Economics Faculty Scholarship

Measurement Error in Earnings Data in the Health and Retirement Study

Jessie Bricker et al. — Tue, 15 Jan 2013 10:37:10 PST

We provide new evidence on the extent of measurement error in respondent-reported earnings data by exploiting detailed W-2 records matched to older workers in the Health and Retirement Study (HRS). Our empirical findings are qualitatively consistent with the findings of previous studies. Mean measurement error in the 1991 HRS earnings data for men is somewhat larger than what has been found in other validation studies, but is still modest, averaging about 0.059 log points, approximately 5.9 percent, or $1,500. For women in 1991, it is 0.067 log points, approximately 6.7 percent, or $916. We find a negative correlation between the measurement error and the true value of earnings as measured by the W-2 records, which indicates the presence of non-classical measurement error. For men and women, this error shows little correlation with a standard set of cross-sectional earnings determinants. The one exception is that the measurement error rises with reported education. The bias on the OLS parameter estimate of the impact of having a college degree or higher (relative to a high school drop-out) from using the respondent-reported rather than the W-2 earnings is positive and estimated to be 0.071 log points, or roughly a bias of 7 percent.

Option Value and Dynamic Programming Model Estimates of Social Security Disability Insurance Application Timing

Rickard V. Burkhauser et al. — Tue, 15 Jan 2013 10:37:09 PST

This paper develops dynamic structural models - an option value model and a dynamic programming model - of the Social Security Disability Insurance (SSDI) application timing decision. We estimate the time to application from the point at which a health condition first begins to affect the kind or amount of work that a currently employed person can do. We use Health and Retirement Study (HRS) and restricted access Social Security earnings data for estimation. Based on tests of both in-sample and out-of-sample predictive accuracy, our option value model performs better than both our dynamic programming model and our reduced form hazard model.

Dynamic Modeling of the SSDI Application Timing Decision: The Importance of Policy Variables

Rickard V. Burkhauser et al. — Tue, 15 Jan 2013 10:37:08 PST

This paper analyzes the importance of policy variables in the context of Social Security Disability Insurance (SSDI) application timing decision. Previously, we explicitly modeled the optimal timing of SSDI application using dynamic structural models. We estimated these models using data from the Health and Retirement Study (HRS). This paper uses option value model estimates to simulate application timing under alternative SSDI policy formulations. We consider changes in three policy variables: benefit levels, acceptance rates, and employer accommodation. Our simulations suggest all these changes would have substantial effects on expected spell lengths until application and on lifetime application rates, and hence on SSDI caseloads.

Pet Econometrics: Ownership of Cats and Dogs

Patrali Chatterjee et al. — Tue, 15 Jan 2013 10:37:07 PST

Using a bivariate ordered probit model, this paper examines the effects of house ownership, housing tenure, income, household size and composition, and the occupation and education of the make and female head of household on ownership of cats and dogs. Large households have more of both, but other factors differ between cat and dog ownership. Women's opportunity costs and missing female adults encourage cat ownership. Cats and dogs are slightly complementary. A GMM demographic specification test does not reject normality. The data are taken from a demographic scanner panel data set developed for a marketing study.

Slippery When Wet: The Effects of Local Alcohol Access Laws on Highway Safety

Reagan Anne Baughman et al. — Tue, 15 Jan 2013 10:37:05 PST

Using detailed panel data on local alcohol policy changes in Texas, this paper tests whether the effect of these changes on alcohol-related accidents depends on whether the policy change involves where the alcohol is consumed and the type of alcohol consumed. After controlling for both county and year fixed effects, we find evidence that: (i) the sale of beer and wine may actually decrease expected accidents; and (ii) the sale of higher alcohol-content liquor may present greater risk to highway safety than the sale of just beer and wine.

Entry Deterring Capacity in the Texas Lodging Industry

Michael Conlin et al. — Tue, 15 Jan 2013 10:37:04 PST

We test whether capacity is used to deter entry and whether the amount invested in entry deterring capacity is related to market concentration and market presence. We use a unique dataset containing all lodging properties in Texas from 1991 through 1997. For each of the 3,830 properties, we have information on occupancy rate, number of rooms, location and ownership. This information is augmented by market level information such as tax rates, travel expenditures and retail wages. We find that there is higher investment in capacity relative to demand (i.e. idle capacity) in markets with larger Herfindahl index and by firms with larger share of market capacity. These results are consistent with the entry deterrence literature that suggests firms in more concentrated markets and firms with larger market share have greater incentive to invest in entry deterring capacity.

Love at What Price? Estimating the Value of Marriage

Michael Conlin et al. — Tue, 15 Jan 2013 10:37:03 PST

Using a law within Social Security that provides clear financial incentives to delay marriage, we estimate the financial value of a month of marriage. Specifically, the law provides that widows who are eligible for Social Security benefits on their deceased spouse's earnings records are eligible for benefits at age 60, unless they remarry before that age. If they remarry before that age, they cannot claim widow benefits and must wait until at least age 62 to claim spousal benefits on their new husband's record, which are typically less generous than widow benefits. To generate an estimate of what this behavior implies about the value of marriage, we use data from five panels of the Survey of Income and Program Participation linked to administrative data from Social Security. We estimate the cost of marrying before age 60 imposed by the Social Security program. We develop a model that reflects the institutional details of Social Security and generate a likelihood function that reflects that model. By taking advantage of the variation in these costs and when or whether widows remarry before age 60, we estimate the benefit of marriage to be $8000/month. These estimates appear to be reasonable in the context of the short length of time widows are willing to wait and the high value of Social Security benefits.

Survive Then Thrive: Talent, Research Motivation, and Completing the Economics Ph.D.

Wayne A. Grove et al. — Tue, 15 Jan 2013 10:37:01 PST

This study investigates the completion of the Ph.D. in Economics. We use ex ante information, based solely upon reviewing a set of individual applications from former doctoral students. Estimation for determining success is done by logit, multinomial logit, and generalized ordered logit. We find that students need different skills and attributes to succeed at each distinct and sequential stage of the doctoral program. Significant determinants for passing the comprehensive exams include high GRE verbal and quantitative scores, a Masters degree, and a prior focus on economics. Research motivation and math preparation play significant roles in completing the dissertation, but having a Masters degree and economics preparation becomes insignificant. GRE scores disappear as a significant determinant for completion in the generalized ordered logit estimates, which emphasize the sequential nature of the Economics Ph.D. program.

Improving Economic Literacy: The Role of Concurrent Enrollment Programs

Donald H. Dutkowsky et al. — Tue, 15 Jan 2013 10:37:00 PST

This paper introduces Concurrent Enrollment Programs (CEPs), within the context of Syracuse University Project Advance (PA) Economics, as a way to improve economic literacy. We describe measures implemented to operate PA Economics as a high-quality CEP, as well as the National Alliance of Concurrent Education Partnerships to establish national standards. This study also contains results from administering to high school students taking PA Economics the nationally normed Test of Economic Literacy (TEL). PA students average nearly one percentage point higher than the AP/Honors Economics Group, and score considerably better than AP/Honors Economics in fundamentals and international economics. By cognitive level, PA Economics students score over four percentage points higher on questions than AP/Honors Economics in the knowledge area, and the findings present evidence of better performance on application questions. PA Economics students also average over seventeen percentage points better than those taking the TEL in AP/Honors Social Studies courses.

Interest on Bank Reserves and Optimal Sweeping

David D. VanHoose et al. — Tue, 15 Jan 2013 10:36:59 PST

This paper utilizes a profit maximizing banking model to analyze sweeping behavior. Comparative statics results indicate that sweeping responds positively to increases in bank loan rates and reserve ratios and negatively to increases in the interest rate on reserves or to exogenous increases in bank deposits or equity. Sweeping generates greater responsiveness in lending to changes in loan rates or the interest rate on reserves and lower responsiveness to exogenous changes in reserve ratios or equity. Empirical analysis of an explicit condition that we derive relating sweeping to the interest rate on reserves suggests with an unchanged reserve requirement, the Fed could eliminate sweeping by setting the interest rate on reserves to no less than 3.67 percentage points below the market loan rate. The range of interest rates on reserves that lead to zero sweeping increases sharply, however, as the required reserve ratio is reduced.

What’s Wrong with Economics? It Ignores the Pogo Principle: 'We Have Met the Enemy and He is Us'

Jerry Evensky — Tue, 15 Jan 2013 10:36:57 PST

The piece begins with the proposition that the economic perspective on human activity must reflect the fact that human beings transact in a world defined for the actors by social norms. An analysis of the crisis of 2008 is offered as a demonstration of the value of adopting such a broader perspective. Part two offers a historical model based on Adam Smith’s Moral Philosophy of such a broader analysis. The piece closes with the case that the history of ideas offers alternative perspectives on the questions we explore in economics today and thus can serve as a valuable resource for constructive skepticism of the current discourse.

Bandwidth Selection and the Estimation of Treatment Effects with Unbalanced Data

Jose Galdo et al. — Tue, 15 Jan 2013 10:36:56 PST

This paper addresses the selection of smoothing parameters for estimating the average treatment effect on the treated using matching methods. Because precise estimation of the expected counterfactual is particularly important in regions containing the mass of the treated units, we define and implement weighted cross-validation approaches that improve over conventional methods by considering the location of the treated units in the selection of the smoothing parameters. We also implement a locally varying bandwidth method that uses larger bandwidths in areas where the mass of the treated units is located. A Monte Carlo study compares our proposed methods to the conventional unweighted method and to a related method inspired by Bergemann et al. (2005). The Monte Carlo analysis indicates efficiency gains from all methods that take account of the location of the treated units. We also apply all five methods to bandwidth selection in the context of the data from LaLonde's (1986) study of the performance of non-experimental estimators using the experimental data from the National Supported Work (NSW) Demonstration program as a benchmark. Overall, both the Monte Carlo analysis and the empirical application show feasible precision gains for the weighted cross-validation and the locally varying bandwidth approaches.

Does the Quality of Training Programs Matter? Evidence from Bidding Processes Data

Alberto Chong et al. — Tue, 15 Jan 2013 10:36:55 PST

We estimate the effect of training quality on earnings using a Peruvian program, which targets disadvantaged youths. The identification of causal effects is possible because of two attractive features in the data. First, selection of training courses is based on public bidding processes that assign standardized scores to multiple proxies for quality. Second, the evaluation framework allows for the identification and comparison of individuals in treatment and comparison groups six, 12, and 18 months after the program. Using difference-indifferences kernel matching methods, we find that individuals attending high-quality training courses have higher average and marginal treatment impacts. External validity was assessed by using five different calls over a nine-year period.

Household Wealth and Heterogeneous Impacts of a Market-Based Training Program: The Case of Projoven in Peru

Jose Galdo et al. — Tue, 15 Jan 2013 10:36:54 PST

This paper analyzes the relationship between households' wealth and heterogeneous treatment impacts for a market-based training program that has benefited more than 40,000 disadvantaged individuals in Peru since 1996. We proxy long-run wealth by a linear index based on 21 household assets, and three main findings emerge. First, we find that voluntary choices among eligibles, rather than administrative choices, play a bigger role in explaining demographic disparities in program participation. Second, quantile treatment effects on the treated suggest important differences in program impacts at different quantiles of earnings, and strong differences in distributional impacts for men and women. Third, both parametric-based and semiparametric regression-matching estimates reveal that the poorest among the poor benefit the same from the program. It is the type of institution that provides the training services that largely accounts for the heterogeneity of the impacts.

The Long-Run Labor-Market Consequences of Civil War: Evidence from the Shining Path in Peru

Jose Galdo — Tue, 15 Jan 2013 10:36:53 PST

This study exploits district-level variation in the timing and intensity of civil war violence to investigate whether early-life exposure to civil wars affects labor-market outcomes later in life. In particular, we examine the impacts of armed conflict in Peru, a country that experienced the actions of a tenacious, brutally effective war machine, the Shining Path, between 1980 and 1995. This study finds that the most sensitive period to early-life exposure to civil war violence is the first 36 months of life. A one standard deviation increase in civil war exposure leads to a four percent fall in adult monthly earnings. Neither fetal, nor pre-school, periods significantly affect long-run earnings. Substantial heterogeneity in the earnings impacts emerge when considering variation in the types of civil war violence. Sexual violations disproportionally affected the wages of women, while torture and forced disappearances disproportionally affected the wages of men. Evidence on intervening pathways suggests that health rather than schooling is the most important channel in connecting early-life exposure to civil war and adult earnings.

Informality and Productivity in the Labor Market: Peru 1986 - 2001

Alberto Chong et al. — Tue, 15 Jan 2013 10:36:52 PST

Peru has one of the highest informality rates in Latin America, with almost 60 percent of the urban labor force working at the margins of labor market legislation or in microenterprises that lack basic labor market standards (Marcouiller, Ruiz de Castilla, and Woodruff, 1997). This paper identifies two factors that can explain the variation in informality rates in the 1990s. First, Peru experienced a steady increase in employment allocation in traditionally “informal” sectors—in particular, retail trade and transport. Second, there was a sharp increase in nonwage labor costs, despite a reduction in the average productivity of the economy. In addition, the paper illustrates the negative correlation between productivity and informality by evaluating the impacts of the PROJOVEN youth training program.

Public Long-Term Care Insurance and the Housing and Living Arrangements of the Elderly: Evidence from Medicare Home Health Benefits

Gary V. Engelhardt et al. — Tue, 15 Jan 2013 10:36:50 PST

We provide empirical evidence on the extent to which long-term care insurance affects the housing and living arrangements of the elderly by examining plausibly exogenous changes in the supply of long-term care insurance through the Medicare program that occurred in the late 1990s. Prior to 1997, Medicare reimbursed home health care agencies on a retrospective-cost basis. Then, starting in October, 1997, as a result of the Balanced Budget Act of 1997 (BBA97), Medicare switched to a system of prospective payments for home health care, which induced state-by-calendar-year variation in the supply of this type of public long-term care insurance. We exploit this variation to econometrically identify the impact on the housing and living arrangements of the elderly, using CPS data from 1995-2000 (before and after the law change). Our estimates indicate that living arrangements are quite responsive to home health care benefits. The estimated elasticity of shared living to benefits is -0.7 over all elderly and -1 for widowed elderly. However, these benefits have little impact on household headship among the elderly. This suggests that the bulk of the shared-living response occurred through co-residents living in elderly households. There is some weak evidence that increases in benefits raised elderly homeownership.

New Wine in Old Bottles: A Sequential Estimation Technique for the LPM

William C. Horrace et al. — Tue, 15 Jan 2013 10:36:49 PST

The conditions under which ordinary least squares (OLS) is an unbiased and consistent estimator of the linear probability model (LPM) are unlikely to hold in many instances. Yet the LPM still may be the correct model or a good approximation to the probability generating process. A sequential least squares (SLS) estimation procedure is introduced that may outperform OLS in terms of finite sample bias and yields a consistent estimator. Monte Carlo simulations reveal that SLS outperforms OLS, probit and logit in terms of mean squared error of the predicted probabilities.

The Value of Statistical Life: Pursuing the Deadliest Catch

Kurt E. Schnier et al. — Tue, 15 Jan 2013 10:36:48 PST

Observed trade-offs between monetary returns and fatality risk identify estimates of the value of a statistical life (VSL), which inform public policy and quantify preferences for environmental quality, health and safety. To date, few investigations have estimated the VSL associated with trade-offs between returns from natural resource extraction activities and the fatality risks they involve. Furthermore researchers have been unable to determine whether or not one’s VSL is stable across multiple decision environments using revealed preference methods. Understanding these tradeoffs (and the VSL that they imply) may be used to inform resource management policy and safety regulations, as well as our general understanding of the value of life. By modeling a commercial fishing captain's choice to fish or not, conditional on the observed risk, this research investigates these topics using data from the Alaskan red king crab and snow crab fisheries. Using weather conditions and policy variables as instruments, our estimates of the mean VSL range from $4.00M to $4.76M (depending on the modeling assumption and fishery analyzed) and are robust to the incorporation of heterogeneous preferences. Furthermore, given the unique nature of the data we are able to conduct an intra-vessel comparison of the VSL and conclude that for roughly 92% of the fishermen observed in the data set their VSL estimates are stable across both fisheries.

Alternative Technical Efficiency Measures: Skew, Bias, and Scale

Qu Feng et al. — Tue, 15 Jan 2013 10:36:47 PST

In the fixed-effects stochastic frontier model an efficiency measure relative to the best firm in the sample is universally employed. This paper considers a new measure relative to the worst firm in the sample. We find that estimates of this measure have smaller bias than those of the traditional measure when the sample consists of many firms near the efficient frontier. Moreover, a two-sided measure relative to both the best and the worst firms is proposed. Simulations suggest that the new measures may be preferred depending on the skewness of the inefficiency distribution and the scale of efficiency differences.