DSpace Repository :: Browsing by Author "Henderson, Daniel J."

Browsing by Author "Henderson, Daniel J."

Now showing 1 - 13 of 13

Applied nonparametric density and regression estimation with discrete data: plug-in bandwidth selection and non-geometric kernel functions
(University of Alabama Libraries, 2017) Chu, Chi-Yang; Henderson, Daniel J.; University of Alabama Tuscaloosa
Bandwidth selection plays an important role in kernel density estimation. Least-squares cross-validation and plug-in methods are commonly used as bandwidth selectors for the continuous data setting. The former is a data-driven approach and the latter requires a priori assumptions about the unknown distribution of the data. A benefit from the plug-in method is its relatively quick computation and hence it is often used for preliminary analysis. However, we find that much less is known about the plug-in method in the discrete data setting and this motivates us to propose a plug-in bandwidth selector. A related issue is undersmoothing in kernel density estimation. Least-squares cross-validation is a popular bandwidth selector, but in many applied situations, it tends to select a relatively small bandwidth, or undersmooths. The literature suggests several methods to solve this problem, but most of them are the modifications of extant error criterions for continuous variables. Here we discuss this problem in the discrete data setting and propose non-geometric discrete kernel functions as a possible solution. This issue also occurs in kernel regression estimation. Our proposed bandwidth selector and kernel functions perform well in simulated and real data.
The distribution of returns to education
(University of Alabama Libraries, 2019) Souto, Anne-Charlotte; Henderson, Daniel J.; University of Alabama Tuscaloosa
In this work, we revisit the traditional human capital framework and infer that risk as measured by the shape of the returns to education's distributions should be included. While education is often considered to be an investment good, human capital models often ignore the impact of risk on education investment decisions. This thesis has two aims. First, we want to find out how our different measures of risk evolved through time and between different groups. Second, we want to find out if those risks impacted education investment decisions through changes in the expected returns. That is, we investigate whether there exist risk-return trade-offs in education. In the first chapter, we overview nonparametric (spline and kernel) regression methods and illustrate how they may be used in labor economic applications. We focus our attention on issues commonly found in the labor literature such as how to account for endogeneity via instrumental variables in a nonparametric setting. We showcase these methods via data from the Current Population Survey. In the second chapter, we estimate the risk-return trade-off in the context of education. If education is treated like any other investment good, risk could play an important role in individual’s educational decisions. As portfolio theory predicts, there could be a trade-off between returns to education and risks concerning the returns: higher risks are generally associated with higher returns. We contribute to the literature by proposing various measures of risk based on the distribution of returns to education, which are in turn based on nonparametric regression results using the Current Population Survey dataset (1980-2015). We infer that risk-averse individuals prefer distributions with positive skewness and low kurtosis. Our results confirm the findings of the literature, i.e. we observe compensation for variance. We also find statistically significant compensation for the higher moments: skewness and kurtosis. Interestingly, we find that the relationship between expected returns and the higher moments skewness and kurtosis is non-linear. In the third chapter, we build on the second chapter to test two hypothesis: first whether there is heterogeneity in the risk of educational investments and if so whether there is compensation for that risk. We use our individual-level estimated rates of return to education and split them in three different ways: by occupation, by region and race, and by region and education-level. We infer that there is heterogeneity, not only in the expected returns (1st moment), but also in the risk faced by individuals (higher moments). We also add to the second chapter by testing whether risk-return trade-offs exist between occupations, whites and non-whites, and different education-level. We expect, for example, occupations that retain higher risk to be compensated by higher mean returns. Generally, we find risk-return trade-offs exist between states, occupations, whites and non-whites, and different education-level, for all three measures of risk. Surprisingly, we find that kurtosis matters more than skewness as a measure of risk. Moreover, the trade-offs between skewness, kurtosis, and expected returns are not always in the directions predicted by theory on decision making under uncertainty.
Essays on Attention Deficit Hyperactivity Disorder
(University of Alabama Libraries, 2018) Hampton, James; Chen, Susan; University of Alabama Tuscaloosa
In the first essay, we use stochastic dominance techniques to understand how the reporting of behavioral problems as well as ADHD prevalence has changed between 2000 and 2004. This time period coincides with changes in national educational policy which we hypothesize may have influenced the reporting of behavioral problems in children and a change in ADHD prevalence. We use stochastic dominance techniques and find that the distribution of behavioral problems in 2004 first-order stochastically dominates that of 2000. We then use decomposition techniques to study the primary drivers of changes in mother reported behavioral problems. We find evidence that changes in the educational policy between 2000 to 2004 led mothers of elementary school children to alter their reporting of child hyperactivity. In the second essay, we explore whether the introduction of school accountability policies can account for changes in ADHD diagnosis. We exploit differences across states and time in the introduction of school accountability laws to estimates differences in mean ADHD diagnosis. The results from our analysis suggest that one policy, state-level rewards given to high-performing schools, leads to approximates a 3 percentage point increase in the probability of an ADHD diagnosis among children. We find that the children most impacted by the policy are those whose mothers’ reported zero behavioral problems in the pre-policy period, perhaps indicating that prior to the policy these mothers did not believe that their child had behavioral problems. In the third and final essay, I study the impact of child ADHD on parental labor market and relationship dissolution outcomes. As unobserved characteristics may simultaneously impact the likelihood of having a child diagnosed with ADHD and outcomes of the parent, results using OLS estimation are likely biased. I mitigate issues of endogeneity using an instrumental variables framework where I utilize state-level educational policy as an instrument for child ADHD diagnosis. To be a valid instrument, the educational policy should be correlated with child ADHD, while exogenous to parental outcomes. While in several specifications, I find negative effects of child ADHD on parental outcomes using OLS, interestingly, IV estimates all lead to a switching of sign and are largely insignificant. Findings indicate that parental labor market and marital status outcomes are not impacted by child ADHD.
Essays on yield curve models with markov switching and macroeconomic fundamentals
(University of Alabama Libraries, 2014) Levant, Jared; Ma, Jun; University of Alabama Tuscaloosa
This dissertation explores the interaction of the term structure of interest rates and the macroeconomy for the United States and United Kingdom. In particular, using a dynamic factor yield curve model, the three essays of this dissertation investigate the macroeconomic sources of parameter instability in the US and UK term structure. First, this dissertation explores if parameter instability in the term structures is reflected in structural breaks in latent yield curve factors - level, slope, and curvature. I test for a single and for multiple structural breaks. The results indicate that parameter instability in the US term structure is adequately captured by the structural breaks in the level and slope factors. Similarly, there is evidence that structural breaks in the level and curvature factors characterize parameter instability in the UK term structure. Next, I assume the dynamics of the US term structure follow a two-state Markov process. This allows interest rate dynamics to switch between the two states as frequently as the data dictates. A switching model is proposed which gives macroeconomic insight into an asymmetric monetary policy effect during expansions and recessions. A second proposed switching model provides evidence of a great moderation in the US term structure where there is a dramatic decrease in the volatility of yields. Lastly, I investigate the interaction of the UK term structure and macroeconomy. In order to establish a definitive one-to-one correspondence between macroeconomic fundamentals and latent yield curve factors, I estimate a dynamic yield curve model augmented with macroeconomic variables. Through impulse response analysis, I find that during the inflation-targeting period for the UK, the curvature factor is directly related to real economic activity. I then use this established interaction between the term structure and macroeconomy to gain macroeconomic insight into regime changes in the UK term structure. Using Markov-switching dynamic yield curve models, I estimate the term structure and find that periods of low volatility correspond to regimes where real economic activity and monetary policy have a greater effect on the bond market. Periods of high volatility are driven by inflation expectations having a greater influence on bond pricing.
Industrial firm and household responses to energy price changes: evidence from energy subsidy reform in Iran
(University of Alabama Libraries, 2020) Hasani, Karim; Henderson, Daniel J.; University of Alabama Tuscaloosa
The rise in global average temperature due to greenhouse gas emissions is a serious international concern. Yet, one of the important barriers for clean energy transition in the world is the existence of energy subsidies. Consequently, examining how industrial and residential sectors in countries with heavily subsidized energy markets, like Iran, would respond to energy price reforms can shed light on developing more effective energy policies. In the first chapter, I discuss a broad range of related issues including the political economy behind the existence of high energy subsidies and the Targeted Subsidies Reform (TSF) conducted by Iran in 2010. Further, I review some of the most important studies on modelling energy demands. In the second chapter, I use a unique firm-level panel (2005-2011) and a translog cost system and estimate both price elasticities of demand and Morishima elasticities of substitution for Iranian manufacturing firms. The price elasticities show that labor is slightly more sensitive than capital to energy price reforms, but energy is more elastic than both. In addition, Morishima elasticities indicate the substitution of capital for energy, labor for energy, and capital for labor dominate the opposite directions. Finally, it is seen that firms exhibit heterogeneity in adjusting to energy price reforms according to labor structure and energy intensity. In the third chapter, I apply an Exact Affine Stone Index (EASI) implicit Marshallian demand system on Iranian Household Expenditure and Income Survey (HEIS) data on 10 commodity groups. I estimate the parameters of interest (elasticities, Engle curves, and welfare index) for 2008-2010 (before the subsidy reform) and 2011-2014 (after the subsidy reform). As household income level increases, the nature of energy changes from being normal to inferior. Further, the estimated nonlinear Engel curves for energy, furnishing, communication, and clothing exhibit the largest change between the two time periods. Finally, a simulated 90% increase in energy prices is estimated to be associated with an average of 4.7% rise in the cost of living before the reform and 2.6% afterwards.
Kernel-based specification testing with skewed and heavy-tailed data
(University of Alabama Libraries, 2019) Sheehan, Alice; Henderson, Daniel J.; University of Alabama Tuscaloosa
In this work, we revisit some of the most common nonparametric specification tests and assess their robustness to real world economic data. Most nonparametric econometric theory is based on a compact support assumption of the data, that is, that the data are well behaved and uniform. In economics, this assumption is regularly violated and the consequences of which are unknown. The intent of this research is twofold. First, we investigate what kind of inference practitioners are obtaining when they apply nonparametric specification tests to economic data. Then, having found that this leads to questionable inference, we modify test statistics and suggest other practices to further improve inference with such data. In the first chapter, we modify a nonparametric test for heteroskedasticity by removing the random denominator in the test statistic, an issue that can be exacerbated when applied to skewed and/or heavy tailed data. We find improvements using our modified test and suggest other methods to improve inference for practitioners applying kernel-based specification tests. In the second chapter, we propose a consistent local-linear test for variable significance that has an asymptotically standard normal distribution. Local-linear estimators are generally the dominant and preferred choice in theoretical and applied kernel regression; however, local-constant estimators are typically employed to construct test statistics. Through Monte Carlo simulations and empirical illustrations, we assess the finite sample performance of the proposed test as compared to a local-constant version. The simulations show that our local-linear test performs well, even with skewed and heavy-tailed regressors and errors, and generally outperforms the local-constant version using a wide range of data generating processes. In the third chapter, using two nonparametric kernel-based specification tests, we investigate the relative performance various auxiliary distributions used to implement the wild bootstrap in the presence of skewed and heavy-tailed regressors. Using a data driven method we identify the most appropriate auxiliary distribution for a given sample. Through Monte Carlo simulations, we show that contrary to popular practice, the Rademacher distribution provides better asymptotic refinements than that of the most commonly employed skew-corrected wild bootstrap for these tests.
Minimum Wage Changes across Provinces in China: Average Treatment Effects on Employment and Investment Decisions
(MDPI, 2021) Luo, Ji; Henderson, Daniel J.; Nankai University; University of Alabama Tuscaloosa; IZA Institute Labor Economics
We exploit data from the China Household Finance Survey to examine the impact of changes in the minimum wage on employment and investment decisions. We are able to non-parametrically identify the average treatment effect on the treated via exogenous variation in the minimum wage across provinces. We find that changes in the minimum wage had no adverse effects on employment (in terms of days worked per month or hours worked per work day) but found evidence that changes in the minimum wage impacted the percentage of families that had a bank account, a family in a rural area owned their home, and whether families (whose highest level of education was primary school) planned to purchase a home.
Nonparametric Estimation and Inference in the Presence of Sample Selection Bias in Experimental Economics Studies
(University of Alabama Libraries, 2021) Zhong, Huizhen; Henderson, Daniel J.; University of Alabama Tuscaloosa
Experimental economics studies usually involve self-selection behaviors. In this dissertation, we explore the use of nonparametric approaches to estimate the treatment effect in these studies in the presence of sample selection bias. The first chapter reviews the econometrics literature on nonparametric estimation of treatment effects under sample selection. Specifically, we focus on the Heckman (1979) two-step correction approach, its nonparametric extensions, and three bounding estimation approaches: Horowitz and Manski (2000), Lee (2009), and Behaghel et al. (2015). We also discuss the different estimands and the relative performance in these studies. The second chapter explores the treatment effect of a higher match ratio on an individual’s donation behavior based on evidence from a field experiment using multiple waves of email solicitations. Since donation decisions are observable only for email openers and opening rates differ between treatment and control groups, we apply the nonparametric bounding estimation approaches of Lee (2009) and Behaghel et al. (2015) to correct for selection bias when estimating the treatment effect. A higher match rate significantly increases an email opener’s likelihood to give and increases the donation amount for those who contributed to the fund in the past 24 months. The third chapter investigates whether randomized advertised show-up fees can be used as an exclusion restriction in the Heckman (1979) correction model to correct for bias caused by individuals’ self-selection into lab experiment studies. We control for the actual participation fee and study the impact of the advertised show-up fee on an individual’s participation decision, subject’s decision making, and the treatment effects in three well-studied lab experiment tasks. We estimate these impacts using nonparametric regressions. For the range of show-up fees in our study, we find no impact on an individual’s participation decision. Also, the advertised show-up fee does not affect the participant’s decision-making or the treatment effect in the tasks related to individuals' social preference and risk attitude. However, the advertised show-up fee impacts subjects’ strategic performance under a higher cognitive load. Therefore, caution should be made when we incorporate the randomized advertised show-up fee in the experiment design to correct for participation bias.
Semiparametric Approaches for Dimension Reduction Through Gradient Descent on Manifold
(University of Alabama Libraries, 2021) Xiao, Qing; Wang, Qin; University of Alabama Tuscaloosa
High-dimensional data arises at an unprecedented speed across various fields. Statistical models might fail on high-dimensional data due to the "curse of dimensionality". Sufficient dimension reduction (SDR) is to extract the core information through low-dimensional mapping so that efficient statistical models can be built while preserving the regression information in the high-dimensional data. We develop several SDR methods through manifold parameterization. First, we propose a SDR method, gemDR, based on local kernel regression without loss of information of the conditional mean E[Y|X]. The method, gemDR, focuses on identifying the central mean subspace (CMS). Then gemDR is extended to CS-gemDR for central subspace (CS), through the empirical cumulative distribution function. CS-OPG, a modified outer product gradient (OPG) method for CS, is developed as an initial estimator for CS-gemDR. The basis B of the CMS or CS is estimated by a gradient descent algorithm. An update scheme on a Grassmann manifold is to preserve the orthogonality constraint on the parameters. To determine the dimension of the CMS and CS, two consistent cross-validation criteria are developed. Our methods show better performance for highly correlated features. We also develop ER-OPG and ER-MAVE to identify the basis of CS on a manifold. The entire conditional distribution of a response given predictors is estimated in a heterogeneous regression setting through composite expectile regression. The computation algorithm is developed through an orthogonal updating scheme on a manifold. The proposed methods are adaptive to the structure of the random errors and do not require restrictive probabilistic assumptions as inverse methods. Our methods are first-order methods which are computationally efficient compared with second-order methods. Their efficacy is demonstrated through numerical simulation and real data applications. The kernel bandwidth and basis are estimated simultaneously. The proposed methods show better performance in estimation of the basis and its dimension.
Solving Cardinality Constrained Quadratic Optimization Problems Using the Framework of Interval Branch and Bound
(University of Alabama Libraries, 2024) Singh, Vikram; Sun, Min
Many real-world problems in engineering, statistics, social sciences, and economics can be modeled as continuous global optimization problems. Sometimes, problems also require us to choose only a handful of parameters from a large number of available parameters, resulting in a cardinality-constraint ensuring that we only select a small number of parameters in the final optimal solution. These problems can be modeled as mixed integer optimization problems by introducing some binary variables to handle the cardinality-constraint. But sometimes this re-formulation may not be equivalent to the original problem with cardinality-constraint. The following two problems can be formulated as global optimization problems subject to a cardinality-constraint.Best Subset Selection is a classical problem in statistics that requires choosing a subset from a larger number of available variables to be included in the model which minimizes the residual sum of squares in the linear regression. Sparse Portfolio Selection requires choosing optimal proportions of the assets to be included in the portfolio while restricting the number of assets included, to minimize the risk for a desired return. In this thesis, we solve the above two problems by dealing with cardinality-constraint in the original primal space itself, using the framework of interval branch and bound. Interval branch and bound is a well-known method to solve global optimization problems even when the feasible set is non-convex and the goal is to find all the minimizers of the problem. Firstly, we modified the main features of the interval branch and bound to treat the cardinality-constraint effectively. The resulting algorithm gives a procedure to enumerate all the candidate solutions and discard the sub-optimal nodes by using deletion conditions to find an optimal solution. The algorithm converges in a finite number of iterations. Secondly, we introduced a procedure to find the lower bound of a convex quadratic function for a given node by solving the linear system resulting from the first-order necessary conditions. The procedure incorporates a strategy of recycling the parent node's echelon form to find the function's lower bounds for a sequence of child nodes while maintaining the global convergence of the algorithm. We also proposed a sub-optimal algorithm to solve the Best Subset Selection problem at a low computational cost. The proposed algorithm has also been used as a feasibility sampling procedure to get good-quality upper bounds within the interval branch and bound.Lastly, we extended our framework to include special linear constraints from the Sparse Portfolio Selection problem and demonstrated the applicability of our procedure to solve it. The numerical experiments validate the effectiveness of our methods to solve these problems.
Three essays in managerial discretion
(University of Alabama Libraries, 2017) Gilstrap, Collin; Mobbs, Houston Shawn; Underwood, Shane E.; University of Alabama Tuscaloosa
This dissertation explores three areas where managerial discretion can impact firm outcomes. The in the first essay examines how CEOs with a career background in sales and marketing, “Sales CEOs”, impact operating outcomes at the firm. We find that firms with a Sales CEO control larger market shares, increase market shares when a Sales CEO is hired, and during industry sales downturns experience lower decreases in sales and market value. We also explore outside perception of Sales CEOs firms. We find that firms with Sales CEOs exhibit increases in institutional ownership when the Sales CEO is hired. We also find that Sales CEOs meet or beat analyst expectations more frequently than their non-sales and marketing peers. The second essay focuses on the market’s perception of uncertainty about a firm when Sales CEOs are present. We document larger decreases in uncertainty when Sales CEOs are hired and when Sales CEOs release earnings announcements, relative to their non-sales and marketing peers. We hypothesize and find that our set of Sales CEOs communicate more frequently with investors, and that communication tends to be more readable and positive. The third essay focuses on managerial discretion when estimating bad debt expense around hospital bond issuance. We hypothesize and find that managers under report bad debt expense prior to bond issuance. Further, we find that hospitals that significantly underreport bad debt expense experience significantly lower cost of debt when issuing bonds. Finally, we identify discretionary charity care as one channel that hospitals may misclassify bad debt expense as charity care.
Three essays on distributive politics: how legislature size and partisan politics impact the distribution of government spending
(University of Alabama Libraries, 2014) Hankins, William Bryce; Pecorino, Paul; University of Alabama Tuscaloosa
This dissertation is composed of three essays that investigate how legislature size, political alignment, and political polarization impact the distribution of government expenditures at different levels of government. The first essay focuses on political alignment and polarization while the last two essays focus on legislature size at the cross-country and US state level, respectively. In the first essay, we find evidence that during times when political polarization in the US Senate is relatively low, states with more senators in the majority receive a larger than average share of federal grant spending per capita. We also find that although states with more senators in the majority receive a larger than average share of federal grant spending per capita when both chambers of the US Congress are aligned, that this amount is smaller than what these states receive when control of Congress is divided. Lastly, we verify that states with the entire Senate delegation in the majority are driving these results. In the second essay, we find significant evidence that countries with bicameral legislatures experience larger levels of central government expenditures as a percentage of GDP when the upper chamber is larger than average. Conversely, we were not able to show any consistent relationship between unicameral legislature size and central government expenditures as a percentage of GDP or between lower chamber size and central government expenditures as a percentage of GDP. In the final essay, we examine the role legislature size has in determining the growth in state-level per capita spending. Overall, we were unable to verify a relationship between lower chamber size, upper chamber size, or the ratio of the lower-to-upper chamber size and the change in total spending per capita. We do find a positive relationship between lower chamber size and the change in per capita welfare spending.
Three essays on the impact of demographic and environmental changes on home sales
(University of Alabama Libraries, 2020) Goodnature, Mia; Ross, Amanda; University of Alabama Tuscaloosa
Gentrification occurs when low-income areas transition into higher-income neighborhoods. Chapter 1 examines one possible driver of gentrification: the influx of same-sex couples into a community. Anecdotal evidence suggests that there is a relationship between same-sex couples and gentrification, but this could be because these couples sort into neighborhoods that are more likely to gentrify. To address the endogeneity problem, we employ an instrumental variables strategy using voting results for the state-level equivalent of the Defense of Marriage Act in Ohio as an instrument for the change in the number of same-sex couples. We find that areas with a higher change in the number of same-sex couples are more likely to experience gentrification. In addition, using semi-parametric techniques, we find there is a tipping point after which gentrification is more likely to occur. Overall, our results suggest that same-sex couples can initiate gentrification, but there is a threshold that has to be met for neighborhood change to be more likely to occur. These findings are important for policy makers because understanding the drivers of gentrification is crucial to designing effective policy to revitalize urban neighborhoods and address any problems attributed to gentrification. Chapter 2 identifies same-sex couple households who purchase homes together and evaluates the concentration of their residential location. We draw upon a novel data set of real estate transactions from Miami-Dade County, Florida; Franklin County, Ohio; and King County, Washington. We are able to separately identify male same-sex couple homebuyers and female same-sex couple homebuyers at the property level by predicting the homebuyers’ sex based on homebuyers’ full names. To show that the method suggested in this paper to identify members of the LGBTQ+ community is identifying same-sex couple homebuyers, we compare distributions from the Decennial Census and look at summary statistics of houses purchased by same-sex couples. As hurricane destruction has become more frequent and more dramatic, it is important to understand how communities respond to this damage. Chapter 3 explores how the selling price of houses responds to spillover effects of living near houses with hurricane-induced damages and repairs. These spillover effects are investigated in Punta Gorda, Florida, which was hit by Hurricane Charley, a Category 4 hurricane, in August 2004. Results indicate that house prices temporarily increase after the hurricane. Nearby damaged houses have no statistically significant effect. Nearby houses that were repaired to a larger square footage have a positive spillover effect while all other repaired houses, like those that do not increase their square footage, have a negative spillover effect on housing prices.

Browsing by Author "Henderson, Daniel J."

Results Per Page

Sort Options