|
on Discrete Choice Models |
| By: | van Praag, Bernard M. S. (University of Amsterdam); Hop, J. Peter (Independent Researcher); Greene, William H. (University of South Florida) |
| Abstract: | In the last few decades, the study of ordinal data in which the variable of interest is not exactly observed but only known to be in a specific ordinal category has become important. In Psychometrics such variables are analysed under the heading of item response models (IRM). In Econometrics, subjective well-being (SWB) and self-assessed health (SAH) studies, and in marketing research, Ordered Probit, Ordered Logit, and Interval Regression models are common research platforms. To emphasize that the problem is not specific to a specific discipline we will use the neutral term coarsened observation. For single-equation models estimation of the latent linear model by Maximum Likelihood (ML) is routine. But, for higher-dimensional multivariate models it is computationally cumbersome as estimation requires the evaluation of multivariate normal distribution functions on a large scale. Our proposed alternative estimation method, based on the Generalized Method of Moments (GMM), circumvents this multivariate integration problem. The method is based on the assumed zero correlations between explanatory variables and generalized residuals. This is more general than ML but coincides with ML if the error distribution is multivariate normal. It can be implemented by repeated application of standard techniques. GMM provides a simpler and faster approach than the usual ML approach. It is applicable to multiple-equation models with K-dimensional error correlation matrices and Jk response categories for the kth equation. It also yields a simple method to estimate polyserial and polychoric correlations. Comparison of our method with the outcomes of the Stata ML procedure cmp yields estimates that are not statistically different, while estimation by our method requires only a fraction of the computing time. |
| Keywords: | ordered qualitative data, item response models, multivariate ordered probit, ordinal data analysis, generalized method of moments, polychoric correlations, coarsened events |
| JEL: | C13 C15 C24 C25 C26 C33 C34 C35 |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:iza:izadps:dp17610 |
| By: | Dong, Hao; Otsu, Taisuke; Taylor, Luke |
| Abstract: | The convergence rate of an estimator can vary when applied to datasets from different populations. As the population is unknown in practice, so is the corresponding convergence rate. In this article, we introduce a method to conduct inference on estimators whose convergence rates are unknown. Specifically, we extend the subsampling approach of Bertail, Politis, and Romano (1999) to situations where the convergence rate may include logarithmic components. This extension proves to be particularly relevant in certain statistical inference problems. To illustrate the practical relevance and implementation of our results, we discuss two main examples: (i) non parametric regression with measurement error; and (ii) intercept estimation in binary choice models. In each case, our approach provides robust inference in settings where convergence rates are unknown; simulation results validate our findings. |
| Keywords: | Binary choice; convergence rate; measurement error; subsampling |
| JEL: | C14 |
| Date: | 2024–12–24 |
| URL: | https://d.repec.org/n?u=RePEc:ehl:lserod:126066 |
| By: | Manzke, Leonie (Friedrich-Alexander University Erlangen-Nuremberg); O'Sullivan, Kevin; Tiefenbeck, Verena |
| Abstract: | Food choices profoundly impact population health and the environment. Related research often relies on self-reported data, which is prone to biases, compromising the accuracy and validity of conclusions about consumer behavior. There are few systematic validations of self-reported data with behavioral data, or examinations of predictors for their accuracy. Consequently, this study compares self-reported with observed food choices, by having participants (N = 290) complete a shopping task in a simulated online grocery store, followed immediately by shopping self-reports and a survey, therefore minimizing recall-related distortions to self-reports due to time delays. Nevertheless, on average, participants had reporting errors in 3.81 out of 29 categories, with accuracy as low as a mean of 44 % for categories with no cues provided. Reporting accuracy significantly increased to 78 % with image-based memory aids for specific product categories (e.g., apples), and to 89 % with text-based memory aids for general categories (e.g., vegetables). Contrary to expectations related to social desirability bias, processed foods, often perceived as unhealthy, were overreported. Regression analysis revealed mental load during shopping, deliberation time per item, and health-related identity as significant predictors of self-report accuracy, with mental load also predicting the accuracy of participants' estimates of the proportion of organic products in their shopping basket. Our findings show that even in conditions that minimize social desirability and recall limitations, substantial self-reporting errors persist. Accounting for mental load and product-specific biases is therefore necessary to enhance the validity of self-reports in food choice research. |
| Date: | 2025–01–15 |
| URL: | https://d.repec.org/n?u=RePEc:osf:osfxxx:bn6tg |
| By: | Kosfeld, Michael (Goethe University Frankfurt); Sharafi, Zahra (Frankfurt School of Finance and Management); Sontag González, Maíra (Goethe University Frankfurt); Zou, Na (University of Bath) |
| Abstract: | Developing reliable and practicable measures of economic preferences is a crucial task for empirical economic research with high value for both theory and applications. Here, we present results from a first comprehensive "behavioral validation analysis" of the Global Preference Survey Module (GPS) and the corresponding Preference Survey Module (PSM) developed by Falk et al. (2018, 2023) that have been widely used for the measurement and analysis of economic preferences on a global scale. Our key questions are how well GPS and PSM modules explain behavior in incentivized choice experiments in other countries than in the original validation in Germany, and to what extent survey items and modules developed from behavioral experiments in different countries and cultures resemble one another. Our current results, which are based on experiments in three very diverse countries—China, Iran, and Kenya—show that many GPS and PSM survey items predict behavior in incentivized choice experiments, but coefficients vary and are not always sizable. Quantitative items, which are based on hypothetical choice experiments, are consistently selected into survey modules, whereas the best qualitative items differ between countries. At the same time, the contribution in terms of explanatory power of these latter items is comparably lower. Our analysis provides a first empirical basis for the development of survey modules that reliably predict behavior in incentivized choice experiments and real-life situations across diverse countries and contexts. Additional results, including principal component analysis and prediction of real-life behavior, highlight important gaps that warrant further investigation in future research. |
| Keywords: | economic preference, measurement, experiment, survey |
| JEL: | C83 C91 D01 |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:iza:izadps:dp17608 |
| By: | Burnitt, Christopher (University of Warwick); Gars, Jared (University of Florida and JILAEE); Stalinski, Mateusz (University of Warwick & CAGE) |
| Abstract: | Addressing rising political polarization has become a focal point for policy makers. Yet, there is little evidence of its economic impacts, especially in contexts where partisanship cannot be easily hidden. To fill this gap, we study a novel channel: the perception of out-group partisan oversight of independent civil service reduces trust in regulation, affecting key markets (e.g., food and medicine). First, we motivate it by demonstrating the salience of the association between the president and expert regulators in US media reporting. Second, in a pre-registered experiment with 5, 566 individuals, we test the channel by exploiting an alignment in the way that the EPA under Trump and Biden defended the safety of spraying citrus crops with antibiotics. This enabled us to randomize the partisanship of the administration, holding the scientific arguments constant. Despite the EPA’s independence, out-group administration reduces support for the spraying by 26%, lowers trust in the EPA’s evaluation, and increases donations to an NGO opposing the spraying by 15%. We find no overall effect on the willingness to pay for citrus products, measured in an obfuscated follow-up survey. However, we document significant differences in effects for elastic vs. inelastic consumers. Taken together, polarization has the potential to affect economic decisions. However, a reduction in trust might not translate into lower demand, especially for inelastic consumers. JEL Codes: D12 ; D83 ; P16 ; Q11 ; Q13 ; Q18 ; Z18 |
| Keywords: | political polarization ; civil service ; trust in regulation ; trust in science ; food policy ; partisan identity ; consumer demand |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:wrk:warwec:1542 |
| By: | Gonzalo, Jesús; Pitarakis, Jean-Yves |
| Abstract: | We propose a two-step procedure to detect cointegration in high-dimensional settings, focusing on sparse relationships. First, we use the adaptive LASSO to identify the small subset of integrated covariates driving the equilibrium relationship with a target series, ensuring model-selection consistency. Second, we adopt an information-theoretic model choice criterion to distinguish between stationarity and nonstationarity in the resulting residuals, avoiding dependence on asymptotic distributional assumptions. Monte Carlo experiments confirm robust finite-sample performance, even under endogeneity and serial correlation. |
| Keywords: | Cointegration; High Dimensional Data; Adaptive Lasso; Unit Roots |
| JEL: | C32 C52 |
| Date: | 2025–01–27 |
| URL: | https://d.repec.org/n?u=RePEc:cte:werepe:45708 |
| By: | Willem De Cort; Kristof De Witte |
| Abstract: | Tutoring programs are highly effective in improving students’ outcomes in higher education. However, little is known about students’ demand for tutoring or the optimal design of tutoring programs. Additionally, while privately provided tutoring can threaten social mobility, little is known about students’ consumption of private tutoring. Using a discrete choice experiment, this paper estimates the preferences, perceptions and consumption of tutoring for 1, 200 Flemish first-year higher education students. We find that students’ willingness-to-pay for tutoring is high, as they perceive it to be highly effective. However, they perceive the added value of having an experienced lecturer and smaller class sizes in these tutoring programs as low. Our simulations suggest that tutoring could significantly improve private and social welfare, but that the optimal size of tutoring groups is larger than commonly considered. Despite its potential, only 8% of students purchase privately provided tutoring. Using an information experiment, we show that biased perceptions of the returns to education cannot explain why students underinvest in tutoring. We argue that supply-side restrictions and behavioral biases such as status-quo bias and social image concerns can explain this low uptake of private tutoring. Finally, the demand for tutoring is highest among students with low socioeconomic status, but they are less able to purchase tutoring on the private market due to its high prices. These findings suggest that public investments in tutoring could increase social welfare and social mobility cost-effectively, while lowering barriers to students seeking academic support. |
| Date: | 2024 |
| URL: | https://d.repec.org/n?u=RePEc:ete:leerwp:757473 |
| By: | Rodier, Caroline |
| Abstract: | This report presents the results of the Ecosystem of Shared Mobility Services in the San Joaquin Valley (Ecosystem) pilot project. The project is part of California Climate Investments (CCI), a statewide initiative that puts billions of Cap-and-Trade dollars to work reducing greenhouse gas emissions, strengthening the economy, and improving public health and the environment — particularly in disadvantaged communities. As the grantee for this pilot project, the San Joaquin Valley Air Pollution Control District implemented the pilot program by partnering and/or subcontracting with several local entities including, but are limited to: Sigala Inc.; UC Davis, Institute of Transportation Studies; Shared-Use Mobility Center (SUMC); Self-Help Enterprises, and MOVE. Funding for the Ecosystem pilot project provided by a grant from the California Air Resources Board (CARB) through the Car Sharing and Mobility Option Pilot Project solicitation. Research for the project was also supported by funding through the University of California via the Public Transportation Account and the Road Repair and Accountability Act of 2017 (Senate Bill 1) and the National Center for Sustainable Transportation, supported by the U.S. Department of Transportation (USDOT) and the California Department of Transportation (Caltrans) through the University Transportation Centers program. View the NCST Project Webpage |
| Keywords: | Social and Behavioral Sciences, Demand responsive transportation, Mode choice, Ridesharing, Rural areas, Shared mobility, Transportation disadvantaged persons, Travel behavior, Vehicle sharing |
| Date: | 2023–03–01 |
| URL: | https://d.repec.org/n?u=RePEc:cdl:itsdav:qt6x38h5ck |
| By: | Hannah Attar |
| Abstract: | This paper investigates the determinants of home prices in Dana Point, California to analyze various factors influencing the real estate market. The results are based on a cross-sectional dataset that incorporates year and month-time dummies to account for temporal trends, as well as spatial variables that capture effects within and between clusters. To address endogeneity issues between square footage and price, parking is employed to instrument square footage and break the reverse causality link. The robustness of the instrument is confirmed through statistical tests, indicating a strong relationship with square footage. Additionally, this study employs the use of Probability models to test Tobit's robustness at estimating the dummy-transformed price variable. Spatial trends are analyzed through fixed effects, random effects, as well as Spatial Autoregressive models absorbing cluster factors, which highlight the differences in price dynamics across various clusters within Dana Point. |
| Date: | 2024–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2412.03583 |
| By: | Tinghan Zhang |
| Abstract: | Consumers are increasingly shopping online, and more and more datasets documenting consumer search are becoming available. While sequential search models provide a framework for utilizing such data, they present empirical challenges. A key difficulty arises from the inequality conditions implied by these models, which depend on multiple unobservables revealed during the search process and necessitate solving or simulating high-dimensional integrals for likelihood-based estimation methods. This paper introduces a novel representation of inequalities implied by a broad class of sequential search models, demonstrating that the empirical content of such models can be effectively captured through a specific partial ranking of available actions. This representation reduces the complexity caused by unobservables and provides a tractable expression for joint probabilities. Leveraging this insight, we propose a GHK-style simulation-based likelihood estimator that is simpler to implement than existing ones. It offers greater flexibility for handling incomplete search data, incorporating additional ranking information, and accommodating complex search processes, including those involving product discovery. We show that the estimator achieves robust performance while maintaining relatively low computational costs, making it a practical and versatile tool for researchers and practitioners. |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2501.07514 |
| By: | Klumpenhouwer, Willem; Karner, Alex |
| Abstract: | Distributive concerns in transportation equity can be evaluated either in terms of inequality (e.g., how equal are distributions?) or sufficiency (e.g., how many and what kinds of people lack access to the transportation resources they need?). Sufficiency analyses offer more actionable insights that can be used to mitigate transportation disadvantage, but related analytical methods are not well developed. To advance this area of research and practice, this paper investigates three approaches to measuring sufficiency through the lens of public transport access to jobs: (i) Fraction of total regional destinations, (ii) Competitiveness with auto access, and (iii) population-weighted percentile measures. We use a class of decomposable Foster-Greer-Thorbecke poverty measures to understand the sensitivity of overall levels of disadvantage to the choice of disadvantage lines and other parameters, in the context of seven U.S. urban regions. We find that fractional and auto competitiveness measures produce similar results and are highly sensitive to the choice of disadvantage line, that population-weighted percentile measures may allow for better comparisons across demographic groups, and that by most reasonable definitions of transport poverty the vast majority of residents (80+%) in an area might be considered to be in transport disadvantage. |
| Date: | 2025–01–06 |
| URL: | https://d.repec.org/n?u=RePEc:osf:osfxxx:95qbv |
| By: | Ziyu Jiang |
| Abstract: | Identifying structural parameters in linear simultaneous equation models is a fundamental challenge in economics and related fields. Recent work leverages higher-order distributional moments, exploiting the fact that non-Gaussian data carry more structural information than the Gaussian framework. While many of these contributions still require zero-covariance assumptions for structural errors, this paper shows that such an assumption can be dispensed with. Specifically, we demonstrate that under any diagonal higher-cumulant condition, the structural parameter matrix can be identified by solving an eigenvector problem. This yields a direct identification argument and motivates a simple sample-analogue estimator that is both consistent and asymptotically normal. Moreover, when uncorrelatedness may still be plausible -- such as in vector autoregression models -- our framework offers a transparent way to test for it, all within the same higher-order orthogonality setting employed by earlier studies. Monte Carlo simulations confirm desirable finite-sample performance, and we further illustrate the method's practical value in two empirical applications. |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2501.06777 |
| By: | Charles F. Manski |
| Abstract: | Researchers cannot definitively interpret what the framers of the United States Constitution had in mind when they wrote of the general Welfare. Nevertheless, welfare economics can contribute to policy choice in democracies. Specifying social welfare functions enables coherent analysis, by formalizing mechanisms for preference aggregation and studying the policies they yield. This paper argues that it is essential for welfare economics to adequately express the richness and variety of actual human preferences over social states. I first discuss devices that economists have used in attempts to circumvent or grossly simplify specification of social welfare functions. I next discuss the common welfare economic practice of assuming that personal preferences are homogeneous, consequentialist, and self-centered. I then call for incorporation of broader forms of personal preferences into social welfare functions. Individuals may hold heterogeneous social preferences, being concerned in various ways with the distribution of outcomes in the population. They may hold heterogeneous deontological preferences, placing value on their own actions and the actions of others. They may have preferences for the mechanism used to aggregate preferences in a social welfare function. These potential aspects of personal preference should be recognized in welfare economics. |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2501.08244 |
| By: | ALAMI CHENTOUFI, Reda |
| Abstract: | This paper introduces a two-step procedure for convex penalized estimation in dynamic location-scale models. The method uses a consistent, non-sparse first-step estimator to construct a convex Weighted Least Squares (WLS) optimization problem compatible with the Least Absolute Shrinkage and Selection Operator (LASSO), addressing challenges associated with non-convexity and enabling efficient, sparse estimation. The consistency and asymptotic distribution of the estimator are established, with finite-sample performance evaluated through Monte Carlo simulations. The method's practical utility is demonstrated through an application to electricity prices in France, Belgium, the Netherlands, and Switzerland, effectively capturing seasonal patterns and external covariates while ensuring model sparsity. |
| Keywords: | Weighted LSE; LASSO estimation; variable selection; GARCH models |
| JEL: | C01 C22 C51 C52 C58 |
| Date: | 2024–12 |
| URL: | https://d.repec.org/n?u=RePEc:pra:mprapa:123283 |
| By: | Cui, Chi; Dai, Ming; Alevy, Jonathan |
| Abstract: | This study explores the way in which social information about giving impacts the stability of distributional preferences. We designed a two-stage treatment which varied the information participants received about the maximum amounts given to recipients. Information on maximum giving can significantly increase giving share compared to the control group, especially when the relative price of giving is low. However, with a rise in the relative price, the giving decreases significantly. Applying measures of consistency with the Generalized Axiom of Revealed Preference (GARP) and non-linear Tobit estimates of preferences, we observe changes in distributional preferences indicating that more fairness and efficiency are considered in distributions when social information is provided. Type changes in distributional preferences at an individual level provide evidence that there is one substitution relationship with context to fairness-selfishness and efficiency-equality tradeoffs. People’s preferences can change due to environmental factors, which are less equalityfocused and more efficiency-oriented. It provides evidence for heterogeneity in preference stability by studying distribution stability causal effect. |
| Keywords: | Distributional Preferences; GARP; Dictator Game; Maximum Information |
| Date: | 2025–01–21 |
| URL: | https://d.repec.org/n?u=RePEc:awi:wpaper:0759 |