Sample Selection Bias Research Articles

Outbound tourism and outward foreign direct investment

Malik and Latif (2021, 2023) have conducted empirical research that demonstrates a positive relationship between Chinese outbound tourism and China's outward foreign direct investment. However, their studies possess significant flaws. Firstly, they fail to account for the heteroscedasticity of random errors, yielding flawed statistical inferences. Secondly, the control variables lack bilateral attributes between countries and overlook time-fixed effects, resulting in omitted endogeneity. Thirdly, their regressions inadvertently exclude a substantial number of observations due to missing values of the independent variable, introducing sample selection biases. This article aims to address these concerns by reanalyzing the empirical data and using a series of econometric methods to handle endogeneity issues. Our findings indicate that Chinese outbound tourism does not significantly influence China's outward foreign direct investment.

Learning With Imbalanced Noisy Data by Preventing Bias in Sample Selection

Learning With Imbalanced Noisy Data by Preventing Bias in Sample Selection

An Assessment of the Relevance of the Mincerian Equation: The Case of Mauritius

This paper assesses the relevance of the Mincerian Equation in Mauritius. Using CMPHS data, the basic equation is augmented to cater for wage responsiveness to changes across different schooling levels. However, the use of the standard linear regression method does not take into consideration the endogeneity of the variable years of schooling and sample selection bias. The IV2SLS Method is adopted, using region as an instrument. The Heckit model is further adopted to address the problem of sample selection bias. This study also employs pseudo panel models as an alternative to IV2SLS as a weak instrument may significantly impact results.

Sample Selection Bias in Non-Traditional Lending: A Copula-Based Approach for Imbalanced Data

Sample Selection Bias in Non-Traditional Lending: A Copula-Based Approach for Imbalanced Data

How does the setup of sample collection influence survey results - an example of new mobility services

Conducting surveys in transportation research is becoming more complex. Depending on the survey subject, the survey format and the circumstances of the sample collection the motivation of respondents to participate and consequently the results can vary substantially. Skewness of samples and sample selection bias occur to different degrees. This study compares different surveys which were created to capture the acceptance regarding shared autonomous minibuses in Germany. By analyzing distributions of behavior, perception and intention to use the services, biases in the datasets are worked out. The results show that voluntary on-site surveys lead to more positive perception of minibuses.

Post-pandemic learning trajectories in oral reading proficiency: Results from twelve Chilean schools

By drawing data from first-to sixth-grade students of 12 Chilean schools, we examined the evolution of COVID-19 pandemic-driven losses in two components of oral reading proficiency: reading decoding skills and reading speed. Our dataset spans from 2018 to 2022, enabling us to compare students’ results after the fully remote 2020, the incomplete resumption of in-person learning in 2021 and the massive return in 2022. We used multivariate regression models corrected for sample selection bias. Reading proficiency losses were more significant and persistent in reading speed, and they were higher among socially disadvantaged students entering first grade in 2020–2022. These findings uphold that school disruption affected the reading proficiency of cohorts beginning school during the pandemic more significantly and long-lastingly.

Does food safety perception matter to farmers’ happiness？ evidence from China

Although the determinants of individuals' happiness have been widely examined in the literature, little is known regarding whether and how food safety perception affects farmers' happiness. To fill this research gap, this paper examines the impact of food safety perception on happiness among Chinese farmers, utilizing open-access data collected through the Chinese Social Survey project in 2013, 2017 and 2021. This study focuses on Chinese farmers as the research subject, attempting to analyze the “happiness code” from the perspective of food safety, which supplements the literature on happiness and provides reference for protecting the rights of low-income groups and promoting food safety strategies in developing countries. To address sample selection bias, this paper employs the recursive bivariate ordered probit (RBOP) model and conditional mixed process (CMP) method. The results reveal that the perception of food safety exerts a positive and statistically significant impact on farmers' happiness in China. In addition, food safety perception is more important among middle-aged and elderly farmers and among those with higher education. Thus, the policy makers should continue to make up for the shortcomings of rural food safety work and extend regulatory measures to rural areas. They also need to take efforts to strengthen food safety promotion, enhance farmers' safety awareness, and safeguard farmers’ “safety on the tongue”.

PhD holders propensity to work in research-intensive sectors: evidence from Italy

PhD holders play an important role in exploring new research areas relevant to the development of knowledge-based economies, and the investment in PhD education is seen as part of a strategy for tomorrows’ society, with doctorates historically being absorbed by an academic career dedicated to teaching and research. However, these days are gone, and the number of PhD holders is far beyond the academic vacancies available. The present study aims at estimating the probability of PhD holders to work in research intensive sectors in Italy. We use data available from the National Institute of Statistics to estimate a probit model with the Heckman correction for the sample selection bias. The study has political consequences related to PhD holders career orientation.

Analysis of gender wage gap and the Nigerian labour market: a new empirical evidence

PurposeThis study investigated the gender wage gap in Nigeria by analysing two waves of household surveys (in 2003–2004 and 2018–2019) in order to understand the dynamics or polarisation of the labour market in Nigeria in terms of the gender wage gap over time.Design/methodology/approachThe study applied an extension of Oaxaca–Blinder decomposition that relies on the re-centred influence function (RIF) regressions to analyse the gender wage gap at all points along the wage distribution.FindingsThe results unambiguously show that there is a significant gender wage gap in Nigeria at all points along the wage distribution, such that for the two surveys used and after nearly two decades, men still earn more than women. That is, the log wage difference between males and females is statistically significant at all points between the 10th and the 90th quantiles. In 2003–2004 period, the authors found that most of the wage difference was significantly accounted for by the wage structure effect, whilst the composition effect was negative and only significant at the bottom of the wage distribution. Since the 2018–2019 period, the authors found that there has been a visible change such that most of the gender wage gap is now accounted for by the composition effect at all points along the wage distribution. Another interesting finding is that there has been a general decline in the gender wage gap along the entire wage distribution, such that inequality was higher in 2003–2004 than in 2018–2019. This decline is bigger at the top than at the bottom of the wage distribution. The authors also found that, contrary to some of the studies on the wage gap, the raw gaps for the two surveys appear to show inverted U-shape, but the gap has fallen quickly since the 2018–2019 period. Thus, the authors found strong evidence of a “sticky floor” compared to a “glass ceiling” effect in both periods, and this becomes more pronounced over time. In terms of the contributions of individual covariates on gender pay gap in Nigeria, the authors found that urban residence, unionisation, education and occupation variables exhibit major influence. However, the effects of covariates on the composition and wage structure components of the wage gap have changed over time.Practical implicationsThe major policy implication of these findings is that to address the gender wage gap in Nigeria, policy should focus more on how labour is rewarded and improving human capital for women.Originality/valueThis study is a novel paper in Nigeria that has investigated the gender wage gap in Nigeria by extending the focus of literature in three ways. First, the authors applied an extension of Oaxaca–Blinder decomposition that relies on the RIF regressions to analyse the gender wage gap at all points along the wage distribution. Second, the authors used sample selection bias to account for the non-randomness of participation in wage employment. And third, the authors applied similar analysis to two waves of household surveys (in 2003/2004 and 2018/2019) in order to understand the dynamics or polarisation of the labour market in Nigeria in terms of the gender wage gap over time.

Does environmental innovation improve investment efficiency?

Using a sample of 12,550 firm-year observations for the US over the period 2009–2021, we empirically examine the impact of environmental innovation on firm-level investment efficiency. Using fixed-effect regressions, our results indicate that eco-innovation has a significant and positive impact on firm-level investment efficiency. We show that eco-innovation raises investment efficiency and reducing information asymmetry enhances this effect. In addition, we show that the impact of eco-innovation on investment efficiency is statistically significant at environmentally sensitive firms, unlike firms that are environmentally insensitive. Further, the impact is higher at firms with low leverage than at firms with high leverage, in terms of both magnitude and economic significance, and the relative significance of this relationship is more pronounced at large firms. Our results are robust to alternative measures of eco-innovation, endogeneity, and sample selection bias.

Likelihood Analysis of the Factors Effecting Public Opinion of Police Performance in Pakistan

This study identifies behavioral and socio economic determinants of public satisfaction regarding police services in Pakistan. We incorporated the behavioral aspects of society through social networking or social awareness and the usage of technology as additional determinants of police performance. The Probit model with sample selection (heckprobit) is employed, in order to avoid sample selection bias. The findings reveal that the likelihood of satisfaction of male household heads is significantly low. Furthermore, wealthier households are more satisfied; which also raises a question about police performance. The results strongly recommend that the police department needs to improve its behavior, as urban society is less satisfied. Furthermore, the government should reduce the transaction cost of police services and other barriers that keep poor people away from the police; as the poor households are less satisfied. Finally, the social awareness and media is also a significant determinant of public satisfaction. The study finding reveals that the access to media and technology variables indicates that the families who have these facilities have a significantly lower probability of satisfaction in Pakistan.

Trial Selection and Estimating Damages Equations

Abstract Many studies have employed regression analysis with data drawn from court opinions. For example, an analyst might use regression analysis to determine the factors that explain the size of damages awards or the factors that determine the probability that the plaintiff will prevail at trial or on appeal. However, the full potential of multiple regression analysis in legal research has not been realized, largely because of the sample selection problem. We propose a method for controlling for sample selection bias using data from court opinions.

Informal sector in India and adoption of digital technologies

PurposeAccording to the Government of India’s Ministry of Labour and Employment Report (2015), almost 90% of the Indian workforce can still be categorized as informally employed, generating approximately 50% of the national product. Challenges with data availability have made a rigorous analysis of the informal economy in India often difficult and inadequate for policy formulations. This study aims to fill the gap by providing an empirical analysis of the informal economy in India using micro-data from the World Bank’s Informal Enterprise Surveys.Design/methodology/approachThe authors contribute by empirically testing the association between the adoption of digital technology (payments) and firm performance proxied by firm sales. Matching models are used to mitigate sample selection bias arising out of simultaneous sample selection.FindingsThe results suggest that the participation in digital platforms, namely, use of digital payment instruments, is associated with higher sales for firms. The results of this study also show that adoption of digital payments helps in both situations – whether a firm has been using digital technology or has just started using it since the outbreak.Research limitations/implicationsMore in-depth data over time, spanning across more cities of India, is needed to conduct a further detailed investigation.Social implicationsThe results should allow policymakers in India to reconsider youth-centric and women-centric business needs, even within the informal sector, which does not often enter the purview of the government but remains responsible for the growth and sustenance of 90% of the country’s workforce. If further research on this issue could engage with the impact of demonetization of currency in 2016 as a lagged shock on sales and reestimate subsequent growth, it would perhaps offer a wider spectrum of how the performance of the informal economy in India affects the entire economy, which has over the last four years and before the onset of Covid reported slower growth.Originality/valueProductivity is measured in terms of sales of informal firms in India in a regular month or in recent period like last month. Adoption of technology such as making payments using digital platforms can enhance productivity of firms by lowering standard transaction costs and time spent for visiting banks or financial institutions. Albeit not extensively, the literature has investigated digital technology adoption in the context of firms achieving comparative advantage (D’Ippolito et al. 2019; Scuotto et al. 2017), firms generating value creation (Magistretti, Dell’Era and Petruzzelli, 2019), and in helping with strategic initiatives and agility of firms (Ghezzi and Cavallo, 2018; Piccoli and Ives, 2005). Nonetheless, it would incur certain fixed costs, including acquiring skills and awareness, to manage digital platforms. In addition, physical access to instruments such as smartphones or computers and internet connectivity are prerequisites for productivity enhancements. Firms belonging to the informal sector in India generally face these challenges but may also benefit significantly following successful adoption. To the best of the authors’ knowledge, this is the first study to conduct a preliminary empirical analysis of the impact of digital technology adoption on the performance of informal sector firms in India.

Revisiting the predisposing, enabling, and need factors of unsafe abortion in India using the Heckman Probit model.

Unsafe abortion refers to induced abortions performed without trained medical assistance. While previous studies have investigated predictors of unsafe abortion in India, none have addressed these factors with accounting sample selection bias. This study aims to evaluate the contributors to unsafe abortion in India by using the latest National Family Health Survey data conducted during 2019-2021, incorporating the adjustment of sample selection bias. The study included women aged 15 to 49 who had terminated their most recent pregnancy within five years prior to the survey (total weighted sample (N) = 4,810). Descriptive and bivariate statistics and the Heckman Probit model were employed. The prevalence of unsafe abortion in India was 31%. Key predictors of unsafe abortion included women's age, the gender composition of their living children, gestation stage, family planning status, and geographical region. Unsafe abortions were typically performed in the early stages of gestation, often involving self-administered medication. The primary reasons cited were unintended pregnancies and health complications. This study underscores the urgent need for targeted interventions that take into account regional, demographic, and social dynamics influencing abortion practices in India.

The political consequences of wartime sexual violence: Evidence from a list experiment

Sexual violence is a prevalent feature of war with severe physical, psychological, and social consequences for survivors. Yet we have a limited understanding of how survivors relate to their political environment after the conflict ends. We analyze individual-level survey data on postwar Sri Lanka to assess whether wartime sexual victimization relates to political activism. Connecting unobtrusive measures from a list experiment to individual survivors’ political action, we show that personal experience of sexual violence increases political participation. This effect is substantial in size, holds for institutionalized and non-institutionalized forms of political action, and is robust to unobserved confounding or sample selection bias. Causal mediation analyses suggest that survivors of wartime sexual violence mobilize politically through their involvement in civic networks. The findings stress the relevance of survivors’ agency and contribute to a better understanding of wartime sexual violence, the role of civil society in post-conflict politics, and of humanitarian policy.

Measuring under-5 mortality and fertility through mobile phone surveys: an assessment of selection bias in 34 low-income and middle-income countries

ObjectivesThis study aims to assess sample selection bias in mobile phone survey estimates of fertility and under-5 mortality.DesignWith data from the Demographic and Health Surveys, we use logistic regressions to...

Investment decisions and small and medium-sized enterprise indebtedness: Heckman’s two-stage approach

PurposeInsufficient internal financing capacities and challenges to accessing external finance are crucial to small and medium-sized enterprises (SMEs) investment and growth. This study aims to investigate how SME leverage of bank financing is related to the investment decision.Design/methodology/approachUsing Heckman’s two-step econometric modelling to correct for sample selection bias, this study investigates the effect of entrepreneur characteristics, firm characteristics and performance on firms’ capital structure choices conditional on new investment decisions.FindingsThe main results reveal that larger firms with growth aspirations tend to make new investments. In the second stage equation, empirical results demonstrate that among SMEs who made a new investment, those SMEs with highly educated owner/managers, on average, use more external financing (i.e. banks loan) rather than internal funds – also, the smaller the company, the less bank leverage. Compared to the limited liability legal form, SMEs registered as individual businesses have less bank financial leverage. These results confirm that internal capacities for funding new investments are limited, and hence small firms must rely on external finance.Originality/valueThis study provides a unique empirical investigation and evidence based on a sample of SMEs in Kosovo. To the best of the authors’ knowledge, this study is the first attempt to empirically analyse investment behaviour in relation to capital structure for SMEs in Kosovo and one of the few, in general, to consider the sample selection bias issues underpinning the other studies in this field. The analysis corrects for sample selection bias, using growth aspiration as an instrumental variable.

Transitions in and out of work across the life course of patient with a NCD diagnosis

Abstract Background Coupled with its impact on quality of life, non-communicable Diseases (NCD) may degrade both the quality and quantity of the workforce, and stunt economic growth and productive employment. However, there is limited information about the social costs of NCDs to identify priorities in resource allocation, particularly in regard to retirement policies. In this paper, we investigate if NCDs affect them employment status, i.e., the patients’ ability to work beyond middle age, using longitudinal data. Methods We used longitudinal data from employed men and women who responded to the sixth and seventh waves of the SHARE project from several European countries and Israel. We restricted the sample to all people working up to 3 years before the expected retirement age for each country. To achieve a representative sample of employed people from each country we calculated calibrated longitudinal individual weights for participants in these two waves. It was to adjust for non-response, sample selection bias, and attrition (N = 9645). We modelled the probability to change employment status on Wave 7 as function of having a NCD, adjusting for age, gender, marital status, job satisfaction, education level, and country, using multinominal logistic regressions. Results Overall, 30.3% of the study population self-reported one NCD and 28.3% mentioned at least two chronic health conditions (multimorbidity). Among multimorbid individuals the risk of unemployment is 1.27 times more (OR = 1.271, 95% CI (1.266-1.276).), while the risk of retirement is multiplied by 1.26 (OR = 1.259, 95%CI (1.255-1.262)). Conclusions Having one NCD or multimorbidity is a common phenomenon in the working middle-aged European population. It is associated with unemployment and retirement, controlling for educational level, age, marital status, gender, job satisfaction, and country. This result highlights that early retirement will persist unless health policies are implemented to decrease the burden of NCDs. Key messages • This study shows that there is a strong association between having one or more NCD with unemployment and retirement, controlling for educational, age, and occupational type. • Results reinforce that we need health policies focused on alerting society about real impact of NCD on employment.

Venture capital directors and corporate debt structure: An empirical analysis of newly listed companies

We investigate how venture capitalists (VCs) serving as directors on corporate boards affect portfolio companies’ debt structure after initial public offerings (IPOs). Using hand-collected data, we find that companies with a higher fraction of VC directors on their boards use significantly fewer types of debt. The impact of VC directorships on debt concentration is more pronounced in companies facing greater expected bankruptcy costs or higher degrees of uncertainty. We further explore the benefits of debt concentration and find that a highly concentrated debt structure is associated with better corporate performance in companies with VC directors. Taken together, our evidence suggests that VC directors influence newly listed companies to adopt a concentrated debt structure, thus minimizing the risk of distress and enhancing company value. Our results are robust to accounting for endogeneity and sample selection bias.

Impact of relocation in response to climate change on farmers’ livelihood capital in minority areas: a case study of Yunnan Province

PurposeThis study aims to construct an evaluation system for farmers’ livelihood capital in minority areas and evaluate the impact of relocation in response to climate change on farmers’ livelihood capital.Design/methodology/approachAccording to the characteristics of Yunnan minority areas, the livelihood capital of farmers in minority areas is divided into natural, physical, financial, social, human and cultural capital. The improved livelihood capital evaluation system measures farmers’ livelihood capital from 2015 to 2021. The net impact of relocation on farmers’ livelihood capital was separated using propensity score matching and the difference-in-difference (PSM-DID) method.FindingsThe shortage of livelihood capital makes it difficult for farmers to resist climate change, and the negative impacts of climate change further aggravate their livelihood vulnerability and reduce their livelihood capital. Relocation has dramatically increased the livelihood capital of farmers living in areas with poor natural conditions by 15.67% and has enhanced their ability to cope with climate change and realise sustainable livelihoods.Originality/valueAn improved livelihood capital evaluation system is constructed to realise the future localisation and development of livelihood capital research. The PSM-DID method was used to overcome endogeneity problems and sample selection bias of the policy evaluation methods. This study provides new ideas for academic research and policy formulation by integrating climate change, poverty governance and sustainable livelihoods.