Instrumental Variable Regression Model Research Articles

Mandatory pension savings and long-run debt accumulation: evidence from Danish low-wage earners

AbstractBased on two decades of Danish register data at the individual level, this paper finds that a 1-dollar increase in pension wealth leads to a 42-cent rise in total debt for a group of low-wage earners. Collective bargaining in the labor market provides time-sector variation in mandatory pension contribution rates, which we exploit in two empirical research designs; an event study and a cross-sectional instrumental variable regression model. Both methods demonstrate that the debt rise is accompanied by increased housing wealth and homeownership rates. Together, the empirical evidence indicates that mandatory pension contributions lead to a significant increase in net wealth, as well as in gross debt.

Birth Weight and Cognitive Development during Childhood: Evidence from India

Health at birth is an important indicator of human capital development over the life course. This paper uses longitudinal data from the Young Lives survey and employs instrumental variable regression models to estimate the effect of birth weight on cognitive development during childhood in India. We find that a 10 per cent increase in birth weight increases cognitive test scores by 0.11 standard deviations at 5–8 years of age. Low birth weight infants experience a lower test score compared with normal birth weight infants. The positive effect of birth weight on a cognitive test score is larger for girls, children from rural households and those with less educated mothers. Our findings suggest that health policies designed to improve birth weight could improve human capital in resource‐poor settings.

Effects of screening for anxiety and depression in patients with ischaemic heart disease - a nationwide Danish register study.

To investigate the effect of screening for anxiety and depression (AD) in patients with ischaemic heart disease (IHD) on the likelihood of receiving treatment for AD. We used a nationwide dataset of all Danish patients with an incident IHD diagnosis in the period 2015-2018 (N = 80,701) of which 20,461 (25%) were exposed to screening for AD as part of cardiac rehabilitation. A binary composite indicator for the use of any AD treatment (prescriptions of AD drugs, general practitioner (GP) counselling or referral to a psychologist), was modelled as the dependent variable. The probability of receiving AD treatment was estimated using linear probability and instrumental variable regression models. Exposure to AD screening was lower for patients with low income (change in probability -0.67, 95% CI -0.76; -0.59), low education (change in probability -0.16, 95% CI -0.20; -0.13), and a high comorbidity burden (change in probability -0.09, 95% CI -0.10; -0.07). Screened patients had a lower conditional probability of AD treatment (change in probability -0.0061, p < 0.001) than non-screened patients. The patient's GP also had an impact on the probability of being referred for AD treatment. Using an instrumental variable approach did not affect the results. Screening for AD was subject to selection at the patient level; patients at lower risk of AD had a higher probability of being screened. Hence, extending systematic screening to cover a larger population may not achieve a noticeable increase in the uptake of AD treatment if it is not supported by appropriate measures to reduce reverse selection into screening.

Impacts of firewood burning for cooking on respiratory health and healthcare utilisation: Empirical evidence from Sri Lankan micro-data.

Despite escalating income levels of Sri Lankan households, 70.3% of them still depend on firewood for cooking. This might lead to bad respiratory health and thereby, increased healthcare utilisation. With this backdrop, this paper examines how firewood burning for cooking influences respiratory health and healthcare utilisation of Sri Lankans. The study proposes a theoretical model to explain effects of firewood burning on respiratory health and healthcare utilisation and empirically tests it using nation-wide micro-data of 79,170 individuals belonging to 21,748 households. The data are drawn from the most recent wave of Sri Lankan household income and expenditure survey. After addressing potential endogeneity by applying instrumental variable regression models, the study finds that firewood burning increases households' probability of asthma prevalence by 10.9 percentage points (P<0.001), out-patient care utilisation by 33.1 percentage points (P<0.001), and in-patient care utilisation by 17.5 percentage points (P<0.001), on average. Our individual level analysis demonstrates that females are more vulnerable to bad respiratory health induced by firewood burning (β=0.055, P<0.001) compared to males (β=0.008, P>0.1). The results imply policies on promoting improved-cooking stoves, separated-kitchen designs, and switching more towards cleaner energy sources, including LP gas and solar power.

The association between polygenic scores for attention-deficit/hyperactivity disorder and school performance: The role of attention-deficit/hyperactivity disorder symptoms, polygenic scores for educational attainment, and shared familial factors.

Polygenic scores (PGS) for attention-deficit/hyperactivity disorder (ADHD) negatively predicts educational attainment (EA), but it remains unclear how ADHD symptoms, PGS for EA, and shared familiar factors influence the associations between PGS for ADHD and school performance. We combined survey data on ADHD symptoms, PGS, and register-based, objective measures of compulsory school performance at age 16 for 6049 twins in the Child and Adolescent Twin Study in Sweden. Linear and instrumental variable (IV) regression models were used to estimate the association between PGS for ADHD and grade point average (GPA), overall and by natural science, humanities, and practically oriented (e.g., sports, arts, music) subject categories. The models were adjusted for parent-rated ADHD symptoms, PGS for EA, and shared familial factors (dizygotic twin comparisons) to examine how these factors influenced the associations between PGS for ADHD and school performance. PGS for ADHD were negatively associated with school performance; β=-0.12, 95% confidence interval = (-0.15, -0.09) for overall GPA with minor differences by subject category. Adjustment for ADHD symptoms attenuated these associations to a small degree compared to PGS for EA, and shared familial factors respectively. Stronger associations were observed using IV regressions compared to linear regression. However, in the IV regression analyses, most associations between PGS for ADHD and GPA in the practically oriented subject category were not significant. Associations between PGS for ADHD and school performance are to a small degree influenced by ADHD symptoms, compared to PGS for EA and shared familial factors. These results highlight important considerations for research using PGS for ADHD to control for genetic factors, and for future clinical applications aiming to determine genetic liability towards ADHD.

Willingness to Take COVID-19 Vaccines in Ethiopia: An Instrumental Variable Probit Approach.

This paper analyzed the factors influencing the willingness of Ethiopia’s population to take COVID-19 vaccines. The data included the COVID-19 High Frequency Phone Survey of Households in Ethiopia that were collected in 2021. This paper relied on the 10th round of the survey, which was comprised of 2178 households. The Instrumental Variable Probit regression model was used to analyze the data. The results showed that majority of the respondents (92.33%) would receiveCOVID-19 vaccines, while 6.61% and 1.06% were, respectively, unwilling and unsure. Across the regions of Ethiopia, Southern Nations, Nationalities, and Peoples’ Region (SNNPR) (99.30%), Oromia (97.54%), Tigray (97.04%) and Gambela (95.42%) had the highest proportions of respondents willing to have the vaccine. Vaccine safety concern was the topmost reason for those unwilling to receive the vaccine. The results of the Instrumental Variable Probit regression showed that currently working, age, engagement with non-farm businesses and region of residence significantly influenced the population’s willingness to take the vaccine (p < 0.05). It was concluded that although the willingness be vaccinated was impressive, without everyone being receiving the COVID-19 vaccine, infection risk can still be high; this is due to the persistent mutation of the viral strains. Thus, there is a need to intensify efforts toward addressing the safety issues of COVID-19 vaccines, while efforts to enhance acceptability should focus on the youth population and those who are unemployed.

Second-order refinements for [formula omitted]-ratios with many instruments

This paper studies second-order properties of the many instruments robust t-ratios based on the limited information maximum likelihood and Fuller estimators for instrumental variable regression models with homoskedastic errors under the many instruments asymptotics, where the number of instruments may increase proportionally with the sample size n, and proposes second-order refinements to the t-ratios to improve the size and power properties. Based on asymptotic expansions of the null and non-null distributions of the t-ratios derived under the many instruments asymptotics, we show that the second-order terms of those expansions may have non-trivial impacts on the size as well as the power properties. Furthermore, we propose adjusted t-ratios whose approximation errors for the null rejection probabilities are of order O(n−1) in contrast to the ones for the unadjusted t-ratios of order O(n−1/2), and show that these adjustments induce some desirable power properties in terms of the local maximinity. Although these results are derived under homoskedastic errors, we also establish a stochastic expansion for a heteroskedasticity robust t-ratio, and propose an analogous adjustment under slight deviations from homoskedasticity.

Health insurance enrollment and vision health in rural China: an epidemiological survey

BackgroundVision health is an important aspect of health worldwide. Visual impairment (VI) is associated with poor quality of life and is usually more prevalent in rural areas. To help rural populations obtain vision care, health insurance policies have emerged throughout the world. However, some existing literatures show that health insurance enrollment’s impact on the overall physical health of rural population has been minimal. Focusing on vision health among adults in rural China, our study aims to investigates the impact of health insurance on vision health, heterogeneity of the effect, and the moderating effect of health insurance enrollment on the impact of chronic physical diseases and basic eye diseases on vision health.MethodsPrimary data were collected through a nation-wide epidemiological survey of vision health conducted in rural China in 2018, with a sample size of 28,787 used in our statistical analysis. Instrumental variables regression and Heckman selection models were conducted to examine the impact of health insurance enrollment and reimbursement ratio adults’ vision health outcomes. Subsample regressions by sex, age, education level, and whether with eye diseases were further conducted to explore the heterogeneity in our results. We then examined whether health insurance enrollment moderates the impact of chronic physical diseases and basic eye diseases on vision health through the method of introducing interaction terms.ResultsParticipating in health insurance reduced the probability of VI by 2.15 %. The reimbursement rate increasing by 1 % point may reduce the probability of worsening VI by 6.12 %. Men (-0.0235, P = 0.0002) benefit more from insurance enrollment than women (-0.0201, P = 0.0082) with respect to vision health. From the young adult group to the oldest group, the marginal effect of health insurance increased from − 0.0068 (P = 0.0394) to -0.0753 (P < 0.0001). The marginal effect on VI was most significant in people with lower education levels and weakened with increased education levels. People with basic eye diseases (-0.0496, P = 0.0033) benefit more from participating insurance than the people without basic eye diseases (-0.0196, P = 0.0001) with respect to vision health. The moderating effects of health insurance enrollment on the impacts of cerebral infarction (-0.1225, P < 0.0001), diabetes (-0.0398, P = 0.0245), hyperlipidemia (-0.1364, P = 0.0271), mental illness (-0.1873, P = 0.0010), glaucoma (-0.1369, P = 0.0073), diabetic retinopathy (-0.1560, P = 0.0043), and retinal vein obstruction (-0.2018, P = 0.0155) on vision health were significantly negative.ConclusionsThe results suggest that participation in health insurance and higher health insurance reimbursement ratios reduced the risk of VI in the sampled adults. Health insurance has the most significant effect in in vulnerable groups. Heath insurance enrollment moderates the impacts of several chronic physical and basic eye conditions on vision health. Our findings have potential implications for reforming health insurance policies to improve vision health conditions in rural areas of developing countries.

How does population aging affect household carbon emissions? Evidence from Chinese urban and rural areas

Population aging has become a salient demographic phenomenon in China, thus causing social and economic concerns. However, little is known about the impact of population aging on the environment. The available literature mainly emphasizes decreases in energy consumption or carbon emissions caused by population aging, conveying limited information about the impact of population aging on household carbon emissions. This paper explores the impact of population aging on household carbon emissions in both urban and rural areas of China, using panel data of 30 provinces over 1997–2017. The relationship between population aging and household carbon emissions is investigated with a threshold regression model and a two-stage instrumental variable regression model. The results show that urban population aging increases urban household carbon emissions at levels both below and above 0.083, while this positive impact is comparatively smaller when urban population aging is above 0.083. However, at the national level, rural population aging has no significant impact on rural household carbon emissions at the level below 0.066, while rural population aging significantly increases rural household carbon emissions at the level above 0.066. Rural population aging has a significant positive effect on the increase in household carbon emissions in northern heating areas, when rural population aging is both below and above 0.077. The analysis of the underlying mechanism identifies consumption structure and consumption level as intermediary factors that affect the nonlinear relationship between urban population aging and urban household carbon emissions. Furthermore, the income level also affects the nonlinear relationship between rural population aging and rural household carbon emissions.

MO980OUTCOMES OF RABBIT ANTI-THYMOCYTE GLOBULIN VERSUS IL-2 RECEPTOR ANTAGONIST INDUCTION THERAPIES IN 2DR MISMATCH RENAL TRANSPLANT RECIPIENTS

Abstract Background and Aims 2DR HLA mismatch indicates high immunological risk renal transplant. Induction therapy with rabbit Anti-thymocyte Globulin (r-ATG) and IL-2 Receptor Antagonist (IL-2RA) resulted in marked reduction of acute allograft rejection rate and improved graft survival. However, the outcomes in 2DR (HLA-DR) mismatched renal transplant recipients (RTRs) in the era tacrolimus-mycophenolate mofetil maintenance immunosuppression remains understudied. Method This was a retrospective cohort study using data from the United States Organ Procurement and Transplantation Network, all 2 DR mismatched RTRs who were maintained on tacrolimus and mycophenolate mofetil immunotherapy between 2005 and 2015 were included. Follow up data was until September 2020. Patients who received transplants from living donors were included in the study. Collected data included recipient factors (age, sex, ethnicity, diabetes, body mass index), transplant factors (delayed graft function, cold ischemia time, number of previous transplants, panel reactive antibodies, HLA-mismatches, induction therapies, maintenance immunotherapy, and donor factors (donor type, donor age). RTRs were divided into 2 groups: based on induction therapy r-ATG and IL-2RA. Instrumental variable regression models were used to assess effect of induction therapy on acute rejection episodes at 6 months post-transplant and on graft survival. Type of induction therapy was instrumented for the transplant centre to reduce the centre effect on the choice of the induction therapy. Cox proportional hazard regression analysis was performed to assess the effect of induction therapy on graft survival. The regression models were adjusted for collected recipient, donor and transplant factors. Results 3052 patients received IL2-RA while 5143 patients received R-ATG induction. Using instrumental variables regression models, there were no significant differences between IL2-RA versus R-ATG induction in acute rejection episodes (OR=1.49, 95% CI ranges from 0.73 to 3.05, P=0.27), or graft survival (coefficient=0.95, 95% CI: -0.18 to 2.10, P=0.10). Using Cox proportional hazards regression, there was no significant difference in graft survival between either induction therapies (HR=0.90, 95% CI: 0.74 to 1.09, P=0.29). Conclusion This study showed no significant difference in acute rejection episodes or graft survival when using ATG or IL-2RA in 2DR HLA mismatched renal transplant recipients in the current tacrolimus-based maintenance immunosuppression era. Therefore, IL2-RA is a safe induction therapy in this group of patients and non-inferior to R–ATG induction therapy. Moreover, this study also highlights the fact that antigen level 2DR mismatches are not a risk factor for rejection and HLA matching should be based on epitope level rather than antigen level.

Remittance inflow and unemployment in Nigeria

Nigeria is unarguably one of the countries with its citizens widely spread across the globe and the income earned forms a huge chunk of remittance back to Nigeria. The study focuses on what implications remittances may have for unemployment in Nigeria. Remittance is treated as being endogenously determined by the number of migrants, the nominal exchange rate (with the Naira as local currency), the inflation rate and the migrants’ income. Data from 1981 to 2019 is calibrated for structural break points and stationarity under conditions of regimes changes. While the data was found to have been affected by regime changes and stationary in levels, an Instrumental Variable Regression model was estimated and it was found that remittance positively and significantly influence unemployment. However, when remittance is interacted with the dependants in Nigeria, unemployment is observed to fall. The study strongly recommends that fiscal planning should take an account of the inflow of remittances when curbing unemployment. The study further recommends that there is the need to deliberately encourage a rise in the demand for the Naira as this would protect the value of locally produced goods from being eroded by remittances.

Financial Literacy and Gambling Behavior in the United States.

Problem gambling is becoming a growing concern in the United States because of the proliferation of, and state support for, gambling opportunities. The economic cost along with the physical and mental health problems associated with problem gambling make it necessary to study how problem gambling can be reduced. Our study examines whether financial literacy could be a means to reducing gambling frequency in the United States. We use data from the Preference Parameter Study of Osaka University, Japan, and apply instrumental variable probit regression models. The results show that, generally, financial literacy does not have a relationship with gambling frequency, but the relationship is significant in the states where electronic gambling machines (EGMs) are available. The results imply that gamblers are irrational and fail to assess the risks of gambling as well as the probabilities that maximize expected payoffs. It appears that gamblers' psychological gain from gambling outweighs the negative expected utility when there is easy access to gambling. Thus, rationality with regard to gambling decisions does not work unless the easy access to EGMs is controlled. Our results further show that males, older people, people with higher household income, and people who have easy access to gambling are likely to be frequent gamblers.

The Effect of Agricultural Commercialization on Household Food Security among Smallholder Farmers in Zhombe North Rural District of Zimbabwe

Background: Achievement of food security has become one of the major challenges of most developing countries, including Zimbabwe. This study was designed to investigate the effect of agricultural commercialization on household food security. Methods: This study used cross-sectional data for the 2017/18 farming season collected from 165 smallholder farmer households in Zhombe north rural district in Zimbabwe. Instrumental variable regression model was applied for data analysis over the effect of agricultural commercialization. Agricultural commercialization was measured using the crop output market participation share (COMPS). Results: The results revealed that COMPS, household head age, household head gender, and income per capita had a significant positive influence on household food security. Conclusions: The results indicate that agricultural commercialization and socio-economic characteristics have a role to play in defining the household’s food security. Therefore, the findings recommend policies that would promote agricultural commercialization, which improve household food security in turn. The study recommends policymakers to promote agricultural commercialization, since it positively contributed towards household food security. In addition, young-headed, female-headed, and low-income earning households that were identified as the food insecure households.

What are the shocks of climate change on clean energy investment: A diversified exploration

Abstract Climate change may affect energy consumption and thus bring shocks to clean energy investment. By employing instrumental variable quantile regression model, this research investigates the effects of climate change on clean energy investment according to global panel data spanning a long period. In addition to the sample of all countries, we also carry out investigations of sub-samples for OECD and non-OECD countries to analyze whether there exists heterogeneity in development levels. Overall, we provide a diversified exploration on how climate change shocks impact clean energy investment. More specifically, we find that effects of climate change on clean energy investment vary significantly in countries with different levels of clean energy investment; i.e., climate change is likely to promote clean energy investment in countries whose clean investments are greater, and there exists significant heterogeneity in the subsamples. From these results, we offer policy implications such as carrying out differentiated clean energy plans based on the specific environment, implementing policies on clean energy, and copying the practices in some OECD countries that have high levels of clean energy investment.

Decentralization estimators for instrumental variable quantile regression models

The instrumental variable quantile regression (IVQR) model (Chernozhukov and Hansen (2005)) is a popular tool for estimating causal quantile effects with endogenous covariates. However, estimation is complicated by the nonsmoothness and nonconvexity of the IVQR GMM objective function. This paper shows that the IVQR estimation problem can be decomposed into a set of conventional quantile regression subproblems which are convex and can be solved efficiently. This reformulation leads to new identification results and to fast, easy to implement, and tuning‐free estimators that do not require the availability of high‐level “black box” optimization routines.

Financial literacy and gambling behavior in the United States

Problem gambling is becoming a growing concern in the United States because of the proliferation of, and state support for, gambling opportunities. The economic cost along with the physical and mental health problems associated with problem gambling make it necessary to study how problem gambling can be reduced. Our study examines whether financial literacy could be a means to reducing gambling frequency in the United States. We use data from the Preference Parameter Study of Osaka University, Japan, and apply instrumental variable probit regression models. The results show that, generally, financial literacy does not have a relationship with gambling frequency, but the relationship is significant in the states where electronic gambling machines (EGMs) are available. The results imply that gamblers are irrational and fail to assess the risks of gambling as well as the probabilities that maximize expected payoffs. It appears that gamblers’ psychological gain from gambling outweighs the negative expected utility when there is easy access to gambling. Thus, rationality with regard to gambling decisions does not work unless the easy access to EGMs is controlled. Our results further show that males, older people, people with higher household income, and people who have easy access to gambling are likely to be frequent gamblers.

Land transfer and food crop planting decisions in China

ABSTRACT This paper proposes a theoretical model to illustrate the relationship between the scale of land transfer-in and farmers’ crop selection behaviour. We use a semiparametric instrumental variable regression model to test this relationship. Data come from a face-to-face interview survey of 2,677 farmers located in nine provinces of China collected in 2014–2015. Results show that the size of land transfer-in has a U-shaped impact on the probability of planting food crops. To guarantee food crop self-sufficiency, the Chinese government should encourage land transfer, foster the mobility of agricultural labourers, and enhance the capacity of the rural financial market.

Debt Stress and Debt Illusion: The Role of Consumer Credit, Reverse and Standard Mortgages

Objectives. This study examines the relationship of debt stress and reverse mortgage borrowing and compares it to stress from standard mortgages and consumer debt. Debt stress is measured as a self-reported response to the amount of debt. Method. Using a unique national data set of 1,026 homeowners who chose whether to obtain a reverse mortgage in 2010, we estimate the relationship of 2014 levels of debt stress with various types of debt, assets, and income. Using an ordered probit model, we address the endogeneity of our measures of mortgage and consumer debt using an instrumental variables regression model. Results. We find that consumer debt causes more stress per dollar of debt compared to mortgage debt. Reverse mortgages cause a relatively low level of stress per dollar of debt compared with standard mortgage debt. The average treatment effect of originating a reverse mortgage indicates statistically significantly higher probability of reporting no and not very much debt stress. Discussion. Reverse mortgage debt causes a complex stress response. Stress per dollar of debt is lower for reverse than standard mortgages four years after origination. However, reverse mortgages’ loan balance grows over time causing total stress to increase, while stress from a standard mortgage decreases as it is repaid. If an older adult uses reverse mortgage funds to repay consumer debt then total stress is reduced.

Changes in corporate governance: Externally dictated vs voluntarily determined

Using the 2003 SEC regulations (following the Sarbanes–Oxley Act) on board independence as an identification for externally imposed governance changes, I compare its influence on firm performance to the effect of voluntarily conducted adjustments. I use publicly listed US firms between 1998 and 2009. In a triple-difference (dif-in-dif-in-dif) analysis setting, I explicitly interact the dictated change in board independence with the identifiers of the shock and non-compliant firms. Controlling for companies with voluntary changes, firms forced to modify their governance by increasing board independence experience a decrease in ROA, asset turnover, and sales growth. Testing the joint influence of dictated and voluntary adjustments in board independence on performance through a cross-sectional logistic-regression model, and controlling further for potential endogeneity through an instrumental variable (IV) regression model, I obtain consistent results. The findings are robust for other mandated provisions and stronger for bigger changes; small, single-segment firms operating in wholesale, retail, and high-tech industries; and constrained companies with financial distress, high leverage, low cash, high volatility, high growth and R&D expenses.

The Influence of Household Refrigerator Ownership on Diets in Vietnam

Refrigerator ownership accompanies socio-economic development, with the potential to change human diets. Household refrigerator ownership in Vietnam has increased from 13% to 59% between 2004-2014. This study estimates changes in food consumption and diet linkages with household refrigerator ownership in Vietnam, while controlling for socioeconomic variables. We use a two-step instrumental variable regression model on two panels of the Vietnam Household Living Standards Survey covering 2004-2014. Our study finds refrigerator ownership to be significantly associated with decreases in per-capita calorie intake over both periods. Refrigerator ownership may be connected with households substituting lower-nutrient foods with higher ones, with substantial decreases in starchy staple food consumption connected with refrigerator ownership in both panels. For both periods, refrigerator ownership is significantly connected with increased dairy consumption, potentially reflecting the refrigerator increasing a household’s ability to store dairy products.