Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references. / This thesis is composed of four studies on the role of annuities and life insurance in the portfolios of elderly households. It has long been known in the economics literature that annuities should be of substantial value in the portfolios of retired individuals. However, outside of private pension plans. the purchase of private annuity contracts is relatively rare. This paper explores the issues associated with this "Annuity Puzzle" using the framework of a life cycle model of consumption, extended to account for mortality uncertainty. Chapter 1 empirically examines household decisions about whether or not to annuitize balances 111 defined contribution pension plans upon retirement. This study calculates a utility based measure of annuity valuation for households nearing retirement, and allows for variation across households to result from variation in mortality rates, risk aversion. marital status, and the presence of pre-existing annuities such as Social Security. It finds that a one percentage point increase in the calculated gain from annuitization is associated with a one percentage point increase in the ex ante probability of annuitizing one's retirement resources. It also finds that the presence of bequest motives has no effect on the disposition of definer.'. contribution plan assets. The finding that bequest motives do not affect marginal annuity decisions is in contrast to some of the previous literature. In particular, previous research has argued that a significant fraction of the elderly are over-annuitized by Social Security, as evidenced by their decision to held life insurance. Chapter 2 re-examines this finding using new and better data on the age 70+ population, and finds little support for the existence of strong bequest motives. One important source of variation in the decision of whether or not to annuitizc is marital status. Chapter 3 explores the demand for joint-life annuity products among marrind couples, and finds that couples should value annuitization less than single individuals. This is because couples have Opportunities to share risk between them, and as such, marriage is a partial substitute for perfect annuity markets. This may help to explain the relative scarcity of private annuity contracts in the U.S. Chapter 4 extends the analysis of annuities to account for inflation risk. It explores the relative value of real versus nominal annuities, and also explores the gains to purchasing annuities that arc linked to an underlying portfolio of risky assets. / by Jeffrey R. Brown. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references. / This dissertation proposes new instrumental variable methods to identify, estimate and test for causal effects of endogenous treatments. These new methods are distinguished by the combination of nonparametric identifying assumptions and semiparametric estimators that provide a parsimoniuous summary of the results. The thesis consists of three essays presented in the form of chapters. The first chapter shows how to estimate linear and nonlinear causal response functions with covariates under weak (instrumental variable) identification restrictions. The second chapter (co-authored with Joshua Angrist and Guido Imbens) applies the identification results of the first chapter to estimate quantile causal response functions, so we can study the effect of the treatment on different parts of the distribution of the outcome variable. The third chapter of this dissertation looks again at distributional effects but focusing directly on the cumulative distribution functions of the potential outcomes with and without the treatment. / by Alberto Abadie. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references. / This thesis consists of three empirical studies that address issues of public economics in developing countries. Two broad questions motivated me in turn: one, the effect of household institutions on the provision of old age security; and two, the effect of political institutions on government policy and performance. Chapter 1, entitled Inter-generational Transfers and Intra-household Bargaining: Evidence from Indonesia, examines whether there is bargaining between husbands and wives within the donor household over transfers to their respective parents, using a household survey from Indonesia. The evidence is that wives' education and income have a significant positive effect on transfers to their parents, and no effect on transfers to their husbands' parents. In addition, women who are gainfully employed and who have greater dowries from the time of their marriage are more likely to make transfers to their parents. This evidence is consistent with a model of collective decisionmaking where education and income influence the bargaining power of women and therefore the allocation of household resources. It also implies that intergenerational transfers may be viewed as returns to dowry and human capital investments made by parents. These interpretations are not incompatible-- parental investment may endow daughters with the bargaining power to channel resources towards their parents. In conclusion, the evidence addresses both gender roles in providing support to elderly parents and returns to parental investment in old age security, in the context of developing countries with no public social security system. Both issues have enormous policy ramifications for income redistribution programs and the design of formal social security institutions for economies with traditional inter-household arrangements. Chapter 2, entitled Effect of Electoral Accountability on Economic Policy in India, studies the effect of state legislative assembly elections on the economic policies of state governments in 14 major states of India, over the period 1960-1994. The effect of the timing of elections on economic policies is identified using an instrument for the electoral cycle that distinguishes between constitutionally scheduled elections and midterm polls. Election years have a negative effect on commodity taxes, a positive effect on capital spending, and a positive effect on road construction by public works departments. The Indian political cycle is of greater magnitude than any comparable cycle in the developed countries. In addition, unlike political budget cycles in OECD countries that are accompanied with higher budget deficits, state elections in India have no effect on state deficits. This evidence is consistent with a moral hazard model where career concerns persuade politicians to improve performance. The electoral cycle in policy is generated by high discounting of the future in an uncertain political environment. The pattern suggests that state governments strategically manipulate economic policies to increase the provision of public services, without increases in taxes and deficits, in order to influence the probability of reelection. Chapter 3, entitled Partisan Politics and Intergovernmental Transfers in India, studies the effect of partisan politics on budgetary transfers from the central to the state governments in India. Using a panel of 14 major Indian states, from 1960 to 1994, it tests for the effect of the party affiliation of a state government on grants and loans from the center, and its share in central taxes. The results indicate that when a state government is politically affiliated with the party governing at the center it receives significantly greater grants per capita, 10 per cent higher than the average per capita grants. The timing of state elections has no significant effect on central grants. These results are consistent with a model of partisan manipulation of grants within a system of centralized decisionmaking. Central governments give greater grants to affiliated states in order to exercise greater control over state spending. / by Stuti Khemani. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references. / This thesis is a collection of three essays in empirical development economics. The first chapter evaluates the effects on education and wages of a large school construction program undertaken by the Indonesian government between 1973 and 1978. I evaluate the effect of this program on education and wages by combining differences across regions in the number of schools constructed with differences across cohorts induced by the timing of the program. The estimates suggest that the construction of primary schools led to a substantial increase in education and earnings. These estimates implie returns to education ranging from 6.4% to 9.1%. This second chapter studies the impact of household resources on child nutrition in South Africa. In the early 1990s, the benefits and coverage of the South African social pension program were dramatically expanded for the black population. About a third of black South African children under age 5 live with an elderly person. This chapter examines whether this large positive income shock was followed by an improvement of anthropometric status. Estimates suggest that pensions received by women had a large and significant impact on the anthropometric status of girls and a smaller and insignificant effect on that of boys. I found no effect of the pension on child nutrition when it is received by men. The third chapter examines the role that reputation plays in determining contractual outcomes, using a data set containing detailed information about 230 projects carried out by 125 software firms that I have collected for this purpose. Ex ante contracts as well as the outcome after ex post renegotiation vary with firms' characteristics plausibly associated with reputation. I propose a model of the industry where reputation determines contractual outcomes, whose predictions are consistent with several facts observed in the data. / by Esther Duflo. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references. / This thesis consists of three essays in public finance and labor economics, with a focus on the economics of higher education. Chapter J estimates the effect of grant aid on college attendance. I exploit the variation in aid created by the elimination of the Social Security Student Benefit Program, which at its peak provided grants totaling $3.3 billion a year to one out of ten college students. I find that $ I 000 ($1997) of grant aid increases educational attainment by about 0.20 years and the probability of attending college by five percentage points. The elasticities of attendance and completed years of college with respect to schooling costs are estimated to be 0.90 to J .0. Given plausible values for the rate of return to a year of college, the aid program examined by this chapter was a cost-effective use of government resources. Chapter 2 explores the response of middle- and upper-income youth to tuition subsidies. Traditionally, federal student aid has focused on low-income students. However, the new federal Hope Scholarship is aimed squarely at middle-and upper-income families. In order to determine how middle-class aid programs will affect college attendance, I examine the impact on college attendance of the Georgia program that is the namesake and inspiration of the federal Hope Scholarship. I find that Georgia's program has increased the college attendance rate of J 8- to J 9-year-olds by 7 .5 to 8.3 percentage points. Among the subset of youth most likely to be eligible for the Georgia program, attendance has risen 10.9 percentage points. The increase is concentrated among Georgia's white students, who have experienced a J 2.3 percentage point rise in their enrollment rate. Black enrollment rates in Georgia are unaffected by the program. Chapter 3, written jointly with Jonathan Gruber, considers the ability of families to smooth consumption in the face of variable income. We find that families are fairly well able to smooth their consumption in the face of variable earnings. Roughly half of this consumption smoothing occurs through offsetting income flows, and, in particular, through the tax system; the other half comes through saving and dissaving. Government transfers play a particularly large role in smoothing consumption against income shocks caused by unemployment. / by Susan Marie Dynarski. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references (p. 70-72). / Three essays are presented which explore how strategic decision-making on a micro level translates into macro effects. Careful attention is paid to how asymmetric information and free-riding exert strong influences on t.he behavior of individuals. In the first chapter, a simple model of open-source software development is presented. It is found that either too little development or redundant development effort can occur. While any redundant research effort grows slowly with the size of the community, projects for which user valuations arc sufficiently extreme, such as solutions to the Year 2000 Computer Problem (Y2K), will result in significant waste relative to a traditional closed-source environment. Correlations between value and cost are shown to resolve the empirical puzzle as to why some extremely useful and fairly simple software docs not get written while more complex software sometimes docs. It is shown that a modular design can improve or worsen the performance of an open-source community. In the second chapter, an industry is considered in which new firms require time to learn whether they have the "right stuff" to grow in size and profitability in the long run. The critical input market ( that for skilled labor) is imperfectly competitive. By extending the literate on nonstationary dynamic bargaining, analysis is performed on a set of intertemporal externalities exerted by future parties on today's parties, and vice versa. The results suggest why, even if firms are able to write detailed contingent contracts with their current employees, inefficient levels of firm entry will generally exist. The theory also sheds some light on the continuing debate over the contribution of small firms to economic growth. In the third and final chapter, players in a. war of attrition care about the identity of the winner, even when they lose. In particular, a three player war of attrition is considered . Two "team" players enjoy a. fraction of their valuation when their partner wins. The remaining, ''solo" player benefits only by winning the war. Imposing team symmetry, the solo player drops out more quickly then either team player. The incentive to avoid fighting costs by free riding on a teammate is outweighed by a strategic commitment effect. Team players thus continue to fight even when they have no chance of winning a subsequent two-player subgame with the solo player. Examining limiting results, when the "caring coefficient" between t.he team players is small, a selection result obtains: The solo player drops out immediately, allowing the team players to then compete in a standard two-player war of attrition. / Justin Pappas Johnson. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references (p. 161-166). / The first chapter derives optimal income tax formulas using the concepts of compensated and uncompensated elasticities of earnings with respect to tax rates. This method of derivation casts new light on the original Mirrlees formulas of optimal taxation and can be easily extended to a heterogeneous population of taxpayers. A simple formula for optimal marginal rates for high income earners is derived as a function of the two elasticities of earnings and the thickness of the income distribution. Optimal income tax simulations are presented using empirical wage income distributions and a range of realistic elasticity parameters. The second chapter derives the non-linear income tax schedule which minimizes dead-weight burden without any regard for redistribution. The features of this problem are shown to be equivalent to the Mirrlees' optimal income tax problem. The tax schedule minimizing dead-weight burden is an optimal income tax schedule in which the government applies particular marginal welfare weights at each income level. In the case of no income effects, these marginal welfare weights are the same for everybody. The last chapter uses a panel of individual tax returns and the 'bracket creep' as source of tax rate variation to construct instrumental variables estimates of the sensitivity of income to changes in tax rates. Compensated elasticities can be estimated by comparing the differences in changes in income between taxpayers close to the top-end of a tax bracket to the other taxpayers. The elasticities found are higher than those derived in labor supply studies but smaller than those found previously with the same kind of tax returns data. / by Emmanuel Saez. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references. / This thesis consists of three essays. The first essay investigates what role employee stock options and CEO compensation have in explaining the surge in corporate share repurchases in the mid 1990s. Corporations may opt to fund options with repurchased shares to avoid the immediate dilution of earnings per share. Whether the top executives receive stock-based compensation may also influence distribution decisions. To test the importance of these two hypotheses, I collect data on stock option programs for over 800 U. S. corporations at the end of 1994. Estimates suggest that a firm with outstanding options representing 10% of shares outstanding will repurchase .9 percentage points more stock in 1995, as opposed to a firm with no option program. Once total outstanding options are controlled for, CEO options and option holdings of the top five executives are if anything negatively correlated with stock buybacks. Firms whose CEOs hold options are significantly more likely to retain earnings. The paper also considers what role the taxation of distributions has in explaining the growth of corporate share repurchases. The second essay examines how participant choice in pension plans affects household portfolios. Some retirement plans allow the participant to choose how funds are invested. Being exposed to historical differences in asset returns may provide the participant with financial education which would otherwise not be received. This paper finds that households covered with pension plans in which the employee must decide upon investments are significantly more apt to hold stock outside of their retirement plan relative to households with plans offering no choice. The third essay investigates the effect of specific features of the U.S. capital gains tax on turn-of-the-year stock returns. Both the fraction of long-term losses that are deductible from Adjusted Gross Income and the required holding period for long-term losses have changed over the past three decades. These changes alter the incentives for year-end capital loss realization for individual investors. This paper presents evidence that is consistent with the hypothesis that detailed provisions of the capital gains tax affect the link between past capital losses and turn-of-the- year stock returns. / by Scott J. Weisbenner. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, 1999. / Includes bibliographical references (p. 97-100). / This thesis consists of an examination of the role of consumer search in explaining price patterns in retail markets for prescriptions drugs {Chapters 1 and 2), as well as a separate study of hospital-insurer bargaining in the state of Connecticut (Chapter 3). Chapter 1 seeks to establish the empirical importance of price dispersion due to costly consumer search by examining retail prices for prescription drugs. Posted prices in two geographically distinct markets are shown to vary considerably across pharmacies within the same market, even after controlling for variation due to pharmacy differences. Pharmacy heterogeneity accounts for at most one third of the observed price dispersion. The empirical analysis hinges on the observation that consumers' incentives to price-shop depend on characteristics of the drug therapy. Cross-sectional patterns in price distributions across drugs are consistent with the predictions of a search model: prices for repeatedly purchased prescriptions (for which the expected benefits of search are highest) exhibit significant reductions in both dispersion and price-cost margins. The second chapter extends the analysis of prescription drug pricing by using detailed data on retail pharmacy transactions to make inferences about the nature and intensity of consumer search. Prescription prices exhibit patterns that should, in principle, induce search: in particular, prices vary widely across stores, and stores' price rankings are inconsistent across drugs (so that the low-price pharmacy is different for one prescription vs. another). Estimates from a model of pharmacy choice suggest that search intensities are generally low: I estimate that for a typical prescription, the fraction of consumers that price-shops is approximately .14. Variation in this estimated search intensity across drugs appears to be consistent with explanations based on rational search; for instance, price-shopping is more prevalent for maintenance medications than for one-time purchases. Comparisons of estimated search intensities with corresponding estimates of benefits imply that the cost of conducting a price-search is roughly $9.50 for the average consumer. Chapter 3 contains a separate study of hospital-insurer bargaining. I use unique data from the state of Connecticut to examine discounting patterns in the state's hospital industry for the years following deregulation (1994-1996). The data provide a rare opportunity to study payer level differences in negotiated discounts for hospital services. In addition to presenting descriptive evidence on how discounts vary across payers, payer types, and hospital types, this study uses matched revenue data to analyze the economic determinants of discount magnitudes. I find that HMOs negotiate discounts that are on average 14 percentage points larger than discounts received by traditional indemnity plans. Payers designated as PPOs negotiate discounts only 2-3 percentage points better than indemnity plans. Payer size appears to affect bargaining power, but the economic significance of the effect is rather small. The data are shown to exhibit patterns consistent with a bargaining explanation, in which discount magnitudes depend critically on payers' abilities to "channel" patients to hospitals with which favorable discounts have been negotiated. In contrast with the effects of payer size, the estimated impact of channeling ability on discount magnitudes is substantial. / by Alan T. Sorensen. / Ph.D.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, c1999. / Includes bibliographical references. / This thesis analyzes the micro-economic effects of different government programs on the savings and labor supply behavior of Mexican households and individuals. The first two chapters deal with household saving behavior and government intervention. Chapter 1 assesses the impact of increasing financial access on low-income people saving rates and on different informal savings instruments. It considers an exogenous expansion of a Mexican savings institute that targets low-income people and uses the 1992 and 1994 National Surveys of Income and Expenditures. Results show that the expansion increased the average saving rate of affected households by more than 3 to almost 5 percentage points of income. Evidence of crowding out of informal savings instruments caused by the expansion is limited. Chapter 2 investigates the effects of the country's financial reform in the early 1990's on households' saving rates. It uses the 1989 and 1992 Mexican National Surveys of Income and Expenditures. Households had different degrees of exposure to the financial reform depending on their income level and location. Results indicate that households located in cities, which are more likely to have financial intermediaries, reduced their saving rate significantly after the financial reform. The effects were significantly stronger among richer households. Findings are consistent with the hypothesis that the financial reform reduced borrowing constraints among younger households. Finally, chapter 3 evaluates the effects of a Mexican training program of the unemployment spells of trainees. This program consists of training courses on several types of activities. Non-parametric estimations and different hazard rate models are estimated. To tackle selection biases, a propensity score procedure was calculated. Parametric results for the complete database show that training increaser. the hazard rate of leaving unemployment by 20 percent to 60 percent, depending on the type of training. Average effects hide substantial variation among men and women. The training effect is in general non-significant for men, while it is highly significant for women. / by Fernando Aportela. / Ph.D.