Marginal Effect Analysis Research Articles

The long run health consequences of rural‐urban migration

Rural‐urban migration is an integral part of the structural transformation as societies move from a traditional agricultural economy to a modern economy. This process has many potential consequences for migrants. Our study focuses on the lifetime health effects of the large mid‐20th century migration out of rural U.S. Northern Great Plains states, primarily to urban locations in the West and Midwest. An analysis of marginal treatment effects (MTEs) shows that (a) migrants are positively selected, and (b) the causal impact of migration is decreased longevity. Our evidence suggests that elevated mortality among migrants is linked to increased smoking and alcohol consumption.

Perceived happiness, perceived trust and perceived income levels: The case of the reunified Germany

This study explored the possible impact of perceived income on individual (perceived) happiness in Eastern and Western Germany in relation to perceived trust and four socio-economic variables, namely gender, age, marital status and employment status. To examine the relationship between these variables, a generalized ordered logit model was applied using the World Values Survey data. Bootstrapping and marginal effects were used to obtain a more robust model. The findings provided insights regarding the impact of perceived income and perceived trust on individual (perceived) happiness in both regions after reunification. Perceived income had a positive effect on all happiness categories in both regions. Perceived trust had a stronger positive impact on individual happiness than that of perceived income, although its significance varied across individual (perceived) happiness categories. Analysis of marginal effects revealed differences between the base models.

Willingness to pay for improved public health care services in Saudi Arabia: a contingent valuation study among heads of Saudi households.

The bulk of health care service provision in Saudi Arabia is undertaken by the public health care sector through the Ministry of Health, which is funded annually by the total government budget, which, in turn, is derived primarily from oil revenue. Public health care services in Saudi Arabia are characterised by an overload, overuse, and shortage of medical personnel, which can result in dissatisfaction with the quality of the current public health care services. This study uses a contingent valuation method to investigate the willingness of Saudi people to pay for improvements to the quality of public health care services. This study also determines the association between the willingness to pay for quality improvements and respondents' demographic and socioeconomic characteristics. A pre-tested interviewer-administered questionnaire was used to collect data from 1187 heads of household in Jeddah Province over a five-month period. Multi-stage sampling was employed to recruit participants. Partial Tobit regression and corresponding marginal effects analyses were used to analyse the data. These empirical analyses show that the majority of the sample was willing to pay for quality improvements in the public health care services. The results of this study might be of use to policymakers to help with both priority setting and fund allocation.

Overeducation, perceived career progress, and work satisfaction in young adulthood

Some people work in occupations that require lower levels of education than their attained education, and these “overeducated” workers tend to be less satisfied than those who work in occupations that match their attained education. This study sought to extend the previous finding by answering the following previously unexplored questions: (1) Does the association depend on the level of attained education?; and (2) Does perceived career delay account for overeducated workers’ lower work satisfaction? Data from U.S. young adults (the National Longitudinal Study of Adolescent to Adult Study) showed that the association between overeducation and work satisfaction was limited to severely overeducated workers with bachelor’s or graduate degrees and moderately overeducated workers with high school degrees. These people also assessed their career progress more negatively, which accounted for their lower work satisfaction. Analysis of marginal effects demonstrated that among people with bachelor’s degree, the effect of severe overeducation was strong enough to cancel out the benefit of holding the degree to improve work satisfaction.

The impact of land ownership, firefighting, and reserve status on fire probability in California

The extent of wildfires in the western United States is increasing, but how land ownership, firefighting, and reserve status influence fire probability is unclear. California serves as a unique natural experiment to estimate the impact of these factors, as ownership is split equally between federal and non-federal landowners; there is a relatively large proportion of reserved lands where extractive uses are prohibited and fire suppression is limited; and land ownership and firefighting responsibility are purposefully not always aligned. Panel Poisson regression techniques and pre-regression matching were used to model changes in annual fire probability from 1950–2015 on reserve and non-reserve lands on federal and non-federal ownerships across four vegetation types: forests, rangelands, shrublands, and forests without commercial species. Fire probability was found to have increased over time across all 32 categories. A marginal effects analysis showed that federal ownership and firefighting was associated with increased fire probability, and that the difference in fire probability on federal versus non-federal lands is increasing over time. Ownership, firefighting, and reserve status, played roughly equal roles in determining fire probability, and were found to have much greater influence than average maximum temperature (°C) during summer months (June, July, August), average annual precipitation (cm), and average annual topsoil moisture content by volume, demonstrating the critical role these factors play in western fire regimes and the importance of including them in future analysis focused on understanding and predicting wildfire in the Western United States.

Beyond LATE with a Discrete Instrument

We show how a discrete instrument can be used to identify the marginal treatment effects under a functional structure that allows for treatment heterogeneity among individuals with the same observed characteristics and self-selection based on the unobserved gain from treatment. Guided by this identification result, we perform a marginal treatment effect analysis of the interaction between the quantity and quality of children. Our estimates reveal that the family size effects vary in magnitude and even sign and that families act as if they possess some knowledge of the idiosyncratic effects in the fertility decision.

An Exploratory Study of Attendee Willingness to Pay for Environmentally Friendly Meetings

ABSTRACTThe study examined meeting attendees’ willingness to pay (WTP) more for green meetings. The survey data were used to measure the impact of demographic characteristics on WTP a price premium for green meetings. The ordered-probit model was used to estimate the impact of the explanatory factors on the probabilities of meeting attendees’ WTP. Analysis of marginal effects suggested that meeting attendees who were willing to sacrifice convenience associated a green meeting were more likely to pay more for green meetings. The study also found that the green meeting attendee profile was female, married, younger age, and with a high level of education, but men and respondents with higher incomes were less willing to pay more for green meetings when respondents paid their own registration fees.

Determinants of Participation of Dairy Farm Women in Women Dairy Self Help Groups in Belagavi district of Karnataka

The present study was conducted in Belagavi district of Karnataka to analyse the determinants of participation of dairy farm women in Women Dairy Self Help Groups (WDSHGs). Three talukas and from each taluka three villages were selected purposively based on the presence of Women Dairy Self Help Groups (WDSHGs). Finally, two WDSHGs were selected randomly from each village and from each WDSHG two members were selected randomly making a sample of 90 members. Also, 90 non-members having similar socio-economic status were selected, thus making the total sample size to 180. Primary data was collected on various variables and analysed by using the logistic regression and marginal effect technique. The result showed that age of the farmer and family size had a positive and significant effect on farm women participation in WDSHGs in the study area. This suggest that as the age and family size of the farm women increases, the more likely they are to become a member of WDSHGs. Between these two variables, the family size was found to have more influence on participation as seen from the marginal effect analysis. Education level of the dairy farm women also positively influenced their participation. On the other hand, the number of milch animals owned had a negative impact implying that large dairy farm women are less likely to join WDSHGs. Alternatively this points out that small and marginal dairy farm women are more likely to be members of WDSHGs

The limitations of negative incomes in the Gini coefficient decomposition by source

ABSTRACTLerman and Yitzhaki (1985) developed a decomposition of the Gini coefficient by income source that has been extensively used in the literature. This method has strong limitations in the presence of negative incomes, which were not discussed by the original authors and have been widely overlooked in successive studies. Through theoretical argumentation and practical examples, this article shows that, when using negative incomes, (1) the original decomposition formulae become inappropriate, (2) the marginal effects analysis may yield erroneous results and (3) the Pigou–Dalton ‘principle of transfers’ is not always met. This has critical implications for policy development, given that strategies based upon incorrect analyses could actually result in undesired greater income inequalities. The Gini source decomposition should be carefully applied by researchers and policymakers, especially in rural developing areas, where negative incomes are common due to financial losses from agricultural activities.

Combination Social Protection for Reducing HIV-Risk Behavior Among Adolescents in South Africa.

Social protection (ie, cash transfers, free schools, parental support) has potential for adolescent HIV prevention. We aimed to identify which social protection interventions are most effective and whether combined social protection has greater effects in South Africa. In this prospective longitudinal study, we interviewed 3516 adolescents aged 10-18 between 2009 and 2012. We sampled all homes with a resident adolescent in randomly selected census areas in 4 urban and rural sites in 2 South African provinces. We measured household receipt of 14 social protection interventions and incidence of HIV-risk behaviors. Using gender-disaggregated multivariate logistic regression and marginal effects analyses, we assessed respective contributions of interventions and potential combination effects. Child-focused grants, free schooling, school feeding, teacher support, and parental monitoring were independently associated with reduced HIV-risk behavior incidence (odds ratio: 0.10-0.69). Strong effects of combination social protection were shown, with cumulative reductions in HIV-risk behaviors. For example, girls' predicted past-year incidence of economically driven sex dropped from 11% with no interventions to 2% among those with a child grant, free school, and good parental monitoring. Similarly, girls' incidence of unprotected/casual sex or multiple partners dropped from 15% with no interventions to 10% with either parental monitoring or school feeding, and to 7% with both interventions. In real world, high-epidemic conditions, "combination social protection," shows strong HIV prevention effects for adolescents and may maximize prevention efforts.

JAM: A Scalable Bayesian Framework for Joint Analysis of Marginal SNP Effects.

ABSTRACTRecently, large scale genome‐wide association study (GWAS) meta‐analyses have boosted the number of known signals for some traits into the tens and hundreds. Typically, however, variants are only analysed one‐at‐a‐time. This complicates the ability of fine‐mapping to identify a small set of SNPs for further functional follow‐up. We describe a new and scalable algorithm, joint analysis of marginal summary statistics (JAM), for the re‐analysis of published marginal summary stactistics under joint multi‐SNP models. The correlation is accounted for according to estimates from a reference dataset, and models and SNPs that best explain the complete joint pattern of marginal effects are highlighted via an integrated Bayesian penalized regression framework. We provide both enumerated and Reversible Jump MCMC implementations of JAM and present some comparisons of performance. In a series of realistic simulation studies, JAM demonstrated identical performance to various alternatives designed for single region settings. In multi‐region settings, where the only multivariate alternative involves stepwise selection, JAM offered greater power and specificity. We also present an application to real published results from MAGIC (meta‐analysis of glucose and insulin related traits consortium) – a GWAS meta‐analysis of more than 15,000 people. We re‐analysed several genomic regions that produced multiple significant signals with glucose levels 2 hr after oral stimulation. Through joint multivariate modelling, JAM was able to formally rule out many SNPs, and for one gene, ADCY5, suggests that an additional SNP, which transpired to be more biologically plausible, should be followed up with equal priority to the reported index.

Rolling stock purchase cost for rail and road public transportation: random-parameter modelling and marginal effect analysis

ABSTRACTPublic transportation rolling stock is a major item of capital expenditure associated with transit facility construction, expansion, or revitalisation. As such, reliable estimates of the expected purchase costs of rolling stock are critical for analysing the financial feasibility of proposed public transportation systems and financial planning for existing systems. Such cost data are often characterised by heterogeneity of observations that often plagues fixed parameter (FP) model estimation. Therefore, in modelling transit rolling stock costs, it is needed to address such limitations of past studies. Using data from the National Transit Database, this paper addresses the issue through the random parameter (RP) estimation modelling technique, where parameter estimates are allowed to vary across observations according to the most appropriate distribution. The paper demonstrates that the RP modelling technique yields superior estimates compared to the FP model. This is because the former can account for unobserved heterogeneity in transit rolling stock purchase cost data due to unobserved factors that otherwise severely impair the reliability or intuitiveness of the model. The paper also carries out marginal effect and sensitivity analyses on the developed model to acquire a deeper insight into the influence of the cost factors. The RP rolling stock cost models developed in this paper can be applied by public transportation agencies to carry out strategic tasks associated with the development or management of new or existing transit systems, respectively. These tasks include analyses of financial feasibility of proposed transit systems, development of budgets, and projections of cash outflows as part of long-range plans, financial planning, and short- and long-term funding needs assessment for existing or proposed transit systems. Moreover, the findings regarding the influential cost factors can be used to guide purchasing policies for transit rolling stock.

A Cross-cultural study of emotional Responses on Colours

This study sought to provide empirical evidence on the effect of social capital indexes on the poverty status of the rural household in Nigeria. Multistage sampling procedure was used to collect primary data using pretested and structured questionnaire. Analytical techniques employed include descriptive statistics, FGT measure of poverty, probit model and marginal effect analysis. Findings from the study showed that the mean age of the household heads in the study area was 50.67 ± 12.76. It was observed that most households were headed by male (89.00%). Mean expenditure among the households in the study area was N11,877.00 ± N7,066.80 per month. About 60.00% of the households in the study area were poor. Membership density index indicated that household heads within the age range of 41 - 60 years (or 44.95%) were members of more social capital networks compared to other age groups. Payback period (β = 0.022, p< 0.05), amount of microcredit received (β =0.001, p<0.1), microcredit processing time (β = -0.338, p< 0.01), cash contribution (β =0.012, p< 0.05), meeting attendance (β =0.007, p< 0.1), membership density (β = -0.047, p< 0.01) and labour contribution indexes (β = -0.008, p< 0.05) were the significant determinants of household’s poverty status among the sampled households. The concluded that poverty level was very high in the study area and that social capital indexes were major factors influencing households’ poverty status. Therefore, encouraging the poor households to join and participate in social capital networks could help to improve their poverty status.

Predicting Accounting Fraud: Evidence from Japan

This study develops a prediction model for identifying accounting fraud by analyzing the accounting information for Japanese firms. In particular, we (1) explore the characteristics of accounting fraud firms by analyzing financial information obtained from annual reports (yukashoken-houkokusho in Japanese) and (2) develop a model for predicting accounting fraud based on the characteristics of Japanese fraud firms. To identify the characteristic of fraud firms, we focus on 39 variables for the eight factors of “accruals quality,” “performance,” “nonfinancial measures,” “off-balance-sheet activities,” “market-related incentives,” “conservatism,” “real-activities manipulation,” and “Japanese-specific factors.” Through our univariate analysis and model building process, we find that “accrual quality,” “market-related incentives,” “real-activities manipulation,” “conservatism” and “Japanese-specific factors” are generally useful for detecting accounting fraud. We also conduct several analyses that test the predictive ability of our models, including (1) the detection rates of fraud firms, (2) Type I and Type II error rates, (3) marginal effect analysis on independent variables, and (4) robustness tests on time periods and industry clustering. We find that our models have generally higher predictive power in detecting accounting fraud. We expect that our models can be used widely in various accounting and finance practices.

Club convergence and factors of digital divide across countries

By examining the digital divide phenomenon among 108 countries, this paper attempts to develop a greater understanding of the factors that can foster a more creative global economy. As part of this investigation, a new ICT development index is constructed using principal component analysis and the inclusion of conventional ICT development indicators. Using Phillips and Sul's (2007) log t convergence test, we found that the level of digitalization convergence that exists among the 108 countries analyzed can be categorized into three different groupings. Within these groups, we found that the 1st group which contains the highest level of convergence showed the slowest speed of convergence, while the results for the 3rd group displayed the lowest convergence level with the highest speed of convergence among its members in digitalization. This study also identified the factors that drive a country's digitalization convergence level, these included: per capita GDP, tertiary education entrance rate, the ratio of urban population, and the share of service trade in GDP. With the exception of urban population, which showed results to the contrary, all of these factors were proven to raise a country's likelihood to belong to a higher convergence level. As for policy measures that can be used to foster a creative economy, marginal effect analysis identifies the tertiary education entrance rate as being the most effective means for raising the digitalization convergence level.

Assessing willingness to pay for health care quality improvements.

BackgroundContingent valuation (CV) is used to estimate the willingness to pay (WTP) of consumers for specific attributes to improve the quality of health care they received in three hospitals in Bangladesh.MethodsRandom sample of 252 patients were interviewed to measure their willingness to pay for seven specified improvements in the quality of delivered medical care. Partial tobit regression and corresponding marginal effects analysis were used to analyze the data and obtain WTP estimates.ResultsPatients are willing to pay more if their satisfaction with three attributes of care are increased. These are: a closer doctor-patient relationship, increased drug availability and increased chances of recovery. The doctor patient relationship is considered most important by patients and exhibited the highest willingness to pay.ConclusionsThis study provides important information to policy makers about the monetary valuation of patients for improvements in certain attributes of health care in Bangladesh.Electronic supplementary materialThe online version of this article (doi:10.1186/s12913-015-0678-6) contains supplementary material, which is available to authorized users.

Factors Shaping Interactions Among Community Health Workers in Rural Ethiopia: Rethinking Workplace Trust and Teamwork

Worldwide, a shortage of skilled health workers has prompted a shift toward community-based health workers taking on greater responsibility in the provision of select maternal and newborn health services. Research in mid- and high-income settings suggests that coworker collaboration increases productivity and performance. A major gap in this research, however, is the exploration of factors that influence teamwork among diverse community health worker cadres in rural, low-resource settings. The purpose of this study is to examine how sociodemographic and structural factors shape teamwork among community-based maternal and newborn health workers in Ethiopia. A cross-sectional survey was conducted with health extension workers, community health development agents, and traditional birth attendants in 3 districts of the West Gojam Zone in the Amhara region of Ethiopia. Communities were randomly selected from Maternal and Newborn Health in Ethiopia Partnership (MaNHEP) sites; health worker participants were recruited using a snowball sampling strategy. Fractional logit modeling and average marginal effects analyses were carried out to identify the influential factors for frequency of work interactions with each cadre. One hundred and ninety-four health workers participated in the study. A core set of factors-trust in coworkers, gender, and cadre-were influential for teamwork across groups. Greater geographic distance and perception of self-interested motivations were barriers to interactions with health extension workers, while greater food insecurity (a proxy for wealth) was associated with increased interactions with traditional birth attendants. Interventions that promote trust and gender sensitivity and improve perceptions of health worker motivations may help bridge the gap in health services delivery between low- and high-resource settings. Inter-cadre training may be one mechanism to increase trust and respect among diverse health workers, thereby increasing collaboration. Large-scale, longitudinal research is needed to understand how changes in trust, gender norms, and perceptions of motivations influence teamwork over time.

Workplace mistreatment and sickness absenteeism from work: Results from the 2010 National Health Interview survey

This study examined the association between workplace mistreatment and occurrence, duration, and costs of sickness absenteeism. We used the 2010 National Health Interview Survey and considered 13,807 employed adult respondents. We used a zero-inflated negative binomial (zinb) model to examine the association between exposure to workplace mistreatment and the occurrence and number of workdays missed due to illness/injury in the preceding 12 months. In 2010, 7.6% of US workers employed at the time of the survey reported having been mistreated at their workplace. Both occurrence and duration of sickness absence were higher for mistreated than for non-mistreated workers. The zinb results showed that being mistreated was associated with a 42% increase in the number of missed workdays, controlling for covariates. The marginal effect analysis showed that lost workdays differed by 2.45 days between mistreated and non-mistreated workers. This implies that workplace mistreatment was associated with $4.1 billion, or 5.5%, of sickness absenteeism costs in 2010. Workplace mistreatment is associated with sickness absence in the United States. While a causal relationship could not be established due to the cross-sectional design of the study, this study reveals the economic importance of developing workplace mistreatment prevention strategies.

Investigating the influence of curbs on single-vehicle crash injury severity utilizing zero-inflated ordered probit models

The severity of traffic-related injuries has been studied by many researchers in recent decades. However, previous research has seldom accounted for the effects of curbed outside shoulders on traffic-related injury severity. This study applies the zero-inflated ordered probit (ZIOP) model to evaluate the influences of curbed outside shoulders, speed limit change, as well as other traditional factors on the injury severity of single-vehicle crashes. Crash data from 2003 to 2007 in the Illinois Highway Safety Database were employed in this study.The ZIOP model assumes that injury severity comes from two distinct sources: injury propensity and injury severity when this crash falls into the injury prone category. The modeling results show that on one hand, single-vehicle crashes that occurring on roadways with curbed outside shoulders are more likely to be injury prone. On the other hand, the existence of a curb decreases the likelihood of severe injury if the crash was in the injury prone category. As a result, the marginal effect analysis implies that the presence of curbs is associated with a higher likelihood of no injury and minor injury involved crashes, but a lower likelihood of incapacitating injury and fatality involved crashes. In addition, in the presence of curbed outside shoulders, the change of speed limit adds no significant impact to the injury severity of single-vehicle crashes.Moreover, the modeling results also highlight some interesting effects caused by vehicle type, light and weather conditions, and drivers’ characteristics, as well as crash type and location. Through a comprehensive evaluation of the modeling results, the authors find that the ZIOP model performs well relative to the traditional ordered probit (OP) model, and can serve as an alternative in future studies of crash injury severity.

Severity of Injuries to Motor Vehicle Drivers at Highway–Rail Grade Crossings in the United States

There are approximately 240,000 highway–rail grade crossings in the United States. High crash frequencies at these locations have led to continued research in safety modeling. Existing crash models for highway–rail grade crossings can be classified into two categories: models for predicting accident frequency and models of the severity of driver injuries. The majority of these studies have focused on the first category. Few studies have focused on the severity of injuries to motor vehicle drivers at highway–rail grade crossings. The objective of this study was to determine the contributing factors that influence the severity of driver injuries in accidents at highway–rail grade crossings. Probit models showed that the following factors were significant: whether the crash occurred during the peak hour, weather, visibility, vehicle type, vehicle speed, annual average daily traffic, train speed, driver age and gender, area type, and type of highway pavement. A marginal effects analysis was also conducted to quantitatively interpret the marginal effects of the contributing factors on each severity level for the highway driver.