional Activities, 2012-2015 Penn Macro Workshop Seminar 2014 Reading Group In Liquidity and Financial Crisis, Wharton Business School Honors, Scholarships, and Fellowships: 2010-2011 Barbara and Edward Netter “Thanks to Scandinavia” Scholarship 2010 Yrjo Jahnsson Foundation Scholarship 2008 Otto A. Malmin Lahjoitusrahasto 2007 Yrjo Jahnsson Foundation Scholarship Research Papers: “Academic Specialization and Misallocation of Skills in the Labor Market” (Job Market Paper) This paper studies academic specialization and misallocation of skills in the labor market. I combine a discrete choice model with a general equilibrium macro model to study occupations where occupation-specific human capital is obtained through university education and people incur considerable upfront costs to work in a particular occupation. The model embeds a market failure: risk-averse individuals face an incomplete markets problem because they are not able to buy insurance against adverse occupation-specific shocks. I compare production efficiency and utilitarian welfare in competitive equilibrium to the outcomes of two social planning problems: (i) output maximization problem (ii) “constrained efficient” planning problem. To get quantitative estimates of the importance of academic specialization, I calibrate the model using data on petroleum, chemical and mechanical engineers. The output loss caused by the lack of insurance depends on model parameters and can potentially be very large. “Team Formation and Learning about Worker-Specific Productivities” The objective of this paper is to study matching of workers into teams within a firm when learning about workerspecific productivity is important. The paper introduces a new Bayesian conjugate prior approach where individuals' contributions are normally distributed but only joint production is observed. This framework can be applied to any problem where a sum of random variables is observed. In this paper, the framework is used to study whether expected continuation values affected by learning are supermodular in expected productivity (prior means) and variances of expected productivity (prior variances). “The Existence of a Unique Solution to the Bellman Equation for Unbounded Return Functions” “Estimating the Gravity Model with Differentiated and Homogeneous Goods – Evidence for the Home-Market Effect?” (Master's Thesis) “The Effect of the Euro on Intra-European Trade” (Undergraduate Thesis) Languages: Finnish (native), English, Swedish Research Statement Vesa Soini My research interests lie in macroeconomics with secondary interests in financial and labor economics. My job market paper combines an occupational choice model with a general equilibrium macro model and studies the effect of alternative labor allocations on output and utilitarian welfare. The focus is on occupations where occupation-specific human capital is obtained through university education and people incur considerable upfront costs to be able to work in a particular occupation. Such occupations include many technical, medical and legal occupations, for example. The model embeds a market failure: risk-averse individuals face an incomplete markets problem because they are not able to purchase insurance against adverse occupation-specific shocks. One ambition of my paper is to develop a tractable model for studying general equilibrium wage movements of these occupations. In my model, initial endowments depend on occupational choices. A social planner is able to manipulate initial endowments by choosing alternative allocations which affects both output and wage inequality in the economy. I solve for two planner’s problems: (i) the output maximizing problem (ii) the ‘constrained efficient’ problem. In the latter problem, the planner recognizes that changing allocations will affect ‘pecuniary externalities’, i.e. externalities transferred through general equilibrium wages. Because markets are incomplete, movements in general equilibrium wages will have a welfare effect. The constrained planner effectively seeks to mitigate inequality across occupations even though he cannot eliminate inequality within occupations. The main ambition of my job market paper is to quantitatively study the output loss caused by the market failure in occupational choices. For this purpose I develop a dynamic version of the model and solve it numerically. I calibrate the model using data on petroleum, chemical and mechanical engineers. The results show that the output loss caused by the market failure can be very large (baseline estimate is 3.7% of output) although the exact magnitude depends on parameter values. The paper assumes that individuals have heterogeneous productivity in each occupation. I assume that each worker’s idiosyncratic productivity for each occupation is drawn from a Fréchet distribution. Therefore, the model embeds the idea that labor is not perfectly substitutable. If more and more workers are needed in a particular occupation, it becomes necessary to attract people who are less talented in that occupation. If these people were more talented in other occupations, there is a loss of total efficiency units of labor. In fact, total labor units are maximized when there is some amount of people going to each occupation based on people’s inherent comparative advantages. I start by studying a one-period version of the model. I derive closed-form solutions for labor allocations, output, utilitarian welfare and optimal taxes for competitive equilibrium and two social planning problems: (i) the output maximization problem (ii) the ‘constrained efficient’ problem. The first of these planning problems yields maximal output by definition but assumes that the planner can transfer income between workers after occupation-specific shocks are realized. The second social planning problem does not allow such transfers. Since the model suffers from an incomplete markets problem, the first welfare theorem does not apply. This means that pecuniary externalities do not offset each other in competitive equilibrium and the outcome is not Pareto efficient. Occupation-specific taxes or education quotas can in theory be used to improve labor allocation in the economy. 1 In my model, workers internalize the mean but not the variance of wages. At the same time, the output losses depend on both the means and variances of wages. To maximize output, the planner wants to allocate more people to more risky occupations. On the contrary, the constrained efficient allocation requires that less people are allocated to more risky occupations. The reason for this is that less people in those occupations increases general equilibrium wages. This increase compensates for the additional risk of the occupation and therefore the expected utilities across various occupations becomes closer to each other. However, there is still wage inequality within occupations because people are paid according to their idiosyncratic productivity. The output losses depend on certain parameter values. If the elasticity of substitution between occupation goods is low, the wage risk becomes less severe since general equilibrium prices increase when output is decreased. This mechanism keeps occupation revenues and also wages relatively stable. The capital share of an occupation increases output losses, as a high capital share amplifies wage movements. Finally, the wage risk depends on the variance of the underlying Bernoulli distribution. In the dynamic version of the model, the state of each occupation changes over time. The main purpose of this section is to get an estimate of the magnitude of potential output losses. The uncertainty is modeled as occupationspecific business cycles where the productivity realizations depend on the hidden state of each occupation. Workers learn about the hidden states over time but the hidden states change according to a Markov process. I calibrate the model using data on petroleum, chemical and mechanical engineers. The output loss caused by the lack of insurance is estimated to be about 3.7% of the GDP. However, this is just an aggregate result. Depending on the chosen occupation there will be winners and losers. An important dynamic result is that the competitive equilibrium features too much variation in occupational choices over time compared to the planner’s outcome. Therefore the model exhibits ‘dynamic misallocation’. If uncertainty is more severe in a downturn, risk-aversion makes individuals ‘over-react’ to perceived changes of the occupation’s state. As a consequence, some people whose comparative advantage lies in another occupation will choose an occupation where they are less productive which reduces total efficiency units of labor. The output maximizing planner takes such possibilities into account in his solution. My job market paper raises several new research questions. For one, the stationarity of the model helps with computations but is somewhat unrealistic. In reality new occupations are born and old occupations die over time. The model could be modified to account for this by choosing a Bayesian process in which new occupations are more uncertain than the established ones but as workers learn about the new occupations over time, the uncertainty disappears. This modification of the model seems appropriate for modeling many interesting occupations in the economy. Secondly, while the utilitarian welfare criteria used in the model is a constrained efficient allocation, the modeling of occupational choice creates a subtle special feature to the general equilibrium theory. When individuals choose a particular occupation, they choose not to go to any other occupations. This implies that they will ‘throw away’ their labor endowments for those other occupations. One research question would be to try to find all Pareto efficient allocations after the idiosyncratic productivity realizations are drawn but before the occupation-specific TFP shocks are realized. In my third year paper, the objective was to study matching of workers into teams within a firm when learning about worker-specific productivity is important. The main novelty in that paper is the Bayesian learning process. That is, I develop a Bayesian conjugate prior approach where individuals’ outputs are normally distributed but only joint production is observed, and the firm owner can choose the fraction of the task performed by each worker. I show in this paper that the posterior is always a bi-variate Normal. Such a learning process implies that if tasks among workers are equally distributed, the firm owner only learns about the average productivity of the team rather than worker-specific productivity. As shown in the paper, when tasks are equally distributed, the posterior covariance will 2 be negative. This implies that if only one worker performs the task in the next period, Bayesian updating can also be applied to other workers even if they did not work on the task. If on the other hand, more of the task is assigned to one worker, the learning process approaches the standard Normal-Normal conjugate process. This implies that the firm owner might find it optimal to first assign the tasks equally but later on give more responsibility to one worker, since the speed of learning will be increased after the previous period’s updating. In principle, this framework can be applied to any problem where a sum of random variables is observed. In my third year paper, a simplified version of the framework was applied to study whether expected continuation values affected by learning are supermodular with respect to expected productivity and the variance of expected productivity. 3 YU WANG https://economics.sas.upenn.edu/graduate-program/candidates/yu-wang wangyu5@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ ECON. UPENN. EDU APOSTLEW@ ECON. UPENN.EDU KQUINN @ ECON.UPENN.EDU 215-573-2057 215-898-7350 215-898-5691 Office Contact Information 160 McNeil Building 3718 Locust Walk Philadelphia, PA 19104 215-919-1358 Personal Information: August 10, 1987, Female, Chinese Citizen, F-1 Visa Undergraduate Studies: B.A. in Finance, Peking University, 2005-2009 Graduate Studies: University of Pennsylvania, 2009 to present Thesis Title: “The Impact of Student Debt on the Education, Career, and Marriage Choices of Female Lawyers” Expected Completion Date: May 2016 Thesis Committee and References: Professor Holger Sieg (Main Advisor) Department of Economics 3718 Locust Walk, RM 459 Philadelphia, PA 19104 215-898-7194 holgers@econ.upenn.edu Professor Kenneth Wolpin Rice University Department of Economics-MS 22 P.O. Box 1892 Houston, TX 713-348-2522 Kenneth.I.Wolpin@rice.edu Professor Hanming Fang Department of Economics 3718 Locust Walk, RM 429 Philadelphia, PA 19104 215-898-7767 hanming.fang@econ.upenn.edu Teaching and Research Fields: Labor, Public, Political Economy, Education Teaching Experience: Fall 2010 Econometrics (Ph.D. 1st year course) Teaching Assistant for Prof. Kevin Song Spring 2011 Statistics for Economists, Teaching Assistant for Prof. Aureo De Paula Fall 2011 Econometrics (Ph.D. 1st year course), Teaching Assistant for Prof. Frank Schorfheide and Prof. Xu Cheng Spring 2012 Econometrics, Teaching Assistant for Prof. Xu Cheng Fall 2012 Fall 2012 Spring 2013 Summer 2013 Fall 2013 Spring 2014 Urban Fiscal Policies, Teaching Assistant for Prof. Holger Sieg Development Economics, Teaching Assistant for Prof. Flavio Cunha Labor Economics, Instructor Econometrics, Instructor Econometrics (Ph.D. 2nd year course), Teaching Assistant for Prof. Petra Todd Econometrics, Teaching Assistant for Prof. Frank Schorfheide Research Experience: Fall 2014 Research Assistant for Prof. Gilles Duranton Professional Activities: Presentation: Spring 2012 U Penn Empirical Micro Lunch Summer 2014 The North American Summer Meeting of the Econometric Society Spring 2015 U Penn Empirical Micro Lunch Summer 2015 EconCon Fall 2015 U Penn Empirical Micro Workshop Referee: International Economic Review (x3) Other: 2011-2012 Organizer, University of Pennsylvania Empirical Micro Lunch Club Honors, Scholarships, and Fellowships: 2014 GPASA and SAS Travel Grant, University of Pennsylvania 2005-2009 Benz Scholarship, Peking University 2006-2007 Zhongji Group Scholarship, Peking University 2006-2007 Dean’s List, Hong Kong University of Science and Technology Job Market Paper: “The Impact of Student Debt on the Education, Career, and Marriage Choices of Female Lawyers” Abstract: Obtaining a J.D. is one of the most expensive possible investments in human capital. As a consequence, law students quite often take on substantial amounts of debt to finance their graduate education. There has been much concern in the legal profession and among policy makers that this debt burden distorts career choices. The purpose of this paper is to study the impact of student debt on the education, career, and marriage choices of female lawyers. The empirical analysis is based on a novel, nationally representative, longitudinal dataset. In contrast to the previous literature that has largely focused on males and finds debt has only small effects, these new data suggest that debt has large and significant negative effects on female career and marriage outcomes. Specifically, women with more debt stay longer in private sector jobs, postpone marriage, marry men with lower earnings, and delay childbearing. To explore the likely causes of these negative debt effects, I develop and estimate a dynamic model of education, labor, and marriage markets. My model accounts for a variety of special features of the labor markets for lawyers, such as the large differences between public and private sector jobs, the considerable uncertainty faced by lawyers regarding promotion to partner, and the need for updating beliefs about potential for success in private law firms. My findings suggest that a large part of the debt effect on schooling and career choices comes from the diminished marriage prospects associated with the debt burden. I then focus on policies that aim to reduce the debt burden while also encouraging female lawyers to pursue careers in the public sector. My policy experiments show that subsidizing student debt repayment earlier in the career is more effective than doing so later. Publications: “The Impact of Unions on Municipal Elections and Urban Fiscal Policies”, with Holger Sieg. Journal of Monetary Economics, 2013, 60(5), 554-567 Abstract: The efficient decentralized provision of public goods requires that special interest groups, such as municipal unions, do not exercise undue influence on the outcome of municipal elections and local fiscal policies. We develop a new political economy model in which a union can endorse one of the candidates in a local election. A politician that prefers an inefficiently large public sector can, therefore, win an election if the union can provide sufficiently strong support during the campaign. We have assembled a unique data set that is based on union endorsements that are published in leading local newspapers. Our empirical analysis focuses on municipal elections in the 150 largest cities in the U.S. between 1990 and 2012. We find that challengers strongly benefit from endorsements in competitive elections. Challengers that receive union endorsements and successfully defeat an incumbent also tend to adopt more union friendly fiscal policies. Research Papers in Progress: “Paved with Good Intentions? The Unintended General Equilibrium Effect of Loan Repayment Assistance Programs on Tuition” Abstract: This paper studies the general equilibrium effects of loan repayment assistance programs on tuition. A prevalent concern is that loan repayment assistance programs have fueled the recent rise in tuition well beyond inflation. Law schools compete for high-ability students by improving education quality, which depends on the amount of monetary investment. Programs subsidizing repayment may encourage borrowing and motivate students to select better and often more expensive schools. This in turn stimulates schools to raise tuition to increase monetary resources. I develop a new general equilibrium model of the legal education market and the lawyer labor market that captures the large degree of quality differentiation among law schools, their tuition and admission policies, schools' monetary investment in education quality, and the return to these investments in the labor market. This work in progress estimates this model using a variety of datasets on the law schools' admission rules, tuition, and school characteristics. “Participation in Municipal Electoral Competition in the U.S.” (with Holger Sieg) Abstract: The purpose of this research is to improve our understanding of individual decisions to run for local public office. While there has been some research focused on the career decisions of politicians that compete for state or federal office, much less is known about local elections. Participation decisions of incumbents are likely driven by the quality of the available outside options. In contrast, participation decisions of challengers are not only influenced by qualifications and skills, but also depend on the incumbency status and the potential vulnerability of the incumbent. To conduct our empirical analysis, we have assembled a data set that covers all major candidates in 128 elections in the largest 25 U.S. cities from 1990 to 2012. A key challenge of the data collection process is to characterize pre-election and post-election career choices. Our preliminary findings suggest that 54 percent of challengers in open elections are elected officials. The remaining challengers are recruited from a pool that consists of appointed politicians (15 percent), public sector employees (7 percent), and private sector workers (24 percent). There are distinct observed differences in skills and qualifications. Challengers are less likely to have law degrees than incumbents, but they are more likely to hold Master’s degrees. Challengers are also less likely to have a recognizable politician in the immediate family. Computer Skills: Stata, Matlab, Fortran, Parallel Programming Language: Chinese, English “The Impact of Student Debt on the Education, Career, and Marriage Choices of Female Lawyers” Yu Wang University of Pennsylvania Obtaining a J.D. is one of the most expensive possible investments in human capital. As a consequence, law students quite often take on substantial amounts of debt to finance their graduate education. There has been much concern in the legal profession and among policy makers that this debt burden distorts career choices. The purpose of this paper is to study the impact of student debt on the education, career, and marriage choices of female lawyers. The empirical analysis is based on a novel, nationally representative, longitudinal dataset. In contrast to the previous literature that has focused largely on males and finds only small effects, these new data suggest that debt has large and significant negative effects on female career and marriage outcomes. Specifically, women with more debt stay longer in private sector jobs, postpone marriage, marry men with lower earnings, and delay childbearing. These facts speak to a long-standing concern of the legal profession: does high student debt impel law graduates to eschew public service jobs in favor of more lucrative positions in private practices? The answer has important policy implications regarding the launch and design of public service loan forgiveness programs. Much of the previous literature suggests that the career choices of male lawyers are not related to student debt, and that the effects of such policies are not pronounced. However, these policies may effectively increase public service employment of females, who account for nearly half of a typical cohort. Another part of the research question, namely how potential debt burden affects schooling choices, is particularly relevant in the legal profession. Law school costs rise substantially with school rankings and induce a palpable quality-price tradeoff for applicants. If the debt burden affects female choices postgraduation, the top law schools may be even more costly for females who resort to student debt in order to finance their education. This paper analyzes the extent to which students of equal ability but different means may choose lower-ranked, cheaper law schools in order to avoid the debt burden. To address these questions, I develop a dynamic model of the education, labor, and marriage markets. Students first need to choose among a set of differentiated law schools to some of which they are admitted. Schools differ by costs and rankings. Upon graduation all students enter the labor market which offers a number of occupational tracks such as private law firms, private companies, and public sector jobs, in addition to the option to stay at home. At the same time, female lawyers are active in marriage markets. Potential marriages depend flexibly on one's debt burden and individual characteristics. Labor markets for lawyers have some special features that deserve attention. First, young lawyers employed at private law firms often face an up-or-out decision regarding promotion to partner. Promotion is highly uncertain. This system is similar to that of tenure in academia. A key feature of the model is thus a learning process which allows young lawyers to infer the likelihood of promotion based on their work experience. I show that a learning model can explain the rapidly declining prevalence of private law firm jobs during the life cycle as observed in the data. Second, this paper models differences in the required number of hours of work in addition to differences in salaries among occupational tracks. This approach is consistent with the fact that lawyers in private law firms tend to work longer hours than those in the public sector or at private companies. Finally, this model accounts for the importance of clerkships for young, aspiring lawyers who may delay entry into the job market so as to gain additional training. The objective of the structural analysis is to evaluate the importance of alternative mechanisms that may explain why debt has such a large impact on female lawyers. In the model, debt affects the lifetime budget constraint and thus primarily influences consumption and labor supply choices over time. In addition, debt affects prospects in the marriage market. The key empirical finding of this paper is that differences in female career choices are driven in large part by marriage market prospects. If one equalizes marriage market prospects for females with and without debt, the differences in labor market choice shrink substantially. The intuition is that females with large debt have fewer opportunities in the marriage market, and thereby experience poorer marriage outcomes. Consequently, it takes them longer to meet a qualified spouse and to have children. The second finding is that females significantly under-invest in education quality in anticipation of the diminished marriage market prospects associated with their future debt burden. Using the estimated parameters from the structural analysis, I finally turn to evaluating loan repayment assistance programs. My main contribution here is to show the importance of timing for such policies. The motivation is that females, on average, graduate at age 28 and, as discussed above, the debt effect through marriage prospects is large. In light of this, the existing Public Service Loan Forgiveness Program offered by the Department of Education, which requires ten years of public service for eligibility for loan forgiveness, offers potential room for improvement. I compare a loan forgiveness policy similar to the Public Service Loan Forgiveness Program to a career-contingent tuition waiver policy. The tuition waiver policy conditionally discharges a fraction of debt immediately upon graduation, which is repayable if a job in the public service is not held for a specified number of years. The main finding is that it is much more effective to subsidize debt sooner rather than later. CHUNZAN WU https://economics.sas.upenn.edu/graduate-program/candidates/chunzan-wu chunzan@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN @ ECON.UPENN.EDU 215-898-6880 215-898-7350 215-898-5691 Office Contact Information 160 McNeil Building, 3718 Locust Walk Philadelphia, PA 19104 Phone: 530-219-7231 Personal Information Citizenship: China Undergraduate Studies B.S., Chemistry, Peking University, 2004 B.A., Economics, Peking University, 2004 Masters Level Work M.A., Economics, Peking University, 2007 Graduate Studies University of California, Davis, 2009 to 2011 University of Pennsylvania, 2011 to present Thesis Title: “Macroeconomic and Fiscal Policy Implications of Household Labor Supply” Expected Completion Date: June 2016 Thesis Committee and References: Professor Dirk Krueger (Primary Advisor) Department of Economics, University of Pennsylvania, 3718 Locust Walk, Philadelphia, PA 19104, USA Phone: (215) 573-1424 Email: dkrueger@econ.upenn.edu Professor Harold L. Cole Department of Economics, University of Pennsylvania, 3718 Locust Walk, Philadelphia, PA 19104, USA Phone: (215) 898-7788 Email: colehl@sas.upenn.edu Professor Guido Menzio Department of Economics, University of Pennsylvania, 3718 Locust Walk, Philadelphia, PA 19104, USA Phone: (215) 898-5170 Email: gmenzio@sas.upenn.edu Teaching and Research Fields Primary fields: Macroeconomics Secondary fields: Public Finance, Labor Economics, Computational Economics Teaching Experience Ph.D. Level Spring, 2014 Macroeconomic Theory I , University of Pennsylvania, Teaching Assistant for Spring, 2013 Professor Dirk Krueger And Professor Jesus Fernandez-Villaverde Undergraduate Level Fall, 2015 Macroeconomic Theory, University of Pennsylvania, Teaching Assistant for Professor Dirk Krueger Spring, 2015 Macroeconomic Theory, University of Pennsylvania, Teaching Assistant for Professor Guido Menzio Fall, 2014 Macroeconomic Theory, University of Pennsylvania, Teaching Assistant for Professor Guillermo Ordonez Fall, 2013 Statistics for Economists ,University of Pennsylvania, Teaching Assistant for Professor Francis J. DiTraglia Fall, 2012 International Trade, University of Pennsylvania, Teaching Assistant for Professor Wilfred J. Ethier Research Experience and Other Employment Fall, 2013 University of Pennsylvania, Research Assistant for Professor Mathieu Taschereau-Dumouchel Spring, 2013 University of Pennsylvania, Research Assistant for Professor Iourii Manovskii Honors, Scholarships, and Fellowships 2015 Joel Popkin Graduate Student Teaching Prize in Economics for outstanding teaching performance by a teaching assistant in economics, University of Pennsylvania 2012 Lawrence Robbins Prize in Economics to the student judged to be the best in the first year class, University of Pennsylvania Research Papers: “More Unequal Income But Less Progressive Taxation: Economics or Politics?” (Job Market Paper) Abstract: Since the 1970s in the US, income inequality has increased sharply. During the same time span, the US federal income tax has become less progressive. Why? I examine this question in a Ramsey optimal tax policy framework. Within this framework, the tax policy is determined by: (1) a set of Pareto weights representing the government's preference over different households; (2) household lifetime utilities encompassing the effects of economic fundamentals. I first quantify the changes of economic fundamentals using an overlapping generations incomplete-markets life-cycle model with heterogeneous households. The model features both endogenous human capital accumulation and household labor supply, and is calibrated to match the major changes of the labor income structure and demographics in the US between 1970s and 2010s. Then I use this economic model to determine whether the change in taxes is the result of changing economic fundamentals or the consequence of a change in Pareto weights. I interpret the later as changes in the political powers of various income groups. I find: (1) The changes of economic fundamentals alone induce a less progressive optimal income tax, and can account for 40% of the reduction in progressivity in the data. In particular, skill-biased technological change, increased female labor productivity, and the aging US population contribute to this reduction in progressivity, but their effects are partially offset by the increase of idiosyncratic income risks. (2) The change of Pareto weights is in favor of the high income households, and imply less valued government services. Finally, using a stylized political economy model, I discuss potential explanations for this change of Pareto weights such as lower cost of conveying information to swing voters and the rising inequality of voter turnout among different socioeconomic groups. “How Much Consumption Insurance in Bewley Models with Endogenous Family Labor Supply?” (with Dirk Krueger) Abstract: We show that a calibrated life-cycle two-earner household model with endogenous labor supply can rationalize the extent of consumption insurance against wage shocks estimated empirically by Blundell, Pistaferri, and Saporta-Eksten (2014) (BPS hereafter) in US data. With additive separable preferences, the model can account for about 94% and 91% of consumption insurance against male and female permanent wage shocks in the data, and only 41% of male and 28% of female permanent wage shocks in the model pass through to household consumption. With non-separable preferences, more consumption insurance is generated, and the pass-through rates are 27% and 18%, respectively. Most notably, the majority of the consumption insurance against permanent male wage shocks is provided through the endogenous labor supply response of the female earner. We also evaluate, using modelsimulated data, whether the empirical approach of BPS delivers unbiased consumption responses to wage shocks. We find that the method overestimates the amount of consumption insurance against male permanent wage shocks, and underestimates that against female permanent wage shocks, but only moderately so. We find larger biases for the outside insurance coefficient which BPS use to capture the insurance provided through channels outside their model. We document that the magnitudes of the biases are not sensitive to the existence of tight borrowing constraints or the presence of an extensive female labor supply margin given their implementation method. Macroeconomic and Fiscal Policy Implications of Household Labor Supply Chunzan Wu Dissertation Abstract chunzan@sas.upenn.edu October 16, 2015 https://economics.sas.upenn.edu/graduate-program/candidates/chunzan-wu More Unequal Income But Less Progressive Taxation: Economics or Politics? (Job Market Paper) Overview Since the 1970s in the US, income inequality has increased sharply. During the same time span, the US federal income tax has become less progressive. Why? Does it imply the policymakers today are more influenced by the high income people than they were in 1970s? Or are there economic forces behind such income tax policy changes? I examine this questions in a Ramsey optimal tax policy framework, and identify the roles of the changing economic fundamentals and the changing preferences of policymakers in shaping the less progressive income tax we observe. I find that skill-biased technological change, increased female labor productivity, and the aging US population require a less progressive income tax, but their effects are partially offset by the increase of idiosyncratic income risks. Overall, economic changes require the income tax to be less progressive, and quantitatively account for 40% of the reduction in progressivity in the data. When interpreted through the lens of my model, the remaining change of the income tax implies a change in preferences of policymakers towards high income households and less valued government services. In the last part of the paper, I provide potential explanations for this change of political preferences of policymakers such as lower cost of conveying information to swing voters and the rising inequality of voter turnout among different socioeconomic groups. Methodology In the Ramsey optimal tax policy framework, the tax policy is determined by: (1) a set of Pareto weights representing the government’s preference over different households; (2) household lifetime utilities encompassing the effects of economic fundamentals. The changes of economic fundamentals since 1970s is quantified using an economic model of households disciplined by the data in 1970s and 2010s. The Pareto weights are harder to measure directly, but can be inferred from the actual tax policies chosen by inverting the Ramsey problem, i.e., finding the Pareto weights which rationalize the actual tax policy as the solution to the Ramsey problem. By construction, the combination of the change of Pareto weights identified this way and the economic changes in the model replicates exactly the change of the income tax in the data. A counterfactual experiment in which economic changes are introduced whereas holding the Pareto weights fixed at the 1970s values identifies the part of income tax change which serves as an optimal response to the underlying economic forces, and the remaining part is then attributed to the change in political powers, as approximated by a change in Pareto weights. Quantitative Life-Cycle Model The economic model employed to capture the changes of economic fundamentals since 1970s is a quantitative overlapping generations incomplete-markets life-cycle model with heterogeneous households. To model skill-biased technological change, a Ben-Porath style human capital accumulation technology is introduced. The return to human capital investment depends on the heterogeneous learning abilities of earners, and increases from 1970s to 2010s to match the widening gap in the upper half of the labor income distribution. To account explicitly for the changing role of female labor supply in the economy, each household in the model consists of two earners, a male and a female, and they make joint decisions on household consumption, savings, labor supply and human capital investment. Female labor productivity increases between 1970s and 2010s to match the declining gender income gap in the data. Finally, earners face uninsurable idiosyncratic risks of labor productivity both at labor market entry and over the span of household life cycle, and are subject to tight borrowing constraints. The amount of idiosyncratic risk in the model is calibrated to match the dispersion of labor income of young earners and the earnings dynamics over the life cycle in the data. 1 Macroeconomic and Fiscal Policy Implications of Household Labor Supply Chunzan Wu Dissertation Abstract chunzan@sas.upenn.edu October 16, 2015 https://economics.sas.upenn.edu/graduate-program/candidates/chunzan-wu Main Findings The first main finding of my quantitative analysis is that the changes of economic fundamentals since 1970s alone require a less progressive optimal income tax policy to be adopted, and can quantitatively account for 40% of the reduction in progressivity we observe. Counterfactual experiments show that skill-biased technological change, increased female labor productivity, and the aging of population all contribute to the less progressive optimal income tax, but the effects of these are partially offset by the increase of idiosyncratic risks which increased the insurance benefits of progressive income taxes. The second main finding is that the Pareto weights have changed between 1970s and 2010s in two dimensions: (1) the Pareto weights of the high income households have increased relative to those at the lower end of the income distribution; (2) the Pareto weights on household private utilities have increased relative to the weight on government services. The first change is most responsible for the remaining 60% of reduction in income tax progressivity, while the second change is most responsible for the significant fall of the overall level of US income taxes since 1970s. Political Economy Explanations Since the model ascribes a significant part of observed changes in the income tax policy to changes in political powers, as approximated by the Pareto weights, in the last part of this paper, I provide potential political economy explanations for this phenomenon. Using a stylized probabilistic voting model with political contributions, I show that a lower cost of conveying information to swing voters due to information technology improvements leads to an increased demand for campaign expenditures, as observed in the data. I show that this induces a change of the Pareto weights benefiting the high income households, consistent with the change of income tax policy studied in the first part. I also show that rising inequality of voter turnout among different socioeconomic groups may also have contributed to such changes of the Pareto weights. How Much Consumption Insurance in Bewley Models with Endogenous Family Labor Supply? (joint with Dirk Krueger) Abstract We show that a calibrated life-cycle two-earner household model with endogenous labor supply can rationalize the extent of consumption insurance against wage shocks estimated empirically by Blundell, Pistaferri, and Saporta-Eksten (2014) (BPS hereafter) in US data. With additive separable preferences, the model can account for about 94% and 91% of consumption insurance against male and female permanent wage shocks in the data, and only 41% of male and 28% of female permanent wage shocks in the model pass through to household consumption. With non-separable preferences, more consumption insurance is generated, and the pass-through rates are 27% and 18%, respectively. Most notably, the majority of the consumption insurance against permanent male wage shocks is provided through the endogenous labor supply response of the female earner. We also evaluate, using model-simulated data, whether the empirical approach of BPS delivers unbiased consumption responses to wage shocks. We find that the method overestimates the amount of consumption insurance against male permanent wage shocks, and underestimates that against female permanent wage shocks, but only moderately so. We find larger biases for the outside insurance coefficient which BPS use to capture the insurance provided through channels outside their model. We document that the magnitudes of the biases are not sensitive to the existence of tight borrowing constraints or the presence of an extensive female labor supply margin given their implementation method. 2 YinYin Yu https://economics.sas.upenn.edu/graduate-program/candidates/yinyin-yu yinyiny@sas.upenn.edu UNIVERSITY OF PENNSYLVANIA Placement Director: Iourii Manovskii Placement Director: Andrew Postlewaite Graduate Student Coordinator: Kelly Quinn MANOVSKI@ ECON.UPENN.EDU APOSTLEW@ECON.UPENN.EDU KQUINN @ ECON.UPENN.EDU 215-898-6880 215-898-7350 215-898-5691 Office Contact Information 3718 Locust Walk Rm 160 Philadelphia, PA 19104 Phone: (410) 979-6345 Citizenship: USA Undergraduate Studies: BA, Economics, Mathematics, University of Pennsylvania, magna cum laude, 2010 Graduate Studies: University of Pennsylvania, 2010 to present Thesis Title: “The Intended and Unintended Consequences of Regulating For-Profit Colleges: A Model of Enrollment and Retention” Expected Completion Date: May 2016 Thesis Committee and Teaching Reference: Prof. Hanming Fang (Primary Advisor) Department of Economics 3718 Locust Walk Rm 160, Philadelphia, PA 19104 hanming.fang@econ.upenn.edu (215) 898-7767 Prof. Jean-Francois Houde Wharton School of Business 1455 Steinberg Hall-Dietrich Hall, 3620 Locust Walk, Philadelphia, PA 19104 houde@wharton.upenn.edu (215) 573-6849 Prof. Petra Todd Department of Economics 3718 Locust Walk Rm 160, Philadelphia, PA 19104 ptodd@econ.upenn.edu (215) 898-4084 Prof. Eduardo Azevedo (Teaching Reference) Wharton School of Business 1455 Steinberg Hall-Dietrich Hall, 3620 Locust Walk, Philadelphia, PA 19104 eazevedo@wharton.upenn.edu (215) 573-9984 Research Fields: Education, Industrial Organizations, Applied Microeconomics Teaching Experience: Spring, 2015 Honors Managerial Economics, Wharton School of Business, University of Pennsylvania, Recitation Instructor for Professor Eduardo Azevedo Fall, 2014 Economic Analysis of the Public Sector (MPA), Fels Institute of Government, University of Pennsylvania, Teaching Assistant for Professor Janice Madden Fall, 2014 Economic Principle of Public Policy (Executive MPA), Fels Institute of Fall, 2012 Fall, 2012 Spring, 2012 Fall, 2011 Government, University of Pennsylvania, Teaching Assistant for Professor Peter Angelides Topics in Development, Department of Economics, University of Pennsylvania, Teaching Assistant for Professor Jere Behrman Law and Economics, Department of Economics, University of Pennsylvania, Teaching Assistant for Professor Camilo Garcia-Jimeno Development Economics, Department of Economics, University of Pennsylvania, Teaching Assistant for Professor Matthias Kredler Topics in Development, Department of Economics, University of Pennsylvania, Teaching Assistant for Professor Jere Behrman Research Experience and Other Employment: 2007-2010 Research Assistant, Department of Accounting, Wharton School of Business 2008-2010 Research Assistant, Center for International Comparisons, University of Pennsylvania Honors, Scholarships, and Fellowships: 2010-2011 University Fellowship Programming Languages: MATLAB, STATA Job Market Paper: “The Intended and Unintended Consequences of Regulating For-Profit Colleges: A Model of Enrollment and Retention” My research takes a closer look at the public debate over for-profit colleges and evaluates the consequences of restricting federal student financial aid to for-profit college students as proposed by policy makers. Opponents of for-profit higher education accuse these schools of gaming the federal student financial aid system by providing extra assistance to their students on financial aid applications and inflating tuition. On the other hand, proponents of for-profit higher education argue that these colleges in fact serve a niche student population that is underserved by community colleges, and that forprofit colleges provide value in terms of higher graduation rates than community colleges. To understand the effect of restricting federal student financial aid to for-profit college students, I develop a two-period discrete choice model of differentiated products much in the spirit of Berry, Pakes and Levinsohn (1995) and Petrin (2002), and estimate it using school-level data collected by the Department of Education. In my model, risk neutral and forward-looking individuals make both enrollment and dropout decisions based on school characteristics, tuition, and financial aid availability. I find that simply eliminating the extra assistance for-profit colleges provide on federal student loan applications induces a 13.7% decrease in for-profit college enrollment. Of those who leave for-profit colleges, 91.7% choose to abandon college plans altogether rather than switch to community colleges. This validates the claim that for-profit colleges are indeed serving a niche student population. Furthermore, I find that for-profit colleges' graduation rates are on average 7 percentage points higher than those of community colleges, which translates into a 0.79 percentage point reduction in average graduation rate under this counterfactual scenario. However, reducing federal student financial aid for for-profit college students is effective in combating student indebtedness. Eliminating for-profit colleges' extra assistance on federal student loan applications results in a decrease in average loan by 24% ($589) and a decrease in tuition paid by 4% ($176), which means that those who forgo the college option are the most credit constrained. Also, among dropouts, who are four times more likely to default on their student loans than those who graduate, average loan decreases by 21% ($200). YinYin Yu University of Pennsylvania "The Intended and Unintended Consequences of Regulating For‐Profit Colleges: A Model of Enrollment and Retention " In recent years for‐profit colleges have been a topic of controversy, both in the media and among policy makers. These institutions provide postsecondary vocational training and offer mostly certificates and associate degrees. But unlike their main competitors, the community colleges, for‐profit colleges are characterized by much higher tuition, higher federal student financial aid uptake, higher federal student loan default rates, and lower instructional spending. Opponents of for‐profit colleges accuse them of gaming the federal student financial aid system by providing extra assistance to students on financial aid applications and inflating tuition. On the other hand, proponents of for‐profit higher education argue that these colleges actually serve a niche student population that is underserved by community colleges, and that for‐profit colleges provide value in terms of higher graduation rates than community colleges. In my research, I closely examine the debate over the merits and flaws of for‐profit higher education. More specifically I answer the following questions: 1) Are proponents of for‐profit colleges correct in their claims that these schools serve a niche student population and that they have higher graduation rates than community colleges? 2) How would restricting federal student financial aid availability to for‐profit college students, as proposed by policy makers, affect various outcomes of interest, including college enrollment rate, graduation rate, average tuition spent, and average federal student financial aid acquired by those who will eventually drop out, in the market for postsecondary vocational training where the major players are for‐profit colleges and community colleges. My research contributes to the nascent but growing literature on for‐profit colleges. Most of the existing studies focus on examining post‐graduation outcomes of for‐profit college students using individual‐level data (Cellini and Chaudhary 2012; Deming et al. 2012; Lang and Weinstein 2013; Liu and Belfield 2014), or on evaluating the effect of government aid on for‐profit colleges' tuition (Cellini 2009, 2010; Cellini and Goldin 2014). My research aims to understand the determinants of students' enrollment and dropout decisions in the market for postsecondary vocational training, and to assess the effects of potential policy interventions directed at for‐profit colleges. Specifically, I study students' responses to (i) eliminating the extra assistance for‐profit colleges provide on federal student financial aid applications, (ii) increasing the federal student loan interest rate for for‐profit college students, and (iii) denying for‐profit college students access to federal student financial aid. A unique advantage of my modeling framework is its ability to accommodate large for‐profit colleges' multi‐state or national operations, including online programs. This allows me to study the for‐profit college industry in its entirety rather than only focusing on local schools within specific states as done in previous works. Being able to model the operations of these large for‐profit colleges is particularly important as these schools constitute the bulk of for‐profit college enrollment. Another attractive feature of my model is that it accounts for non‐pecuniary benefits of college. Most of the existing literature on labor market returns to for‐profit colleges find no wage difference from that of community colleges despite the sizable difference in cost. This result alludes to the presence of non‐pecuniary benefits. Yu 1 YinYin Yu University of Pennsylvania I develop a two‐period discrete choice model of differentiated products much in the spirit of Berry, Pakes and Levinsohn (1995) and Petrin (2002), and estimate it using school‐level data collected by the Department of Education. My sample of colleges consists of the set of Title‐IV eligible postsecondary institutions with no admission standards, most of which are for‐profit colleges and community colleges. I define the set of potential students as those between ages 18‐44, with at least a high school diploma but less than a Bachelor's degree. An individual's choice set consists of the set of colleges available in his/her Core‐Based Statistical Area (CBSA). A college is available in the CBSA's in which it has campuses, with the exception of a few colleges with extensive nation‐wide online offerings that are available in every CBSA regardless of their campus presence. Individuals in my model are forward‐looking and risk neutral, and they make decisions in two periods. In the first period, individuals choose among the schools in their choice sets based on school characteristics, tuition, and financial aid availability. They also have the outside option of not going to school. At this stage, individuals know the payoff of not attending college but only know the payoff distribution associated with graduating from each college in their choice sets, where different colleges' payoff distributions differ in mean but have the same variance. Individuals may have different tastes for school characteristics depending on their demographics. In the second period, those enrolled in colleges learn about the payoff from graduating with a degree/certificate, decide whether or not to drop out, and receive the corresponding payoffs. They are also liable to start repayment on their student loans upon leaving school. I combine school‐level data from the Integrated Postsecondary Education Data System (IPEDS) with data from the Postsecondary Education Participants Systems (PEPS) to construct a complete picture of the 2009‐2010 school year, which marked the height of the for‐profit college industry. I estimate my model using GMM, matching moments related to grants and loans, graduation rate, market shares, and demographic‐specific market shares. My estimates show that for‐profit college enrollment is heavily reliant on federal student financial aid and that these institutions actually do serve a niche student population. For instance, simply eliminating the extra assistance for‐profit colleges provide on federal student financial aid applications induces in a 13.7% decrease in for‐profit college enrollment (3.5% of total enrollment). Of those who leave for‐profit colleges, 91.7% choose to abandon college plans altogether rather than switch to community colleges. This validates the claim that for‐profit colleges are indeed serving a niche student population that is underserved by community colleges. Furthermore, I find that for‐profit colleges' graduation rates are on average 7 percentage points higher than those of community colleges, which translates into a 0.79 percentage point reduction in average graduation rate under this counterfactual scenario. However, restricting federal student financial aid access to for‐profit college students is effective in reducing student indebtedness. Eliminating for‐profit colleges' extra assistance on federal student loan applications results in a decrease in average loan by 24% ($589) and a decrease in tuition paid by 4% ($176), which means that those who forgo the college option are the most credit constrained. Also, among dropouts, who are four times more likely to default on their student loans than those who graduate, average loan decreases by 21% ($200). Yu 2