Sample Selection Model Research Articles

Some identification results in a correlated random coefficients sample selection model

This paper studies the linear correlated random coefficients model subject to sample selection. The random coefficients can be correlated with one or more covariate components. To address the endogeneity of the covariates, we employ the control function approach, while correcting for sample selection by applying the method of local instrumental variables. We establish that, under particular regularity conditions, the average partial effects and several counterfactual parameters can be identified.

Consumer knowledge and intention-behavior consistency

PurposeIn marketing, most research on intention–behavior consistency (IBC) is dedicated to improving the predictive ability of stated intentions for future behaviors, with relatively less exploration into the precursors of IBC, especially those linked to regular durable goods void of ethical consumption characteristics. This study aims to focus on the antecedents of IBC for such products, specifically examining category-level and product-level IBC in light of consumer knowledge.Design/methodology/approachThe authors conducted a two-round survey to collect 3,560 Chinese consumers' vehicle purchase intentions and behaviors. The authors have also leveraged a large vehicle database (containing detailed vehicle attribute information) to measure consumer product knowledge (i.e. product judgment accuracy). A trivariate probit model was proposed to account for the potential selection bias arising from sample attrition while examining the effects of consumer knowledge on category- and product-level intention-behavior consistency.FindingsFindings reveal that 47% of the participants displayed category-level IBC, and within this group, a further 39% exhibited product-level IBC. Notably, product knowledge, manifested as accurate product judgment, correlates negatively with category-level IBC but positively with product-level IBC. Intriguingly, the negative association between inaccurate judgment and product-level IBC is less pronounced for consumers overestimating the target product than for those underestimating it. Furthermore, consumers with direct experience are less prone to show category-level IBC, but are more inclined to display product-level IBC.Practical implicationsVehicle marketers should prioritize consumers who show interest in their products but possess inaccurate knowledge, to retain whom companies can nurture their product knowledge. As for consumers with accurate knowledge, companies should try to expedite their purchase. Vehicle marketers also need to devise suitable advertising strategies to prevent consumers from undervaluing their products. For those overestimating competitors' products, companies can provide information to correct their overestimation and draw attention to possible confirmation biases. Vehicle marketers should encourage potential buyers who have shown interest in their product to participate in test-drive events, exhibitions, and other direct experience opportunities. Yet, for consumers still in the “whether-to-buy” decision-making phase, companies should not rush them into a test drive.Social implicationsIn the policy-making realm, governmental administrators can implement extensive consumer education programs, with a focus on the importance of product knowledge. This may involve providing consumers with accurate information and buying guides through various channels, which can help consumers make informed purchase decisions. Moreover, to foster healthy competition among vehicle companies, governmental administrators can establish regulations that require vehicle companies and other relevant industries to provide accurate and transparent product information, including performance, safety, and environmental aspects. Finally, in order to protect consumer rights, governmental administrators can also strengthen regulations to ensure fair treatment and safeguards for consumers throughout the purchasing process. This includes cracking down on false advertising and fraudulent practices, maintaining market order, and enhancing consumer confidence and purchase consistency.Originality/valueThis study is among the first attempts to examine the relationship between consumer knowledge and intention-behavior consistency, especially for regular durable products void of ethical consumption characteristics. Responding to the call of previous literature (e.g. Morwitz, 1997), the authors distinguish between and examine two forms of intention-behavior consistency simultaneously (using a sample selection model) and obtain more reliable conclusions. Moreover, the study's large-scale two-round survey had obtained individual-level purchase behavioral outcomes, which allowed the authors to measure each consumer's IBC at both category and product levels. More importantly, the authors show the opposite effects of consumer knowledge on the two forms of intention-behavior consistency.

Analysis of Determining Factors of Agricultural Income among Farm Households Participating in Farming Cooperative Using a Sample Selection Model

Employing the Heckman model and drawing on the national database, this study scrutinized the determinants influencing the selection of participation in farming cooperatives, as well as the impact of such participation on the agricultural income of farm households. Participation probabilities were observed to rise notably among males in their 50s and 60s with larger business scales, while protracted periods of farming experience and education correlated positively with the likelihood of engagement in farming cooperatives. Farm households actively participating in farming cooperatives demonstrated a notable elevation in their income levels compared to those who abstained from participation. Among the subset of farm households engaged in agricultural cooperatives, the utilization of ICT was linked to a more agricultural income. Given the discernible positive impact of farming cooperatives on the amplification of agricultural income, there arises a compelling need for the government to fortify policies aimed at nurturing the growth and development of these cooperatives.

Total Bias in Income Surveys when Nonresponse and Measurement Errors are Correlated

Abstract Household surveys on income might suffer from quality limitations mainly due to the difficulty of enrolling households (unit nonresponse) and retrieving correct information during the interview (measurement error [ME]). These errors are likely to be correlated because of latent factors, such as the threat of disclosing personal information, the perceived sensitivity of the topic, or social desirability. For survey organizations, assessing the interplay of these errors and their impact on the accuracy and precision of inferences derived from their data is crucial. In this article, we propose to use a standard sample selection model within a total survey error framework to deal with the case of correlated nonresponse error (NR) and ME in estimating average household income. We use it to study the correlation between the two errors, quantify the ME component due to this correlation, and evaluate ME among nonrespondents. Using the Italian Survey on Income and Wealth linked with administrative income data from tax returns, we find a positive correlation between the two errors and that households at the extremes of the income distribution mainly cause this association. Our results show that ME contributes more to the total error than unit nonresponse and that it would be larger in absence of the correlation between the two errors. Finally, efforts to reduce nonresponse rates are worthwhile only for nonrespondents in the lowest estimated response propensity group. If these households participate, the bias decreases because of the reduction in NR that offsets the increase in ME.

Public Agricultural Extension, Pest and Disease Experience, and Adoption of Improved Wheat Varieties

Improved varieties are considered critical for increasing crop yields worldwide. This study explored the effects of public agricultural extension and pest and disease experience on adoption of improved varieties using survey data on 525 wheat farmers in Anhui Province, China, to which the Heckman sample selection model was applied. The results showed that public agricultural extension had a significant positive relationship with adoption of improved varieties. Demonstration and training, as different forms of agricultural extension, both increased the probability of farmers adopting improved varieties, but only demonstration had a marked effect on degree of adoption. Pest and disease experience increased the probability of adoption of improved varieties by farmers and significantly enhanced the effect of public agricultural extension, but did not affect degree of adoption. Further analysis using an endogenous switching regression model revealed that adoption of improved varieties raised wheat yields by around 337.83 kg/ha. Public agricultural extension should thus be strengthened, especially for farmers with pest and disease experience, and a diversified, well-functioning agricultural extension system should be provided.

표본 선택 모형을 통한 밀레니얼 세대의 인턴십 경험이 취업 및 임금에 미치는 영향 분석

The aim of this study is to analyze the influence of internship experiences on the employment and wages of millennials during their college years to understand the actual labor market outcomes of internships. For this purpose, we utilized data from the 2020 Graduate Occupational Mobility Survey (GOMS 2019) to estimate a probit model and a wage equation using Heckman two-stage model. First, the model validity analysis confirms that the Heckman's sample selection model is statistically significant for this study, exhibiting a negative correlation, with smaller wages having a higher probability of being selected into the sample. Second, the first-stage of Heckman’s probit model analyzed the effects of internship experiences on employment, revealing that internships positively impact employment prospects. This impact was statistically significant with regard to gender, academic grade, metropolitan region, and the father's level of education. Third, in Heckman's two-stage wage equation estimation, it was found that internship experience did not influence wages, wherein the study showed that regular employees earned approximately 22% more than non-regular employees.

The determinants of household electricity demand in Turkey: An implementation of the Heckman Sample Selection model

Residential areas have a significant share in electrical energy consumption in Turkey. At this point, determining electricity consumption demand in residential areas will significantly contribute to ensuring energy efficiency throughout the country. Accurate econometric modelling of the relationships between the various socio-economic characteristics of households and their energy consumption/expenditure can serve as a basis for policymakers when planning investments aimed at efficiency on primary energy consumption in residences. Considering these issues, this study aims to examine the electricity consumption demand in residential areas in Turkey in terms of various socio-economic and building and property characteristics variables. At this point, the fact that some households have zero electricity consumption has led to the need to use censored regression models. The Heckman Sample Selection model was used, which makes it possible to estimate the consumption decision of individuals in two stages and eliminates the parameter bias problem caused by linear regression analysis. Heckman Sample Selection modelling was carried out in the study using Turkish Statistical Institute (TURKSTAT) 2019 Household Budget Survey data. According to the results, it has been determined that annual disposable income, various demographic variables (age, education level, marital status), characteristics of the house (size) and the presence of various electrical household appliances (air conditioner, tumble dryer, deep freezer) have significant effects on electricity consumption. The systematic preparation of informative brochures on the effective use of household electrical appliances can significantly contribute to efficiency. In addition, in countries like Turkey that depend on foreign energy, providing incentives towards renewable energy sources is considered necessary to contribute to the country's economy positively.

Spatial Heterogeneity in Hedonic Price Effects for Lake Water Quality.

This study uses Zillow's ZTRAX property transaction database to investigate variation in hedonic price effects of water clarity on single-family houses throughout the United States. We consider five spatial scales and estimate models using different sample selection criteria and model specifications. Our results indicate considerable spatial heterogeneity both within and across the four U.S. Census regions. However, we also find heterogeneity resulting from different types of investigator decisions, including sample selection and modelling choices. Thus, it is necessary to use practical knowledge to consider the limits of market areas and to investigate the robustness of estimation results to investigator choices. (JEL Q51).

공동주택 관리방식을 고려한 관리인원 결정요인 연구

In this study, the determinants of management personnel considering selection bias according to the management method were empirically analyzed for management personnel, which have been reported as a factor of increase or decrease in common area management fee in previous studies related to apartment management. This study empirically analyzed a total of 14,020 apartment complexes in Korea using a Sample Selection Model based on K-apt's management information as of 2019. The main empirical results are as follows. First, this study shows that the overall complex characteristics, which are the management targets of apartment complexes, have a significant effect on the decision of the management personnel. In the case of multi-use apartment complexes, it was analyzed that the number of management personnel increased due to the complexity and diversification of management targets, and the size of apartment complexes was found to be economical of scale. In addition, the proposition of exclusive area to total complex area, the number of parking spaces per dwelling unit, the exclusive area per dwelling unit, and major brand apartment complexes were also significant influencing factors. Second, this study confirmed that the characteristics of facilities and equipment such as CCTVs per dwelling unit and community facilities related to the infrastructure of apartment complexes are factors that increase the number of management personnel for professional maintenance and operation. Third, this study found that the regional characteristics set as control variables had a significant effect on the decision of the management personnel in the order of Seoul, Gyeonggi-do, and Incheon.

Does directors' and officers' liability insurance improve corporate ESG performance? Evidence from China

AbstractWe examine the causal effect of directors' and officers' liability insurance (D&O insurance) on environmental, social and governance (ESG) performance in China's listed firms. Using a unique data set on D&O insurance purchases in China, we document a positive relationship between D&O insurance and ESG performance, which is robust to a comprehensive battery of robustness checks including Heckman two‐step sample selection model, propensity score matching (PSM) model, instrumental variable approach, and fixed effects model. Further analyses suggest that the positive association is more salient for state‐owned enterprises (SOEs). Moreover, the path analysis provides evidence for the mediating effect of non‐financial information disclosure and corporate governance quality on the relationship between D&O insurance and ESG performance. Collectively, this study contributes to the literature on the economic benefit of D&O insurance by providing initial evidence that D&O insurance appears to enhance corporate governance and improve corporate ESG performance.

Corporate Social Responsibility and the Willingness to Eco-Innovate among Chilean Firms

Adopting social responsibility can be a key strategy for firms to mitigate the impact of production on the environment, contributing to a more sustainable business model. Based on the triple bottom line perspective, we analyze the effect of Corporate Social Responsibility (CSR) on the willingness to allocate resources to eco-innovation among companies in a developing country. Firm-level data from the Fifth Longitudinal Survey of Chilean Firms are used to estimate a binary probit model for the willingness to eco-innovate and a Heckman sample-selection model for total expenditures and investment in eco-innovation. Results confirm that legal regulations and R&D efforts are drivers of eco-innovation among Chilean firms. Larger-sized and export-oriented firms also exhibit a higher willingness to eco-innovate. The main findings show a positive influence of CSR policy on the willingness to eco-innovate and on the resulting resource allocation decision. Interestingly, the evidence reveals that while financial and environmental CSR dimensions only affect the probability of adopting eco-innovation, the social CSR dimension also increases the amount firms spend on eco-innovation. This result suggests that social CSR may act as a complement to other CSR dimensions. These results are robust when controlling for firm-level specific effects under sample-selection models.

Prepayment and credit utilization in peer-to-peer lending

PurposeThis research analyzes borrowers' credit utilization through prepayment behavior in peer-to-peer (P2P) lending. The authors investigate factors influencing the decision to prepay and assess the role of P2P lending as an alternative source of consumer credits.Design/methodology/approachThe authors use individual loan-level data from the LendingClub, one of the largest P2P platforms in the USA. The authors use a Logit model and a sample selection model estimated by the two-stage Heckman method. The empirical analysis considers borrower-specific and loan-specific characteristics as well as macroeconomic factors.FindingsThe authors present a number of significant findings that can enhance understanding consumers' financing decisions. The authors offer evidence that borrowers are able to take advantage of cheaper loans offered by P2P lending to better manage credit card balance and consolidate debt. The authors find that borrowers tend to prepay P2P loans quickly when the aggregate cost of borrowing is low, suggesting that P2P lending offers an efficient alternative to obtain credit. This is particularly true for creditworthy borrowers that are able to take advantage of competing sources of finance. The authors' results provide evidence that P2P lending can improve consumers' optimal credit utilization.Originality/valueP2P lending has grown exponentially and has become a significant credit supplier to consumers and small businesses. While the existing literature mostly focuses on default risks, prepayment has received much less attention. This research fills in the gap and investigates borrowers' prepayment behavior in P2P loans and the role of P2P lending as an alternative source of consumer credits.

A censored regression analysis of household-level demand for pastured eggs in the United States

Purpose This paper empirically investigates the demand for pastured eggs in the United States and evaluates the welfare consequences of Japan's egg import tariff reductions for the US consumers. Design/methodology/approach Using household-level Nielsen Homescan panel data, a fixed-effects Heckman two-stage sample selection model is estimated. Findings The estimation results ascertain the importance of a set of household socioeconomic characteristics, which are found to influence both the purchase probabilities and the consumption amounts associated with pastured eggs. In addition, demand for pastured eggs is estimated to be inelastic, and pastured eggs are found to be a normal good, more specifically a luxury. Research limitations/implications The dataset used in this study reflect purchases only for at-home consumption, lacking information on away-from-home purchases. Originality/value Building upon previous research, this study makes the following distinct contributions to the current literature. To the best of our knowledge, it constitutes the first study to empirically examine the demand for pastured eggs, using household-level panel data and an estimation model that not only allows for left-censoring but also controls for regional and time fixed effects. Second, the present study reflects a unique effort in analyzing the adverse welfare consequences of the increased egg prices in the United States brought by a reduction of Japanese import tariffs on US-supplied eggs, focusing specifically on pastured eggs.

Modelling the maize marketed surplus behaviour under risk and time preference conditions: The case of Zvimba and Mokonde districts of Zimbabwe

The Heckman sample selection technique was applied in modelling the households’ maize marketed surplus decision under risk and time preference assumption. We assumed that maize returns depend on farmers’ market selection choices. We then applied the sample selection model to explain why maize surplus households shunned the uncertain, high-return state-owned Grain Marketing Board (GMB), with delayed payment system in favour of the more certain, low-return private buyer market, with immediate payment system. In the process, we tested whether the marketed surplus decisions followed a sequential or simultaneous decision process. Using survey data collected from 433 households from Zvimba and Makonde districts, the results confirmed that households followed a sequential decision process when selling through the private buyer market and a simultaneous decision process when selling through the GMB market. These findings implied two things; firstly maize surplus households could exercise some bargaining power in the private buyer market even if they got less. Secondly, in the presence of uncertainty and payment delays, high-return incentive offered through government parastatals may not be sufficient to attract maize surplus, as households may not sufficiently respond, due to risk aversion attitude and time impatient behaviour.

Universal robust regression via maximum mean discrepancy

Abstract Many modern datasets are collected automatically and are thus easily contaminated by outliers. This has led to a renewed interest in robust estimation, including new notions of robustness such as robustness to adversarial contamination of the data. However, most robust estimation methods are designed for a specific model. Notably, many methods were proposed recently to obtain robust estimators in linear models, or generalized linear models, and a few were developed for very specific settings, for example beta regression or sample selection models. In this paper we develop a new approach for robust estimation in arbitrary regression models, based on maximum mean discrepancy minimization. We build two estimators that are both proven to be robust to Huber-type contamination. For one of them, we obtain a non-asymptotic error bound and show that it is also robust to adversarial contamination, but this estimator is computationally more expensive to use in practice than the other one. As a by-product of our theoretical analysis of the proposed estimators, we derive new results on kernel conditional mean embedding of distributions that are of independent interest.

Can financial literacy be a substitute for financial advisers? Evidence from China

This study investigates the influence of financial literacy on the utilization of financial advisers in China, using 2019 data from the China Household Finance Survey and a probit with sample selection model. The findings indicate that elevated levels of financial literacy have a beneficial impact on both the demand for and employment of financial advisers. The impact is more significant among younger or middle-aged individuals with debt and awareness of financial threshold changes. The research also reveals that financial literacy improves information attention and comprehension of risky financial products, resulting in greater demand for and employment of financial advisers.

Sustainable Use of Soil and Water Conservation Technologies and Its Determinants: The Case of the Handosha Watershed, Omo-Gibe River Basin, Ethiopia

For the past forty years, Ethiopia has been promoting sustainable land management activities to enhance agricultural productivity. This study was intended to identify the factors determining farmers’ adoption and continued use of soil bund measures in the Handosha watershed, Omo-Gibe river basin. A multistage sampling technique was used to select 340 households using the Heckman sample selection model. A total of 235 (69.12%) households adopted soil bunds, but only 89 (37.87%) of them were sustainably practicing soil bunds on their farm plots. Most adopters widely practiced soil bunds (49.42%), followed by stone bund (15.9%), and Fanyajuu (10%). The empirical results of the Heckman sample selection model showed that the farming experience, land tenure security, and perception of profitability of conservation measures were significantly positively affected the adoption of soil bund. Whereas, farm plot size and participation in off farm activities significantly negatively influenced the adoption of soil bund. Sustainable use of soil bund measures were significantly positively influenced by land tenure security, family size, and frequency of extension contact, whereas the distance between farm plots and home, and farm plot size were negatively affected. As a result, a design of agro-ecological-based soil and water conservation (SWC) measures was essential in reducing farmland vulnerability to soil erosion and food insecurity. It has been concluded that conservation practices should not only focus on the implementation and biophysical factors but also consider the socioeconomic interests of the farmers to improve the sustainable use of conservation technologies.

The effects of environmental information disclosure on stock price synchronicity in China

Environmental information disclosure has attracted the attention of the capital market because it can convey corporate characteristic information. But there needs to be direct evidence that environmental information disclosure can improve the market's overall efficiency. This study examines whether corporate environmental information disclosure can improve the information efficiency of the capital market. This study takes Chinese listed companies from 2008 to 2021 as samples, constructs a panel fixed effect model, and adopts multiple linear, instrumental variable method and Heckman sample selection model. We find that the disclosure of environmental information in the Chinese market reduces the information efficiency of the capital market represented by the synchronicity of stock prices. The reason is that the information after greenwashing by enterprises needs to be of better quality and more obscure, which disrupts market information. Environmental information disclosure by enterprises prone to greenwashing significantly affects stock price synchronicity, such as low institutional ownership, non-state-owned, growth, and manufacturing firms. Finally, this paper discusses the impact mechanism and confirms that stock liquidity and analyst coverage are the two paths through which environmental information disclosure impacts stock price synchronicity. This study is significant in encouraging the government to strengthen market supervision, promoting enterprises to disclose high-quality environmental information, and improving the pricing efficiency of the capital market.

Improving ICESat‐2‐based boreal forest height estimation by a multivariate sample quality control approach

Abstract Boreal forest heights are associated with global carbon stocks and energy budgets. The launch of the Advanced Topographic Laser Altimeter System (ATLAS) onboard the NASA's Ice, Cloud and Land Elevation Satellite (ICESat‐2) enables canopy vertical structure measurement at a global scale. However, with a photon‐counting laser system, ICESat‐2 contains high uncertainties in the estimated canopy heights, requiring appropriate quality control before being applied to canopy height modelling. We adopted a multivariate quality control approach (i.e. the Cook's distance) to improve the quality of ICESat‐2 samples. The controlled ICESat‐2 data were then input as the response variable for predicting boreal forest heights based on spatially continuous satellite data and machine learning (ML) regression models. The examined ML regressors include artificial neural networks (ANN), gradient boosting machine (GBM), random forest (RF) and support vector regression (SVR). The proposed quality control effectively removes low‐quality ICESat‐2 samples and enhances the correlations between ICESat‐2 and airborne laser scanning (ALS) observations. Moreover, the controlled ICESat‐2 samples help achieve a trade‐off between sample quality and quantity for all ML regressors, generating close canopy heights to ALS‐derived counterparts. Overall, RF and GBM make better canopy height predictions than ANN and SVR. Of all explanatory variables, the normalized difference vegetation index calculated based on the first red edge band of Sentinel‐2 (NDVIredEdge1) is considered the most important. The proposed quality control on ICESat‐2 sample selection and canopy height model (CHM) development workflow will greatly benefit forest structure investigations in the Arctic community.

The formation history of our Galaxy’s nuclear stellar disc constrained from HST observations of the Quintuplet field

Context. Until recently it was thought that the nuclear stellar disc at the centre of our Galaxy was formed via quasi-continuous star formation over billions of years. However, an analysis of GALACTICNUCLEUS survey data indicates that > 80% of the mass of the stellar disc formed at least 8 Gyr ago and about 5% roughly 1 Gyr ago. Aims. Our aim is to derive new constraints on the formation history of the nuclear stellar disc. Methods. We analysed a catalogue of HST/WFC3-IR observations of the Quintuplet cluster field. From this catalogue, we selected about 24 000 field stars that probably belong to the nuclear stellar disc. We used red clump giants to deredden the sample and fit the resulting F153M luminosity function with a linear combination of theoretical luminosity functions created from different stellar evolutionary models. Results. We find that ≳70% of the stellar population in the nuclear disc probably formed more than 10 Gyr ago, while ∼15% formed in an event (or series of events) ∼1 Gyr ago. Up to 10% of the stars appear to have formed in the past tens to hundreds of Myr. These results do not change significantly for reasonable variations in the assumed mean metallicity, sample selection, reddening correction, or stellar evolutionary models. Conclusions. We confirm previous work that changed the formation paradigm for stars in the Galactic Centre. The nuclear stellar disc is indeed a very old structure. There seems to have been little star formation activity between its formation and about 1 Gyr ago.