• Selected
• Entire List

Abstract:
This paper uses data on juvenile offenders released from correctional facilities in Florida to explore the effects of facility management type (private for-profit, private nonprofit, public state-operated, and public county-operated) on recidivism outcomes and costs. The data provide detailed information on individual characteristics, criminal and correctional histories, judge-assigned restrictiveness levels, and home zip codes - allowing us to control for the non-random assignment of individuals to facilities far better than any previous study. Relative to all other management types, for-profit management leads to a statistically significant increase in recidivism, but relative to nonprofit and state-operated facilities, for-profit facilities operate at a lower cost to the government per comparable individual released. Cost-benefit analysis implies that the short-run savings offered by for-profit over nonprofit management are negated in the long run due to increased recidivism rates, even if one measures the benefits of reducing criminal activity as only the avoided costs of additional confinement.

Juvenile Crime, Juvenile Correctional Facilities, Recidivism, Prison Privatization, Provision of Public Goods: Nonprofit, For-profit, Public

Abstract:
This paper presents a new equilibrium framework for analyzing
economic and policy questions related to the sorting of
households within a large metropolitan area. We estimate the
model using restricted-access Census data that precisely
characterize residential and employment locations for
households in the San Francisco Bay Area, yielding accurate measures of preferences for a wide variety of housing and
neighborhood attributes across different types of household.
We use these estimates to explore the causes and consequences
of racial segregation in general equilibrium. Our results
indicate that, given the preference structure of households in
the Bay Area, the elimination of racial differences in income
and wealth would significantly increase the residential
segregation of each major racial group, as the equalization of
income leads, for example, to the formation of new wealthy,
segregated Black and Hispanic neighborhoods. We also provide
evidence that sorting on the basis of race itself (whether
driven by preferences or discrimination) leads to large
reductions in the consumption of housing, public safety, and
school quality by Black and Hispanic households.

Segregation, Sorting, Housing Markets, Locational Equilibrium, Residential Choice, Discrete Choice

Abstract:
This paper analyzes the influence that juvenile offenders serving time in the same correctional facility have on each others subsequent criminal behavior. The analysis is based on data on over 8,000 individuals serving time in 169 juvenile correctional facilities during a two-year period in Florida. These data provide a complete record of past crimes, facility assignments, and arrests and adjudications in the year following release for each individual. To control for the non-random assignment to facilities, we include facility and facility-by-prior offense fixed effects, thereby estimating peer effects using only within-facility variation over time. We find strong evidence of peer effects for burglary, petty larceny, felony and misdemeanor drug offenses, aggravated assault, and felony sex offenses; the influence of peers primarily affects individuals who already have some experience in a particular crime category. We also find evidence that the predominant types of peer effects differ in residential versus non-residential facilities; effects in the latter are consistent with network formation among youth serving time close to home.

Abstract:
This paper analyzes the influence that juvenile offenders serving time in the same correctional facility have on each other's subsequent criminal behavior. The analysis is based on data on over 8,000 individuals serving time in 169 juvenile correctional facilities during a two-year period in Florida. These data provide a complete record of past crimes, facility assignments, and arrests and adjudications in the year following release for each individual. To control for the non-random assignment to facilities, we include facility fixed effects, thereby estimating peer effects using only within-facility variation over time. We find strong evidence of peer effects for various categories of theft, burglary, and felony drug and weapon crimes; the influence of peers primarily affects individuals who already have some experience in a particular crime category. We also find evidence that peer effects are stronger in smaller facilities and that the predominant types of peer effects differ in residential versus non-residential facilities. Effects in the latter are consistent with network formation among youth serving time close to home.

social learning, peer effects, social interactions, recidivism, juvenile crime, human capital accumulation

Abstract:
This paper sheds new light on the forces that drive residential segregation on the basis of race, assessing the extent to which across-race differences in other household characteristics can
explain a significant portion of observed racial segregation.
The central contribution of the analysis is to provide a
transparent new measurement framework for understanding
segregation patterns. This framework allows researchers to
characterize patterns of segregation, to decompose them in
meaningful ways, and to carry out partial equilibrium
counterfactuals that illuminate the contributions of a variety
of non-race characteristics in driving segregation. We
illustrate our approach using restricted micro-Census data
from the San Francisco Bay Area that provide a rich joint
distribution of household and neighborhood characteristics not
previously available to the research community. In contrast to
findings in the prior literature, our analysis indicates that
individual household characteristics can explain a considerable fraction of segregation by race, explaining almost 95% of
segregation for Hispanic, over 50% for Asian, and 30% for
White and Black households.

Residential Segregation, Racial Segregation, Sorting, Housing Markets

Abstract:
This paper develops a comprehensive framework for estimating household preferences for school and neighborhood attributes in the presence of sorting. It embeds a boundary discontinuity design in a heterogeneous residential choice model addressing the endogeneity of school and neighborhood characteristics. The model is estimated using restricted-access Census data from a large metropolitan area, yielding a number of new results. First, households are willing to pay less than 1 percent more in house prices - substantially lower than previous estimates - when the average performance of the local school increases by 5 percent. Second, much of the apparent willingness to pay for more educated and wealthier neighbors is explained by the correlation of these sociodemographic measures with unobserved neighborhood quality. Third, neighborhood percent black is not capitalized directly into housing prices; instead, the negative correlation of neighborhood race and housing prices is due entirely to the fact that blacks live in unobservably lower-quality neighborhoods. Finally, there is considerable heterogeneity in preferences for schools and neighbors, with households preferring to self-segregate on the basis of both race and education.

Abstract:
This paper develops a comprehensive framework for estimating household preferences for school and neighborhood attributes in the presence of sorting. It embeds a boundary discontinuity design in a heterogeneous model of residential choice to address the endogeneity of school and neighborhood attributes. The model is estimated using restricted-access Census data from a large metropolitan area, yielding a number of new results. First, households are willing to pay less than one percent more in house prices -- substantially lower than previous estimates -- when the average performance of the local school increases by five percent. Second, much of the apparent willingness to pay for more educated and wealthier neighbors is explained by the correlation of these sociodemographic measures with unobserved neighborhood quality. Third, neighborhood race is not capitalized directly into housing prices; instead, the negative correlation of neighborhood race and housing prices is due entirely to the fact that blacks live in unobservably lower quality neighborhoods. Finally, there is considerable heterogeneity in preferences for schools and neighbors: in particular, we find that households prefer to self-segregate on the basis of both race and education.

Abstract:
This paper sets out a framework for estimating household preferences over a broad range of housing and neighborhood characteristics, some of which are determined by the way that households sort in the housing market. This framework brings together the treatment of heterogeneity and selection that has been the focus of the traditional discrete choice literature with a clear strategy for dealing with the correlation of unobserved neighborhood quality with both school quality and neighborhood sociodemographics. We estimate the model using rich data on a large metropolitan area, drawn from a restricted version of the Census. The estimates indicate that, on average, households are willing to pay an additional one percent in house prices - substantially lower than in prior work - when the average performance of the local school is increased by 5 percent. There is also evidence of considerable preference heterogeneity. We also show that the full capitalization of school quality into housing prices is typically 70-75 percent greater than the direct effect as the result of a social multiplier, neglected in the prior literature, whereby increases in school quality also raises prices by attracting households with more education and income to the corresponding neighborhood.

Capitalization, Local Public Goods, School Quality, Discrete Choice Models, Hedonic Price Regression, Education Demand

Abstract:
We use a novel dataset and research design to empirically detect the effect of social interactions among neighbors on labor market outcomes. Specifically, using Census data that characterize residential and employment locations down to the city block, we examine whether individuals residing in the same block are more likely to work together than those in nearby blocks. We find evidence of significant social interactions operating at the block level: residing on the same versus nearby blocks increases the probability of working together by over 33 percent. The results also indicate that this referral effect is stronger when individuals are similar in sociodemographic characteristics (e.g., both have children of similar ages) and when at least one individual is well attached to the labor market. These findings are robust across various specifications intended to address concerns related to sorting and reverse causation. Further, having determined the characteristics of a pair of individuals that lead to an especially strong referral effect, we provide evidence that the increased availability of neighborhood referrals has a significant impact on a wide range of labor market outcomes including employment and wages.

Neighborhood Effects, Job Referrals, Social Interactions, Social Networks, Labor

Abstract:
We use a novel research design to empirically detect the effect of social interactions among neighbors on labor market outcomes. Specifically, using Census data that characterize residential and employment locations down to the city block, we examine whether individuals residing in the same block are more likely to work together than those in nearby blocks. We find evidence of significant social interactions operating at the block level: residing on the same versus nearby blocks increases the probability of working together by over 33 percent. The results also indicate that this referral effect is stronger when individuals are similar in socio-demographic characteristics (e.g., both have children of similar ages) and when at least one individual is well attached to the labor market. These findings are robust across various specifications intended to address concerns related to sorting and reverse causation. Further, having determined the characteristics of a pair of individuals that lead to an especially strong referral effect, we provide evidence that the increased availability of neighborhood referrals has a significant impact on a wide range of labor market outcomes including labor force participation, hours and earnings.

Abstract:
In many theoretical public finance models, school quality plays a central role as a determinant of household location choices and in turn, of neighborhood stratification. In contrast, the recent empirical literature has almost universally concluded that the direct effect of school quality on housing demand is weak, a conclusion that is robust across a variety of research designs. Using an equilibrium model of residential sorting, this paper closes the gap between these literatures, providing clear evidence that the full effect of school quality on residential sorting is significantly larger than the direct effect - four times as great for education stratification, twice for income stratification. This is due to a strong social multiplier associated with heterogeneous preferences for peers and neighbors; initial changes in school quality set in motion a process of re-sorting on the basis of neighborhood characteristics that reinforces itself, giving rise to substantially larger stratification effects.

Abstract:
In many theoretical public finance models, school quality plays a central role as a determinant of household location choices and in turn, of neighborhood stratification. In contrast, the recent empirical literature has almost universally concluded that the direct effect of school quality on housing demand is weak, a conclusion that is robust across a variety of research designs. Using an equilibrium model of residential sorting, this paper closes the gap between these literatures, providing clear evidence that the full effect of school quality on residential sorting is significantly larger than the direct effect - four times as great for education stratification, twice for income stratification. This is due to a strong social multiplier associated with heterogeneous preferences for peers and neighbors; initial changes in school quality set in motion a process of re-sorting on the basis of neighborhood characteristics that reinforces itself, giving rise to substantially larger stratification effects.

Tiebout sorting, neighborhood stratification, social multiplier, school quality, education demand, capitalization, neighborhood choice, general equilibrium

Abstract:
Standard intuition suggests that residential segregation in the United States will decline when racial inequality narrows. In this paper, we hypothesize that the opposite will occur. We note that middle-class black neighborhoods are in short supply in many U.S. metropolitan areas, forcing highly educated blacks either to live in predominantly white high-socioeconomic status (SES) neighborhoods or in more black lower-SES neighborhoods. Increases in the proportion of highly educated blacks in a metropolitan area may then lead to the emergence of new middle-class black neighborhoods, causing increases in residential segregation. We formalize this mechanism using a simple model of residential choice that permits endogenous neighborhood formation. Our primary empirical analysis, based on across-MSA evidence from the 2000 Census, indicates that this mechanism does indeed operate: as the proportion of highly educated blacks in an MSA increases, so the segregation of blacks at all education levels increases. Time-series evidence provides additional support for the hypothesis, showing that an increase in black educational attainment in a metropolitan area between 1990-2000 significantly increases segregation. Our analysis has important implications for the evolution of both residential segregation and racial socioeconomic inequality, drawing attention to a negative feedback loop likely to inhibit reductions in segregation and racial inequality over time.

segregation, racial inequality, racial sorting, neighborhood formation

Abstract:
In contrast to conventional wisdom, this paper identifies a powerful mechanism which can lead to persistent and even increasing residential segregation when racial differences in education and other sociodemographics narrow. We document that middle-class black neighborhoods are in short supply in many U.S. metropolitan areas, forcing highly educated blacks either to live in white neighborhoods with high amenity levels or in more black neighborhoods with lower amenity levels. A simple model then shows that increases in the proportion of highly educated blacks in a metropolitan area may lead to the emergence of new middle-class black neighborhoods, relieving the prior neighborhood supply constraint and causing increases in residential segregation. Cross- MSA evidence from the 2000 Census indicates that this mechanism does in fact operate: as the proportion of highly educated blacks in an MSA increases, so the segregation of educated blacks and blacks more generally goes up. Our empirical findings are robust and have important implications for the evolution of residential segregation.

Abstract:
Black households in the United States with high levels of income and education (SES) typically face a stark tradeoff when deciding where to live. They can choose neighborhoods with high levels of public goods or a high proportion of blacks, but very few neighborhoods combine both, a fact we document clearly. In the face of this constraint, we conjecture that racial sorting may dramatically lower the consumption of local public goods by high-SES blacks. To shed light on this, we estimate a model of residential sorting using unusually detailed restricted Census microdata, then use the estimated preferences to simulate a counterfactual world in which racial factors play no role in household residential location decisions. Results from this exercise provide the first evidence that sorting on the basis of race gives rise to significant reductions in the consumption of local public goods by black and high-SES black households in particular. These consumption effects lead to significant losses of welfare and are likely to have important intergenerational implications.

Abstract:
This paper studies the causes and consequences of racial segregation using a new general equilibrium model that treats neighborhood compositions as endogenous. The model is estimated using unusually detailed restricted Census microdata covering the entire San Francisco Bay Area, and in combination with a rich array of econometric estimates, serves as a powerful tool for carrying out counterfactual simulations that shed light on the causes and consequences of segregation. In terms of causes, and contrasting with prior research, our GE simulations indicate that equalizing income and education across race would be unlikely to result in significant reductions in racial segregation, as minority households would sort into newly formed minority neighborhoods. Indeed, among Asian and Hispanic households, segregation increases. In terms of consequences, this paper provides the first evidence that sorting on the basis of race gives rise to significant reductions in the consumption of local public goods by minority households and upper-income minority households in particular. These consumption effects are likely to have important intergenerational implications.

Segregation, General Equilibrium, Endogenous Sorting, Urban Housing Market, Locational Equilibrium, Counterfactual Simulation, Discrete Choice

Abstract:
While there is growing interest in measuring the size and scope of local spillovers, it is well understood that such spillovers cannot be distinguished from unobservable local attributes using solely the observed location decisions of individuals or firms. We propose an empirical strategy for recovering estimates of spillovers in the presence of unobserved local attributes for a broadly applicable class of equilibrium sorting models. Our approach relies on an IV strategy derived from the internal logic of the sorting model itself. We show practically how the strategy is implemented, provide intuition for our instruments, discuss the role of effective choice-set variation in identifying the model, and carry-out a series of Monte Carlo simulations to demonstrate performance in small samples.

Abstract:
With the growing recognition of the role played by geography
in all sorts of economic problems, there is strong interest
in measuring the size and scope of local spillovers (i.e.,
simple anonymous agglomeration or congestion effects, or more
complicated interactions between individuals or firms of
specific types). It is well-understood, however, that such
spillovers cannot be distinguished from unobservable local
attributes using just the observed location decisions of
individuals or firms. We propose an empirical strategy for
recovering estimates of spillovers in the presence of
unobserved local attributes for a broadly applicable class
of equilibrium sorting models. This approach relies on an
instrumental variables strategy derived from the internal
logic of the sorting model itself. We show practically how
the strategy is implemented, provide intuition for our
instrumental variables, and discuss the role of effective
choice-set variation in identifying the model, and carry-out
a series of Monte Carlo experiments to demonstrate the
instruments' performance in small samples.

Local Spillovers, Location Choice, Economic Geography, Natural Advantage, Social Interactions, Network Effects, Endogenous Sorting, Discrete Choice Models, Agglomeration, Congestion

Abstract:
A central feature of many models of location choice - whether
of firms or households, within or across cities - is the role
of local interactions or spillovers, whereby the payoffs from
choosing a location depend in part on the number or attributes
of other individuals or firms that choose the same or nearby
locations in equilibrium. The main goal of this paper is to
develop the equilibrium properties of a broadly applicable and
readily estimable class of sorting models that allow the
location decision to depend on both fixed local attributes
(including unobserved attributes) and such local interactions.
In particular, we prove uniqueness in the case of congestion
effects and use a series of simulations to demonstrate that a
unique equilibrium is more likely to obtain (i) the smaller are any agglomeration effects, (ii) the larger are the set of
choices available to the agents, (iii) the more "meaningful
variation" there is in those choices, and (iv) the more
heterogeneous are the agents themselves. This is encouraging
for the use of our model to describe the sorting of individuals and firms over geographic space, where the number of choices is usually large and variation in exogenous fixed attributes can
be important. Moreover, these results conveniently coincide
with the conditions required for econometric identification of
our model.

Local Spillovers, Social Interactions, Economic Geography, Natural Advantage, Endogenous Sorting, Discrete Choice Models, Agglomeration, Congestion, Random Utility

Abstract:
Researchers have long recognized that the non-random sorting of individuals into groups generates correlation between individual and group attributes that is likely to bias naïve estimates of both individual and group effects. This paper proposes a non-parametric strategy for identifying these effects in a model that allows for both individual and group unobservables, applying this strategy to the estimation of neighborhood effects on labor market outcomes. The first part of this strategy is guided by a robust feature of the equilibrium in the canonical vertical sorting model of Epple and Platt (1998), that there is a monotonic relationship between neighborhood housing prices and neighborhood quality. This implies that under certain conditions a non-parametric function of neighborhood housing prices serves as a suitable control function for the neighborhood unobservable in the labor market outcome regression. This control function converts the problem to a model with one unobservable so that traditional instrumental variables solutions may be applied. In our application, we instrument for each individual's observed neighborhood attributes with the average neighborhood attributes of a set of observationally identical individuals. The neighborhood effects model is estimated using confidential microdata from the 1990 Decennial Census for the Boston MSA. The results imply that the direct effects of geographic proximity to jobs, neighborhood poverty rates, and average neighborhood education are substantially larger than the conditional correlations identified using OLS, although the net effect of neighborhood quality on labor market outcomes remains small. These findings are robust across a wide variety of specifications and robustness checks.

Abstract:
Researchers have long recognized that the non-random sorting of individuals into groups generates correlation between individual and group attributes that is likely to bias naïve estimates of both individual and group effects. This paper proposes a non-parametric strategy for identifying these effects in a model that allows for both individual and group unobservables, applying this strategy to the estimation of neighborhood effects on labor market outcomes. The first part of this strategy is guided by a robust feature of the equilibrium in vertical sorting models - a monotonic relationship between neighborhood housing prices and neighborhood quality. This implies that under certain conditions a non-parametric function of neighborhood housing prices serves as a suitable control function for the neighborhood unobservable in the labor market outcome regression. This control function transforms the problem to a model with one unobservable so that traditional instrumental variables solutions may be applied. In our application, we instrument for each individuals observed neighborhood attributes with the average neighborhood attributes of a set of observationally identical individuals. The neighborhood effects model is estimated using confidential microdata from the 1990 Decennial Census for the Boston MSA. The results imply that the direct effects of geographic proximity to jobs, neighborhood poverty rates, and average neighborhood education are substantially larger than the conditional correlations identified using OLS, although the net effect of neighborhood quality on labor market outcomes remains small. These findings are robust across a wide variety of specifications and robustness checks.

Abstract:
We examine the effects of education on financial decision-making skills by identifying an interesting source of variation in pertinent training. During the 1990s, an increasing number of individuals were exposed to programs of financial education provided by their employers. If, as some have argued, low saving frequently results from a failure to appreciate economic vulnerabilities, then education of this form could prove to have a powerful effect on rates of behavior. The current paper undertakes an analysis of these programs using a previously unexploited survey of employers. We find that both participation in and contributions to voluntary savings plans are significantly higher when employers offer retirement seminars. The effect is typically much stronger for non-highly compensated employees than for highly compensated employees. The frequency of seminars emerges as a particularly important correlate of behavior. We are unable to detect any effects of written materials, such as newsletters and summary plan descriptions, regardless of frequency. We also present evidence on other determinants of plan activity.

Abstract:
This paper introduces an equilibrium framework for analyzing residential sorting, designed to take advantage of newly available restricted-access Census microdata. The framework adds an equilibrium concept to the discrete choice framework developed by McFadden (1973, 1978), permitting a more flexible characterization of preferences than has been possible in previously estimated sorting models. Using data on nearly a quarter of a million households residing in the San Francisco Bay Area in 1990, our estimates provide a precise characterization of preferences for many housing and neighborhood attributes, showing how demand for these attributes varies with a household's income, race, education, and family structure. We use the equilibrium model in combination with these estimates to explore the effects of an increase in income inequality, the findings indicating that much of the increased spending power of the rich is absorbed by higher housing prices.

Abstract:
Conventional hedonic techniques for estimating the value of local amenities rely on the assumption that households move freely among locations. We show that when moving is costly, the variation in housing prices and wages across locations may no longer reflect the value of differences in local amenities. We develop an alternative discrete-choice approach that models the household location decision directly, and we apply it to the case of air quality in U.S. metro areas in 1990 and 2000. Because air pollution is likely to be correlated with unobservable local characteristics such as economic activity, we instrument for air quality using the contribution of distant sources to local pollution - excluding emissions from local sources, which are most likely to be correlated with local conditions. Our model yields an estimated elasticity of willingness to pay with respect to air quality of 0.34 to 0.42. These estimates imply that the median household would pay $149 to $185 (in constant 1982-1984 dollars) for a one-unit reduction in average ambient concentrations of particulate matter. These estimates are three times greater than the marginal willingness to pay estimated by a conventional hedonic model using the same data. Our results are robust to a range of covariates, instrumenting strategies, and functional form assumptions. The findings also confirm the importance of instrumenting for local air pollution.

Abstract:
Prompted by widespread concerns about public school quality, a growing empirical literature has measured the effects of greater choice on school performance. This paper contributes to that literature in three ways. First, it makes the observation that the overall effect of greater choice, which has been the focus of prior research, can be decomposed into demand and supply components: knowing the relative sizes of the two is very relevant for policy. Second, using rich data from a large metropolitan area, it provides a direct and intuitive measure of the competition each school faces. This takes the form of a school-specific elasticity that measures the extent to which reductions in school quality would lead to reductions in demand. Third, the paper provides evidence that these elasticity measures are strongly related to school performance: a one-standard deviation increase in the competitiveness of a school's local environment within the Bay Area leads to a 0.15 standard deviation increase in average test scores. This positive correlation is robust and is consistent with strong supply responsiveness on the part of public schools, of relevance to the broader school choice debate.

Abstract:
In cities throughout the United States, blacks tend to live in significantly poorer and lower-amenity neighborhoods than whites. An obvious first-order explanation for this is that an individual's race is strongly correlated with socioeconomic status (SES), and poorer households can only afford lower quality neighborhoods. This paper conjectures that another explanation may be as important. The limited supply of high-SES black neighborhoods in most U.S. metropolitan areas means that neighborhood race and neighborhood quality are explicitly bundled together. In the presence of any form of segregating preferences, this bundling raises the implicit price of neighborhood amenities for blacks relative to whites, prompting our conjecture - that racial differences in the consumption of neighborhood amenities are significantly exacerbated by sorting on the basis of race, given the small numbers of blacks and especially high-SES blacks in many cities. To provide evidence on this conjecture, we estimate an equilibrium sorting model with detailed restricted Census microdata and use it to carry out informative counterfactual simulations. Results from these indicate that racial sorting explains a substantial portion of the gap between whites and blacks in the consumption of a wide range of neighborhood amenities - in fact, as much as underlying socioeconomic differences across race. We also show that the adverse effects of racial sorting for blacks are fundamentally related to the small proportion of blacks in the U.S. metropolitan population. These results emphasize the significant role of racial sorting in the inter-generational persistence of racial differences in education, income, and wealth.

Abstract:
In traditional signaling models, education provides a way for individuals to sort themselves by ability. Employers in turn use education to statistically discriminate, paying wages that reflect the average productivity of workers with the same given level of education. In this paper, we provide evidence that education (specifically, attending college) plays a much more direct role in revealing ability to the labor market. We use the NLSY79 to examine returns to ability early in careers; our results suggest that ability is observed nearly perfectly for college graduates but is revealed to the labor market much more gradually for high school graduates. As a result, from very beginning of the career, college graduates are paid in accordance with their own ability, while the wages of high school graduates are initially completely unrelated to their own ability. This view of ability revelation in the labor market has considerable power in explaining racial differences in wages, education, and the returns to ability. In particular, we find no racial differences in wages or returns to ability in the college labor market, but a 6-10 percent wage penalty for blacks (conditional on ability) in the high school market. These results are consistent with the notion that employers use race to statistically discriminate in the high school market but have no need to do so in the college market. That blacks face a wage penalty in the high school but not the college labor market also helps to explains why, conditional on ability, blacks are more likely to earn a college degree, a fact that has been documented in the literature but for which a full explanation has yet to emerge.

Abstract:
This paper considers nonparametric identification and estimation of a generalized Roy model that includes a non-pecuniary component of utility associated with each choice alternative. Previous work has found that, without parametric restrictions or the availability of covariates, all of the useful content of a cross-sectional dataset is absorbed in a restrictive specification of Roy sorting behavior that imposes independence on wage draws. While this is true, we demonstrate that it is also possible to identify (under relatively innocuous assumptions and without the use of covariates) a common non-pecuniary component of utility associated with each choice alternative. We develop nonparametric estimators corresponding to two alternative assumptions under which we prove identification, derive asymptotic properties, and illustrate small sample properties with a series of Monte Carlo experiments. We demonstrate the usefulness of one of these estimators with an empirical application. Micro data from the 2000 Census are used to calculate the returns to a college education. If high-school and college graduates face different costs of migration, this would be reflected in different degrees of Roy-sorting-induced bias in their observed wage distributions. Correcting for this bias, the observed returns to a college degree are cut in half.

Abstract:
We examine the effects of education on financial decision-making skills by identifying an interesting source of variation in pertinent training. During the 1990s, an increasing number of individuals were exposed to programs of financial education provided by their employers. If, as some have argued, low saving frequently results from a failure to appreciate economic vulnerabilities, then education of this form could prove to have a powerful effect on behavior. The current article undertakes an analysis of these programs using a previously unexploited survey of employers. We find that both participation in and contributions to voluntary savings plans are significantly higher when employers offer retirement seminars. The effect is typically much stronger for nonhighly compensated employees than for highly compensated employees. The frequency of seminars emerges as a particularly important correlate of behavior. We are unable to detect any effects of written materials, such as newsletters and summary plan descriptions, regardless of frequency. We also present evidence on other determinants of plan activity.