Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Economics, 1992. / Includes bibliographical references (leaves 191-202). / by Fred Anthony Bellemore. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Economics, 2001. / Includes bibliographical references. / This dissertation presents three essays in Corporate Finance. In the first essay, I study managerial incentives in internal capital markets. In particular, I develop a two-tiered agency model to study division managers' incentives within internal capital markets. Division managers try to influence the external capital market's assessment of the firm and the internal capital market's assessment of their divisions in order to increase their level of funding. I show that, as the number of divisions increases, the external capital market's assessment of the firm becomes a public good for division managers, and the internal capital market replaces the external capital market in the provision of managerial incentives. I also show that, while diversified firms have an advantage in allocating resources, this may come at the expense of managerial incentives. Based on the analysis, the paper relates the value of diversification to characteristics of the firm, the industry, the external capital market, and the internal capital market. In the second essay, I propose a model of entrepreneurship in which investors decide whether to become venture capitalists or to form firms and entrepreneurs decide whether to join a firm or seek financing in the venture capital market. The venture capital market allows better matching between investors and entrepreneurs, but this comes at the cost of adverse selection. The model suggests that as a sector matures, innovation takes place first within firms, then in ventures backed by venture capitalists backed ventures, and finally within firms again. / (cont.) In addition, I analyze the relationships between the venture capital market and investors' diversity, investors' scope of expertise and entrepreneurial incentives. The third essay, which is co-authored with Andres Almazan, examines how the trading activities of institutional investors can help to mitigate agency conflicts in corporations. The access of institutional investors to privileged information produces an adverse selection effect that reduces the trading activity of institutional investors and generates a free-rider problem that affects the intensity with which institutional investors wish to "vote with their feet". We also study ownership implications, incentives to acquire information and the interaction of the Wall Street Rule with other mechanisms of governance (i.e. capital structure). / by Adolfo de Motta Gregori. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, 1998. / Includes bibliographical references (p. 127-131). / This thesis investigates the relationship between demographics and asset prices. More specifically it examines the effect of changes in the age distribution of the U.S. population on housing, stock, and bond prices over the post World War II period in the U.S. This is done in two steps. First, survey data on household asset holdings is used to construct age profiles of household demand for housing, stocks, bonds, and debt. These asset demand profiles are combined with data on the age distribution of the U.S. population to construct time series measures of aggregate demographic demand for housing, financial assets net of debt, and stocks in excess of bonds, which are then used to analyze the effects of demographically driven changes in aggregate asset demand on equilibrium asset prices over the period from 1946 through 1997. The results of this exercise suggest several interesting findings. With respect to the microeconomic issue of life cycle investment behavior, one finds that the scale and composition of household asset demand changes dramatically over the course of the economic life cycle. Young households, that is, households with heads under age 40, tend to draw credit out of financial markets, primarily by issuing mortgage contracts for the purchase of houses. The extent of this and other borrowing done by young households tends to exceed any gross contributions they make to financial markets through transactions accounts, mutual funds, retirement plans, etc., making them net negative investors in financial assets on average. In contrast, households with heads between ages 40 and 60, tend to provide substantial amounts of credit to financial markets. Much of this saving is, at least nominally, retirement saving, held in personal retirement accounts and employer provided pensions. Households with heads over age 60 tend, like younger households, to drain credit from financial markets. However, unlike young households, older households draw credit out of financial markets not by borrowing, rather, by using previously accumulated assets to fund consumption during retirement. Due to large shifts in the age distribution of the U.S. population since 1946, these life cycle investment patterns appear to have had significant macroeconomic consequences. Tests of the correlation between the constructed demographic demand variables and corresponding asset price series, suggest a statistically significant link between demographic changes in the U.S. population and observed long run movements in housing, stock, and bond prices. This is true even after controlling for the effects of other factors such as fluctuations in real GDP (in the case of housing and bond prices) and dividends (in the case of stock prices). Estimated elasticities of real housing prices with respect to the demographic demand for housing suggest that demographic factors can account for approximately 59% of the observed annual increase in real housing prices between 1966 and 1986. Similarly, demographically driven changes in the demand for financial assets can account for approximately 77% of the observed annual increase in real stock prices between 1986 and 1997 and can account for at least 81 % of the observed annual increase in real bond prices. As for the future, current Census Bureau population projections suggest that annual growth in demographic housing demand will provide a positive stimulus of about 0.35% per year to real housing price appreciation between 1997 and 2007, down from about 0.98% per year for the period between 1986 to 1997, and 1.02% per year for the period between 1966 and 1986. Growth in the demographic demand for financial assets is expected to provide a positive stimulus to real stock and bond price appreciation of about 8.76% per year between 1997 and 2007, up from about 6.62% per year for the period between 1986 and 1997, and -1.34% per year for the period between 1966 and 1986. / by Steven M. Bergantino. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, 2002. / Includes bibliographical references (p. 113-117). / Chapter 1 For most US households, labor income is the most important source of wealth and housing is the most important risky asset. A natural intuition is thus that households whose incomes covary relatively strongly with housing prices should own relatively little housing. Under plausible assumptions on preferences and distributions, this result holds theoretically. Empirically, I find a significant effect: among US households, a one standard deviation increase in income-house price covariance is associated with a decrease of approximately $25,000 in the value of owner occupied housing. This empirical result implies greater cognizance of the interaction between labor income and asset risk on the part of households than suggested by most analyses of stock market behavior. The analysis also suggests that many homeowners enter financial markets in a riskier position than typically thought, and reinforces the intuitive appeal of proposals for market- or tax-based risk sharing in housing prices. Chapter 2 extends the theory of annuitization with no bequest motive in two directions. First, we derive sufficient conditions, in a more general setting than Yaari (1965), under which complete annuitization is optimal, and weaker conditions under which partial annuitization is better than zero annuitization. Second, we explore how incremental and complete annuitization affect consumer welfare in these more general conditions. When markets are complete, all savings are optimally annuitized as long as there is no bequest motive and annuitized assets have greater returns than conventional assets. / (cont.) Consumers' utility need not satisfy intertemporal additive separability nor the expected utility axioms, and annuities need not be actuarially fair. The result is weakened if annuities markets are incomplete, so that there are some assets which do not exist in annuitized form: as long as trade occurs all at once and consumption is positive in every state of nature, a small degree of annuitization is better than no annuitization. When conventional asset markets are incomplete, if annuities are illiquid, then it is possible that no savings are annuitized. We present numerical calculations of the financial benefit and optimal degree of annuitization for consumers with standard CRRA preferences, and compare these results to results where otherwise identical consumers have utility that depends both on present consumption and a standard of living to which they have grown accustomed. In our specification, the effect of adding intertemporal dependence hinges on the size of initial standard of living relative to resources. Chapter 3 addresses the measurement of income sorting and the attribution of observed sorting to different causes. In terms of measurement, I show that a standard decomposition of variance of household income into within jurisdiction and between jurisdiction components understates sorting in the presence of measurement error. Using 1990 US Census data, I find that adjusting for this error approximately doubles the estimated extent of sorting. On average, across all US metropolitan areas (MSAs) I find that approximately ten percent of the variation in household income can be explained by differences across jurisdictions ... / by Thomas Davidoff. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, 2000. / Includes bibliographical references (p. 141-149). / Chapter 1: This chapter examines the logical consequences of the rather unsurprising notion that humans care about and manage their self-image, a notion long taken for granted by psychologists. I model this by assuming that decisionmakers derive utility from positive views about the self, holding constant standard utilitarian outcomes usually assumed relevant in economics. Other than this, agents are timeconsistent expected utility maximizers, are constrained in their updating by Bayes' rule, and can manipulate their beliefs only by controlling the flow of information that they receive. The motive to maintain a favorable self-image leads to a systematic rejection of free information about the self in certain states of the world and eventually to overconfident beliefs. Economically relevant decisions are affected by this overconfidence as well as the incentive to gather information about and make decisions so as to optimally manage beliefs. Agents might avoid informative actions when satisfied with their current beliefs ('self-image protection'), and seek out activities in which they can prove themselves when they are not ('self-image enhancement'), even if these choices are otherwise poor. These motives lead to a whole host of effects on behavior that other models have trouble explaining in a unified framework. The model can also make testable predictions on how these effects play themselves out across different categories of tasks and within a category of tasks over time. Applications to stock market participation, the choice between salaried and self-employment, career choice, manager behavior, and employee motivation are discussed. Chapter 2: This chapter starts from the same premise as the previous one, the assumption that agents care about their self-image, but examines its consequences in a different information structure. Agents can improve financial decisions by making subjective judgments about their payoffs, while they derive ego utility from their perceptions regarding this ability. If the agent has a self-image protection motive, she will as a result be averse to making a subjective judgment and reviewing it later, since this combination is informative about ability. The consequence could be a sluggishness in responding to new information, procrastination in making up one's mind, or the reliance on inferior objective information. Possible remedies and applications are discussed, with particular attention to anxiety about health. Chapter 3 (with Peter Diamond): There is overwhelming psychological evidence that some people run into self-control problems regularly, yet the effect of these findings on major life-cycle decisions hasn't been studied in detail. This paper extends Laibson's quasi-hyperbolic discounting savings model, in which each intertemporal self realizes that her time discount structure will lead to preference changes, and thus plays a game with her future selves. By making retirement endogenous, savings affect both consumption and work in the future. From earlier selves' points of view, the deciding self tends to retire too early, so it is possible that the self before saves less to induce her to work. However, still earlier selves think the pre-retirement self may do this too much, leading to possible higher saving on their part and eventual early retirement. Thus, the consumption path exhibits observational non-equivalence with exponential discounting. Observational non-equivalence also obtains on a number of comparative statics questions. For example, a self could have a negative marginal propensity to consume out of changes in future income. The outcome with naive agents, who fail to realize their self-control problem, is also briefly discussed. In that case, the deciding self's potential decision to retire despite earlier selves' plans results in a downward updating of available lifetime resources, and an empirically observed downward jump in the consumption path. / by Botond Kőszegi. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, 2000. / Includes bibliographical references (p. 139-144). / This thesis studies four different aspects of balance of payments crises. Chapter l provides a dynamic asymmetric-information model of the timing of crises. It focuses on investors' learning process and its interaction with interest rate policy. The model shows that the presence of private information delays the onset of BOP crises, giving rise to large drops in asset prices when crises finally take place. It also shows that raising interest rates can be an effective defense against speculative attacks: the optimal policy consists of raising interest rates sharply as fundamentals become very weak. However, this policy is time inconsistent, suggesting a role for commitment devices such as currency boards or IMF pressure. Chapter 2 studies the relationship between macroeconomic fundamentals and asset prices during crises. Key findings are that fundamentals can account for a significant part of the cross-sectional variance of stock returns during crises, and that credit market conditions play a crucial role during crises. Chapter 3 studies the behavior of spreads on emerging-market sovereign bonds of different maturities, focusing on the supply side of funds. Spreads on long-term bonds are shown to be too volatile to be reconciled with investors' being risk-neutral and financially unconstrained. An explanation for this volatility is proposed, based on the fact that investors holding long-term bonds are subject. to substantial price risk. A study of the expected returns and volatility of holding bonds of different maturities after drops in bond prices provides empirical support. Chapter 4 examines the degree of real exchange rate misalignment in seven Latin American countries and in the U.S. between 1960 and 1998. In all cases there is a long-run relationship among the CPI-based real exchange rate, stock of net foreign assets and relative price of nontradable goods. The results suggest that in 1998 the real exchange rate in Peru was in equilibrium, in Chile slightly undervalued, in Venezuela overvalued by about 8% and in the US overvalued by about 16%. In Argentina, Brazil, Colombia and Mexico, the exchange rate was overvalued by over 20%. / by Fernando A. Broner. / Ph.D.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Economics, 2018. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 147-156). / This thesis consists of three chapters on the economics of health and healthcare. The first and third chapters explore geographic variation in health outcomes within the United States. The second chapter focuses on empirical methods for obtaining causal estimates of treatment effects with an application to healthcare settings. In the first chapter I study geographic variation in health care utilization under two different insurance systems: traditional Medicare and employer-provided private insurance. For each system, I use patient migration as a source of identification combined with empirical Bayes methods to construct optimal linear forecasts for the causal effects of place on utilization. These place effects measure the causal differences in treatment intensity across areas. I find similar levels of variation in the causal place effects for the publicly and privately insured patients, with a correlation of .39 across the two systems. These findings emphasize that insurance systems are affecting the forces that drive the causal component of geographic variation in utilization. In the second chapter, Liyang Sun and I explore event studies, a model for estimating treatment effects using variation in the timing of treatment. Researchers often run fixed effects regressions for event studies that implicitly assume treatment effects are constant across cohorts first treated at different times. In this paper we show that these regressions produce causally uninterpretable estimands when treatment effects vary across cohorts. We propose alternative estimators that identify convex averages of the cohort-specific treatment effects, hence allowing for causal interpretation even under heterogeneous treatment effects. We illustrate the shortcomings of fixed effects estimators in comparison to our proposed estimators through an empirical application on the economic consequences of hospitalization. In the third chapter, Raj Chetty, Michael Stepner, Shelby Lin, Benjamin Scuderi, Nicholas Turner, Augustin Begeron, David Cutler and I use newly available administrative data to quantify the relationship between income and mortality in the United States. Although it is well known that there are significant differences in health and longevity between income groups, debate remains about the magnitudes and determinants of these differences. We use new data from 1.4 billion anonymous earnings and mortality records to construct more precise estimates of the relationship between income and life expectancy at the national level than was feasible in prior work. We then construct new local area (county and metro area) estimates of life expectancy by income group and identify factors that are associated with higher levels of life expectancy for low-income individuals. Our study yields four sets of results. First, higher income was associated with greater longevity throughout the income distribution. The gap in life expectancy between the richest 1% and poorest 1% of individuals was 14.6 years for men and 10.1 years for women. Second, inequality in life expectancy increased over time. Between 2001 and 2014, life expectancy increased by 2.34 years for men and 2.91 years for women in the top 5% of the income distribution, but increased by only 0.32 years for men and 0.04 years for women in the bottom 5%. Third, life expectancy varied substantially across local areas. For individuals in the bottom income quartile, life expectancy differed by approximately 4.5 years between areas with the highest and lowest longevity. Changes in life expectancy between 2001 and 2014 ranged from gains of more than 4 years to losses of more than 2 years across areas. Fourth, geographic differences in life expectancy for individuals in the lowest income quartile were significantly correlated with health behaviors such as smoking, but were not significantly correlated with access to medical care, physical environmental factors, income inequality, or labor market conditions. Life expectancy for low income individuals was positively correlated with the local area fraction of immigrants, fraction of college graduates, and local government expenditures. Additional information on this project is available at https: //healthinequality. org/. / by Sarah Marie Abraham. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Economics, 2013. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 154-159). / This dissertation investigates the different roles that networks play in political economy. In the first chapter, I study how a political party uses electoral data to monitor and incentivize the political brokers who control its clientelistic networks. I study networks organized around rural communal lands in Mexico, which are largely controlled by the Institutional Revolutionary Party (PRI). I use the fact that the level at which brokers operate (the communal land) does not necessarily coincide with the level at which the electoral data is disclosed (the electoral section). Guided by a simple model, I compute a measure of how informative the available electoral data is about the performance of the PRI's political brokers, as a function of the degree of overlap between communal lands and electoral sections. I compare the vote share for the PRI in communal lands where the electoral data is more or less informative, both when the PRI does and does not have access to resources to fund and incentivize brokers. The results suggest that clientelistic networks contribute significantly to the enforcement of clientelistic transactions. In the second chapter, which is co-authored with Joana Monteiro, we study the role of media in compensating political biases. In particular, we analyze how media presence, connectivity and ownership affect the distribution of federal drought relief transfers to Brazilian municipalities. We find that municipalities that are not aligned with the federal government have a lower probability of receiving funds conditional on experiencing low precipitation. However, we show that the presence of radio stations compensates for this bias. This effect is driven by municipalities that have radio stations connected to a regional network rather than by the presence of local radio stations. In addition, the effect of network-connected radio stations increases with their network coverage. These findings suggests that the connection of a radio station to a network is important because it increases the salience of disasters, making it harder for the federal government to ignore non-allies. We show that our findings are not explained by the ownership and manipulation of media by politicians. In the third chapter, which is co-authored with Arun Chandrasekhar and Emily Breza, we shed light on the relationship between network characteristics and investment decisions through a lab experiment in the field. We focus on the role for third parties to act as informal contract enforcers. Our protocol builds on a basic two-party trust game with a sender and receiver, to which we introduce a third-party to serve as either a monitor or punisher. The ex-ante benefits of a third party judge are ambiguous. On one hand, a third party may result in larger sender transfers due to her ability to punish. On the other hand, the punisher might act in a way to build reputation or may crowd-out intrinsic motivation. Importantly, these costs and benefits of a punisher might vary with her centrality in the network. Our findings are consistent with both the role for the punisher to induce efficiency and to crowd out intrinsic motivation. They are also consistent with the effects of reputation-building by the punisher. Importantly, we find that very network-peripheral punishers are detrimental to efficiency, while network-central individuals may improve outcomes when given the technology to punish. We also show that these results cannot be explained by either the fact that the punisher also acts as a monitor, or by the punisher's characteristics such as elite status, educational attainment, caste, or proxies for wealth. / by Horacio Alejandro Larreguy Arbesú. / Ph.D.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Economics, 2017. / Cataloged from PDF version of thesis. / Includes bibliographical references. / In this thesis, I take up the question of how government regulations impact the supply side of the US health care market. In Chapter 1, I exploit the Medicare rural floor, a discontinuity in geographic adjustments to Medicare payments to hospitals, in a regression kink design to estimate the impact of Medicare reimbursement rate changes on the level and mix of hospital services provided. I find that hospitals respond to higher Medicare reimbursement by admitting more Medicare patients, but that the average duration of a patient stay declines. I also document a previously unstudied spillover of Medicare reimbursement on the volume of admitted patients with non-Medicare insurance. For the remainder of my thesis, I turn my attention to Certificate of Need regulation-restrictions on capital investment-in North Carolina. I use newly collected data on the MRI market in North Carolina to estimate the dynamic effects of marginal relaxations of CON restrictions on provider investment. Thresholds in the regulatory approval process enable me to employ a regression discontinuity model to estimate the effects of CON on hospitals and patients. In Chapter 2 I document three key findings. First, health care providers essentially always adopt newly allowed MRI scanners within two years of a CON approval. This suggests that CON regulations are a binding constraint on investment. Second, there is an active market for unregulated mobile MRI scanners that enable health care providers to mitigate the impact of CON restrictions. Finally, MRI utilization increases after new machines are acquired, indicating CON rules affect not only hospital investment but also medical practice. In Chapter 3 I combine the North Carolina regulatory data with patient-level Medicare claims data to estimate the effects of CON on patient care. Hospitals do not offset reductions in MRI services with increased usage of alternative diagnostic imaging technologies. CON regulations reduce the number of scans patients with lower back pain receive by more than a third in the month following their first hospital encounter, reducing medical spending by roughly $400. There is no evidence that this decrease in the intensity of care comes at the cost of increased future lower back pain. / by Bryan J. Perry. / 1. Hospital Responses to Medicare Reimbursement Rate Changes: New Evidence from Medicare's Rural Floor -- 2. Certificate of Need Regulation and Hospital Behavior: Evidence from MRIs in North Carolina -- 3. The Impacts of Certificate of Need Regulation on Patient Care: MRI Utilization in North Carolina. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Economics, 2012. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 189-198). / This thesis consists of four essays on finance, learning, and macroeconomics. The first essay studies whether learning can explain why the standard consumption-based asset pricing model produces large pricing errors for U.S. equity returns. I prove that under learning standard moment conditions need not hold in finite samples, leading to pricing errors. Simulations show that learning can generate quantitatively realistic pricing errors and a substantial equity risk premium. I find that a model with learning is not rejected in the data, producing pricing errors that are statistically indistinguishable from zero. The second essay (co-authored with Anna Mikusheva) studies the properties of the common impulse response function matching estimator (IRFME) in settings with many parameters. We prove that the common IRFME is consistent and asymptotically normal only when the horizon of IRFs being matched grows slowly enough. We use simulations to evaluate the performance of the common IRFME in a practical example, and we compare it with an infrequently used bias corrected approach, based on indirect inferences. Our findings suggest that the common IRFME performs poorly in situations where the sample size is not much larger than the horizon of IRFs being matched, and in those situations, the bias corrected approach with bootstrapped standard errors performs better. The third essay (co-authored with Ricardo Caballero) documents that, in contrast with their widely perceived excess return, popular carry trade strategies yield low systemicrisk- adjusted returns. In contrast, hedging the carry with exchange rate options produces large returns that are not a compensation for systemic risk. We show that this result stems from the fact that the corresponding portfolio of exchange rate options provides a cheap form of systemic insurance. The fourth essay shows that the documented overbidding in pay-as-you-go auctions relative to a static model can be explained by the presence of a small subset of aggressive bidders. I argue that aggressive bidding can be rational if users are able to form reputations that deter future competition, and I present empirical evidence that this is the case. In auctions without any aggressive bidders, there is no evidence of overbidding in PAYGA. / by Joseph Buchman Doyle, Jr. / Ph.D.