Propensity Score Matching Estimator Research Articles

Propensity score matching for estimating a marginal hazard ratio.

Propensity score matching is commonly used to draw causal inference from observational survival data. However, its asymptotic properties have yet to be established, and variance estimation is still open to debate. We derive the statistical properties of the propensity score matching estimator of the marginal causal hazard ratio based on matching with replacement and a fixed number of matches. We also propose a double-resampling technique for variance estimation that takes into account the uncertainty due to propensity score estimation prior to matching.

The Effects of Spousal Caregiving on Middle-Age and Older Caregivers’ Health and Well-Being: Evidence From Vietnam

Background: Population aging is escalating globally, intensifying the demand for long-term care (LTC), primarily met by informal caregivers, notably spouses. Evidence from developed countries suggests potential adverse effects on caregivers’ well-being. Yet, research on this topic is scarce in developing nations. We investigate the effect of informal caregiving on older spousal caregivers’ health and well-being in Vietnam, a rapidly aging country with an early stage of LTC system development. Methods: Utilizing the national survey on aging in Vietnam with propensity score matching estimations to mitigate potential endogenous problems of the decision to provide care between caregivers and non-caregivers. Results: Findings showed caregiving increased poor psychological well-being, life dissatisfaction, and functional limitations by 7.3%, 9.7%, and 8.6%, respectively. The caregiving effects are heterogenous by demographic characteristics. Conclusions: We are the first to examine spousal caregiving in Vietnam, highlighting the urgency of addressing its negative impacts and suggesting several potential policy interventions.

Performance Gap between Best-bet System of Rice Intensification and Farmers’ Practices in Zamfara State, Nigeria: An Additive-Multiplicative Translog Stochastic Frontier Profit Function

Inadequate information about the performance of a technology can hinder its uptake. The aim of this paper was to provide a thorough examination of the performance gap between the best bet system of rice intensification (BB-SRI) and farmers’ practices (FP) in Zamfara State, Nigeria. Using a multi-stage sampling technique, a sample of 300 rice farmers were selected from the Bakolori Irrigation Scheme, Zamfara State, among which 40 farmers (13%) used the BB-SRI while the remaining 260 farmers (87%) used FP. The data were analyzed using descriptive statistics, farm budget technique, additive-multiplicative stochastic frontier translog profit function and propensity score matching (PSM) estimator. The findings revealed that yield increased from 5,285 kg/ha with FP to 12,735 kg/ha with BB-SRI, an increase of 141%. Rice production was profitable under both FP and BB-SRI with return on investment (ROI) increasing from 2.64 using FP to 6.66 under BB-SRI, an increase of 152%. There was evidence of profit inefficiency in rice farming since the profit could still be raised by 7% and 21% using FP and BB-SRI, respectively. The yield, ROI and profit efficiency gaps between the BB-SRI and FP ranged from 7,452 kg/ha to 7,510 kg/ha, 4.02 to 4.05, and from 0.08 to 0.09, respectively. We recommend therefore that important investments and training in SRI should be considered as a top priority for the transformation and sustainability of the rice sector in order to ensure its successful promotion, uptake and diffusion across other rice producing States in Nigeria.

Effect of chronic lung diseases on angina pectoris among Indian adults: longitudinal ageing study in India

The study aimed to evaluate the effect of chronic lung diseases, namely chronic obstructive pulmonary diseases (COPD) and asthma, on angina pectoris in individuals aged 45 years and above. Identifying vulnerable subpopulations suffering from COPD and asthma at higher risk of future cardiovascular events using the rose angina questionnaire is imperative for tailored primary and secondary prevention approaches. The present study utilizes the data from the Longitudinal Ageing Study in India, wave 1, conducted during 2017–2018. The sample size included 58,830 individuals aged 45 years and above. Angina was measured based on seven questions from Rose's questionnaires. Descriptive statistics and bivariate analysis were employed to examine the prevalence of angina among individuals with COPD and asthma. Further, multivariable logistic regression and propensity score matching (PSM) methods were used to assess the independent effect of COPD and asthma on angina after controlling the selected background characteristics. We employed PSM in two different models and included various additional factors in model 2, such as smoking, chewing tobacco, alcohol use, ADL, IADL, body mass index, physical activity, high cholesterol, hypertension, diabetes, and chronic heart disease. The current study shows that the prevalence of angina pectoris, COPD and asthma was 6.0%, 2.3% and 4.7%, respectively, among individuals aged 45 years and above in India. The prevalence of angina pectoris was higher among individuals with COPD (9.6% vs. 5.8%) and asthma (9.9% vs. 5.7%) than those without COPD and asthma, respectively. Additionally, angina pectoris was more prevalent among females and rural respondents with COPD (10.8% and 11.0%) and asthma (10.3% and 10.3%) compared to males and urban respondents with COPD (8.0% and 5.7%) and asthma (8.9% and 7.9%). Moreover, in the adjusted model, individuals with COPD (AOR 1.43, 95% CI 1.03 1.98) and asthma (AOR 1.44, 95% CI 1.17 1.77) had nearly 1.5-fold higher odds of having angina pectoris than those without COPD and asthma. The PSM estimates showed that individuals with COPD had 8.4% and 5.0% increased risk of angina pectoris compared to those without COPD in model 1 and model 2, respectively. We observed that, after adjusting to lifestyle, health-related and morbidity factors in model 2, both average treatment effect on untreated (ATU) and average treatment effect (ATE) values decreased by nearly 3.5%. Additionally, the PSM estimates demonstrated that individuals with asthma had a 3.4% and 2.9% increased risk of angina pectoris compared to those without asthma in model 1 and model 2, respectively. The study suggests that COPD and asthma are significantly associated with angina pectoris, and individuals with COPD and asthma have a higher risk of developing angina pectoris. Additionally, angina pectoris was more prevalent among females, rural respondents and adults aged 45–54 with COPD compared to males, urban respondents and those aged 65 and above, respectively, with COPD. Moreover, the findings of our study underscore the targeted primary and secondary interventions and team-based care approach among individuals with COPD and asthma to reduce the risk of CVD events in future.

Climate smart agriculture impact on food and nutrition security in Ethiopia

IntroductionThe objective of this study was to analyze how innovations in Climate-smart agriculture (CSA) contribute to improving the food and nutrition security of smallholder households, thereby promoting sustainable food systems.MethodsA cross-sectional household survey was conducted among a multi-stage sample of 424 smallholder farmers drawn from five different agroecosystems. To examine households’ food and nutrition security, we used food consumption score (FCS) and modified household dietary diversity score (HDDS) in propensity score matching (PSM) and endogenous switching regression (ESR) estimation models.ResultsPSM results showed that crop residue management, compost, and agroforestry have a significant effect on improving households’ food and nutrition security by 21.3, 13.6, and 16.6%, respectively, whereas Soil and water conservation (SWC) has reduced adopters’ food security by 12.9%. However, the conditional average treatment effect, or ESR result, reveals that households’ food and nutrition security has improved as a result of the adoption of crop residue management, compost, SWC, and agroforestry.DiscussionAlthough the impact of crop residue management, compost, and agroforestry is positive, the effect of SWC on household food security has been inconclusive. Hence, it is important to upscale the adoption of multiple CSA innovations to improve smallholder household’s food security in the face of climate change.

The effect of weather index insurance on social capital: Evidence from rural Ethiopia

AbstractWe study the effect of weather index insurance (WII) uptake on social capital. We measure individual social capital using experimental and survey‐based measures and relate it to the actual purchase of WII. We use propensity score matching (PSM) and an instrumental variable (IV) to address endogeneity concerns. Our descriptive and PSM estimates show that WII uptake negatively and significantly affects social capital. We find that insured households contribute less to the public good than uninsured households. We also find qualitatively and quantitatively similar results using our IV approach, yet insufficient power renders most of our IV estimates insignificant. We however report robust significant effects of instrumented WII uptake on sociopsychological outcomes—WII uptake increases perceptions of self‐sufficiency and free‐riding behavior: these are potential channels through which a negative effect on social capital comes about. Although far from conclusive, our paper provides several pieces of evidence that suggest WII uptake negatively affects social capital.

International labour migration, farmland fallowing, livelihood diversification and technology adoption in Nepal

This article investigates the effect of temporary international labour migration on farmland fallowing, adoption of agricultural intensification technologies and livelihood diversification. Using nationally representative data, combined with empirical methods that allow causal inference, the authors find that households with international migrants are over 50 per cent more likely (based on propensity score matching estimates) to have fallow land than those without. Temporary international migration promotes the adoption of some agricultural intensification technologies and causes rural households to diversify their livelihoods. Land fallowing may increase food insecurity, while agricultural intensification may improve it, for an uncertain net effect.

Goodhart’s law in China: Bank branching regulation and window dressing

After the removal of geographic restrictions on branching in 2006, China’s city commercial banks (CCBs) can apply for permission to branch outside their province. This paper shows that CCBs report a higher provision coverage ratio (PCR) before filing an application, thereby making the bank look safer to regulators. Our finding is robust to controlling for possible endogeneity of the branching application decision by employing propensity score matching estimators, and it is confirmed when we consider a quasi-natural experiment of deregulation reversal. Tests of the dynamic effects show evidence of reversals in PCR adjustment after applications. Higher PCR before branching applications cannot be explained by alternative rationales for manipulating loan loss reserves such as fundamental provisions, earnings management, capital management, and market signaling. Window dressers receive more supervisory penalties after filing applications relative to other branching banks. Our finding of window dressing in response to bank branching regulation confirms Goodhart’s insight that when a regulatory metric becomes a target, it ceases to be a good measure.

The shift in ideal family size: Examining the impact of marriage and parenthood

Motivated by role theory and the concept of the “self-fulfilling prophecy”, this study examines how entries into marriage and parenthood affect the fertility ideals of men and women in China. To address this question, we apply propensity score matching estimators to data from the China Family Panel Studies (CFPS) within a counterfactual framework. We find that marriage and parenthood have a positive effect on fertility ideals, and this result remains robust when we apply difference-in-differences matching estimates. In terms of gender asymmetry in marriage, men and women respond differently to marriage and parenthood. Entry into marriage has a positive effect on men's fertility ideals, while parenthood influences women's ideals. Beyond the average treatment effect, we further examine heterogeneous treatment effects as a function of estimated propensity scores. The study reveals that the life transition treatment effect is the highest among those with a moderate propensity for entering into marriage and parenthood.

Can internet adoption promote the productivity of firms in China: evidence from a longitudinal study

PurposeThis paper aims to conduct research to examine the impact of Internet adoption on the productivity of firms in non-urban China.Design/methodology/approachThe study investigates the impact of Internet adoption on firms' productivity in non-urban China. More specifically, the authors conduct a comprehensive and rigorous study while addressing concerns related to firm-level endogeneity by utilizing firm-level panel data. Information on firms in non-urban areas is collected from China's Annual Surveys of Industrial Firm data. For robustness, the authors implement the instrumental variables approach and propensity score matching estimations to strengthen the evidence for suggestive causal inference. Furthermore, the authors also examine the mechanisms and group heterogeneity.FindingsEvidence indicate that the adoption of Internet technology positively impacts the total factor productivity (TFP) of firms in non-urban areas. According to the heterogeneity analysis, the marginal effect of Internet adoption is more significant and pronounced for labor-intensive, private and small-scale manufacturing firms. Moreover, additional evidence suggest that Internet adoption is beneficial for non-urban firms in expanding their business and enlarging their market. It has also been found that the positive effect of Internet adoption on firms' TFP is amplified by expanding public infrastructure.Originality/valueThe current study supports that the informatization strategy benefits non-urban firms and promotes rural revitalization. The findings suggest the possibility of firms borrowing market size from the closest cities and supporting the ongoing policies of investing in broadband infrastructure to narrow the urban-rural digital gap in China.

Firm investment steering through state-level policy uncertainty

Does state-level policy uncertainty (SPU) stifle firm investment? We document fresh evidence of a strong negative association between SPU and firms’ capital spending. This finding remains robust after excluding states with overrepresentation, through alternative measures of investment and SPU, as well as during low- and high-SPU periods. Furthermore, this effect is stronger for firms facing low market competition. Interestingly, however, firms intensifying lobbying efforts can maintain higher investments during policy turmoil. Additional analyses propose that the mitigating role of lobbying may be explained by the lobbying firms’ greater access to debt financing. Using matched samples based on propensity score matching (PSM) estimation ensures the empirical resilience of these findings.

How sensitive are matching estimates of active labor market policy effects to typically unobserved confounders?

Using a rich and unique combined administrative-survey dataset, this paper explores how sensitive propensity score (PS) matching estimates of Active Labor Market Policies (ALMPs) based on the selection-on-observables assumption are to typically unobserved covariates. Using a sample of German unemployed welfare recipients, the analysis shows that typically unobserved factors such job search behavior, concessions willing to make for a job as well as (mental) health are in fact relevant confounders. However, results also show that matching on the PS using only typically observed covariates reduces imbalance in terms of typically unobserved covariates by about 46 percent in this setting. In line with this finding, the inclusion of typically unobserved covariates yields very similar estimates to estimates based on a standard specification. Hence, a standard matching approach based on rich and high quality administrative data appears to be sufficient to obtain estimates that are rather robust to unobserved confounding.

Adoption of Korra tef ( Eragrostis tef ) and its impact on farm households welfare: A propensity score matching estimation in North-Western Ethiopia

Adoption of <i>Korra tef</i> ( <i>Eragrostis tef</i> ) and its impact on farm households welfare: A propensity score matching estimation in North-Western Ethiopia

Does Community-Based Adaptation Enhance Social Capital? Evidence from Senegal and Mali

Climate change and extreme climate shocks pose a significant threat to resource-dependent rural communities. Successfully supporting households to anticipate and adapt to climate variability and shocks, as well as build long term climate resilience, is essential to facing these changes. Given the importance of social capital in facilitating collective action and adaptation, the development community has focused on bottom-up, participatory adaptation projects. This article explores the social capital impacts of a pilot community-based adaptation project in Senegal and Mali that aims to encourage inclusive decision making around public goods investments. The analysis uses both difference-in-differences and propensity score matching estimates to evaluate whether households that participated in the project realized enhanced social capital, as measured through participation in community development, and acts of reciprocity and community support. The findings indicate that engaging in the participatory process through the project increased the likelihood of future collective action and providing help to other community members in Mali.

Energy security and economic stability: The role of inflation and war

This paper investigates the impact of energy security risk (ESR) on economic stability. Using multiple global datasets, we provide empirical evidence from a panel of 68 countries spanning over a period from 1980 to 2021. Our results indicate that high ESR reduces GDP growth rate (GDPG) from a global perspective. In robustness tests, this effect remains valid across several specifications based on non-U.S. samples, national income-level, alternative measure of economic stability, and a set of endogeneity tests based on propensity score matching estimation. Countries with pre-existing low GDPG mainly suffer from heightened energy insecurity. Further tests reveal that numerous country-specific institutional quality estimates moderate the effect. Finally, the damaging impact of ESR worsens during years of high inflation, geopolitical risk and acts, as well as of escalated war threats. We encourage international collaborations to develop a more sustainable energy system, which enhances the security of energy supply and the stability of economy.

FINANCIAL INCLUSION OF HOUSEHOLDS IN INDONESIA

There is extensive attention on financial inclusion and its benefits, recently strategy to stimulate financial inclusion has focused on innovation and technology penetration. This paper analyzes the effect of financial inclusion on a household’s income and the role of cellphone and internet access in household financial inclusion in Indonesia. We develop the financial deprivation of each household to calculate financial exclusion. The Indonesia Family Life Survey year 2014 data are utilized in this paper. Our methodology is Ordinary Least Square (OLS) to examine how the role of financial institutions affects a household’s income. In the second model, we use Probit estimation to determine the likelihood of household financial deprivation due to cellphone and internet access. We also check the robustness of previous models using Propensity Score Matching (PSM) estimation. Our estimation results found that financial exclusion has a highly significant impact on a specific group of households. With middle-level income households, the financial exclusion would deprive income of almost 80 percent. This finding explains that middle welfare households benefited more from financial inclusion than households in the poorest and the wealthiest group. Another result found that cellphone and mobile banking significantly impact decreasing financial deprivation, respectively.

Return to intended oncologic therapy after colectomy for stage III colon adenocarcinoma: Does surgical approach matter?

Return to intended oncologic treatment (RIOT) is an important paradigm for surgically resected cancers requiring multimodal treatment. Benefits of minimally invasive colectomy (MIC) may allow earlier initiation of adjuvant chemotherapy (ACT) and have associated survival benefits. We sought to determine if operative approach affects RIOT timing in resected stage III colon cancer. NCDB identified pathological stage III colon adenocarcinoma patients who underwent resection and received ACT. Propensity score matching and kernel density estimation compared operative approaches and conversion impact on intervals to RIOT. A total of 15,132 open colectomies (OC) versus 14,107 MIC were included. MIC patients had two-days shorter median length of stay (LOS) (4 vs. 6 days; p < 0.001), one-week shorter median time to RIOT (6 vs. 7 weeks; p = 0.015) comparing 12,867 matched pairs. There was no difference in time interval to RIOT between the LC versus RC, converted MIC vs. OC groups. MIC was a favourable predictor of earlier RIOT (HR 1.14 [1.07-1.22]; p < 0.001). MIC in stage III colon cancer is associated with a shorter time to RIOT when compared to OC. Since timely initiation of ACT may influence cancer outcome, MIC may be oncologically preferable. Prospective studies are needed to assess RIOT and survival outcomes in stage III colon cancer.

Effects of maternity on labor outcomes and employment quality for women in Chile

ABSTRACT Parenthood have a significant impact on engagement in the labor force. Parents may decide to look for jobs with specific characteristics, and in some cases, they may decide to not participate in the labor market. Using survey data from Chile, we use a propensity score matching estimation to compare various labor outcomes of women and men with and without children. We found that maternity has an impact on several dimensions of labor engagement. To study how these multidimensional effects are compounded at the individual level, we use a quality of employment (QoE) index to estimate the drop in job quality post-childbirth. We found a drop in quality only for mothers. Moreover, for less educated women, the drop in QoE is equivalent to no longer being hired under contract. For single mothers, the drop is equivalent to a reduction in wages that would make the family fall below the poverty line.

Evaluation of the Causal Association of Risk Factors for Death From COVID-19 Patients Admitted to Golestan Hospitals by Propensity Score Estimation Method

Background and Purpose: Evaluating the causal association effect of risk factors is essential in estimating the mortality rate among COVID-19 patients. Much research has been conducted to assess the impact of COVID-19 on death in various countries worldwide. However, few studies have addressed the effect of causal association of risk factors. This study aims to address this gap by estimating the impact of COVID-19 on death by evaluating the causal association with the risk factors. Materials and Methods: The research population included all inpatients with initial COVID-19 symptoms, confirmed by their PCR test results. They were admitted to hospitals affiliated with Golestan University of Medical Science, Golestan, Iran, in 2020. We employed the propensity score method, an effective statistical technique for evaluating the causal association effect of risk factors in observational studies. We also used the student and chi-squared tests to compare the differences between the two study groups. Results: We used the propensity score matching estimation approach and logistic regression analysis for comparison. Of 6379 inpatients, 5581 (87.5%) were discharged or recovered, and 798 (12.5%) died. The causal association between treatment results (discharged vs died) and variables of PCR test, SpO2, gender, age, and hospitalization duration in ICU were statistically significant. Conclusion: The propensity score matching estimation method revealed a high risk of death in patients with PCR+ test diagnosis. Specifically, using this approach, the above-measured risk factors increased the chance of death in patients with PCR+ to 72%. However, the traditional multiple logistic regression model estimated the risk of death at 46%, suggesting potential underestimation. This disparity might be due to better control of the effect of the above-measured risk factors by the propensity score matching. Therefore, the former estimating approach is more effective in assessing the impact of COVID-19 on death.

Place‐based subsidies and employment growth in rural America: Evidence from the broadband initiatives programme

AbstractThis paper studies the labour market effects of the Broadband Initiatives Program (BIP), a programme authorized by the American Recovery and Reinvestment Act of 2009 to promote broadband deployment, mainly in rural areas. The BIP is one of the largest USDA broadband programmes implemented to date, providing more than $3.4 billion in grants and loans in FY 2010. We investigate the impacts of BIP investments on employment outcomes in BIP‐recipient Census tracts compared to similar tracts outside of BIP project service areas between the inception of the programme in 2010 and 2019. We use a quasi‐experimental research design that combines difference‐in‐difference regression with propensity score matching estimation to identify the causal effect of the BIP investments on employment outcomes. We find that the BIP investments had a positive overall effect on employment growth that increased over time. The subsidized investments had a greater effect on employment in startups than in incumbent businesses, in the goods‐producing sector and the information and communications technology sector than other sectors, and a greater effect in micropolitan census tracts than tracts located in metropolitan areas or in small town and remote rural locations.