No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Estimation of Average Treatment Effects Based on Propensity Scores
Abstract
In this paper, we give a short overview of some propensity score matching estimators suggested in the evaluation literature, and we provide a set of Stata programs,which we illustrate using the National Supported Work (NSW) demonstration widely known in labor economics. Copyright 2002 by Stata Corporation.
... Seguendo (Becker & Ichino, 2002) e (Caliendo & Kopeinig, 2008), se per ogni impresa siamo in grado di osservare X, e se inoltre possiamo effettuare la seguente assunzione 37 : ...
... Se tale ipotesi è rispettata, infatti, "le osservazioni con lo stesso propensity score devono avere la stessa distribuzione di caratteristiche osservabili (e non osservabili), indipendentemente dallo stato del trattamento. In altre parole, per un dato propensity score, l'esposizione al trattamento è casuale, e di conseguenza, le unità trattate e di controllo dovrebbero essere, in media, identiche all'osservazione" (Becker & Ichino, 2002). ...
... Nel secondo stadio si effettua l'abbinamento statistico (matching) per ogni individuo trattato con uno o più controlli, sulla base del valore assunto dal propensity score, e si calcola l'effetto medio del trattamento sui trattati (ATT) secondo l'equazione (8). La tecnica utilizzata è quella del propensity score di tipo kernel (Becker & Ichino, 2002). 38 Questa condizione è chiamata Unconfoundedness dato il propensity score o CIA dato il propensity score. ...
... The estimated propensity scores portray the likelihood of a household being an RD-beneficiary while taking into consideration both RD-beneficiaries and non-beneficiaries, with reference to a set of observable variables (Becker & Ichino, 2002;Li, 2013). The scores are estimated using an appropriate algorithm for matching RD-beneficiaries with non-beneficiaries. ...
... The scores are estimated using an appropriate algorithm for matching RD-beneficiaries with non-beneficiaries. Any RD-beneficiaries possessing a propensity score lying between specified minima and maxima are eliminated from the sample (Becker & Ichino, 2002;Li, 2013). We further employ the balancing property test (Becker & Ichino, 2002;Li, 2013), to ensure that households from distributions of selected observable covariates had equal likelihood of selection. ...
... Any RD-beneficiaries possessing a propensity score lying between specified minima and maxima are eliminated from the sample (Becker & Ichino, 2002;Li, 2013). We further employ the balancing property test (Becker & Ichino, 2002;Li, 2013), to ensure that households from distributions of selected observable covariates had equal likelihood of selection. The choice of variables for estimation of the propensity score for matching must not be influenced by participation in RD's interventions, but may influence the social impact indicator variables selected. ...
In developing countries the most marginalised and disadvantaged people are in rural, remote and regional areas and social enterprises in these areas-rural social enterprises (RSEs)-have been identified as key development actors in this context. However, their impacts are rarely rigorously measured. Our study fills this gap by measuring an RSE's social impact in a developing country. A smallholder farmers' survey (n = 1021) is utilised in a propensity score-based method which allowed us to generate counterfactual and estimate outcomes between members and non-members of an RSE. This method was complemented by a stakeholder focus group discussion. Predictors of participation and social impacts of the RSE are identified besides an evaluation of its interventions. Results generate implications for social enterprise practitioners, supporters and policymakers interested in applying RSEs as local and regional development actors as well as researchers involved in social impact measurement.
... The better the health, the lower the relative poverty depth. In this paper, heterogeneity analysis is conducted for the eastern area, central area and western area in China respectively, and the empirical results are shown in columns (3) to (5). The poverty reduction effect of digital financial inclusion has regional heterogeneity. ...
... In terms of control variables, age, The mechanism of encouraging entrepreneurship in digital financial inclusion for poverty reduction is that the development of digital financial inclusion not only improves the local economy and provides opportunities for entrepreneurship, but also provides credit funds for the entrepreneurship of rural families, so that they can put capital into production with labor forces. This paper takes advantage of the answer of FM2 "How many private activities are you engaged in" in the CFPS questionnaire and uses the two-way fixed effects model to test the entrepreneurship mechanism, as shown in formula (5). entrepreneur jt is the number of private enterprises owned by the rural family j in the year of t. entrepreneur jt = α 0 + α 1 DIF it * Internet jt + α 2 X ijt + ∅ ij + φ t + u ijt ...
Many policy measures are included in China's 2020 poverty reduction schedule by 2020. The advantages of digitization have been used to break through the realistic difficulties of traditional financial inclusion and effectively solve the problem of the last mile of financial inclusion. Using a generalized difference-indifferences method,this paper selects the digital financial inclusion indicator and China family panel studies, and analyzes the impact of the digital financial inclusion indicator on rural family poverty reduction. The following conclusions have been drawn through empirical study: First, the development of digital financial inclusion aids in the reduction of poverty among rural families. Second, there is regional heterogeneity in the impact of digital financial inclusion on poverty reduction among rural families. Third, the mechanisms of digital financial inclusion to reduce poverty in rural families include increasing wage income, easing credit constraints and encouraging entrepreneurship.
... If the neighbourhood dimension is confgured to be very tiny, likely, certain treated units will not be matched because the neighbourhood lacks a control unit. Conversely, the smaller the neighbourhood, the higher the quality of the matches [24]. One issue with calliper matching is determining which option for the tolerance level is the most realistic. ...
... Tis is another matching approach in which all treated units are matched with a weighted average of all controls, with weights inversely proportional to the distance between the treatment and control propensity scores [24]. Kernel weights each comparison group member's input, giving greater weight to those comparators that provide a better match. ...
Irrigation development, particularly small-scale irrigation, is one of the most important projects for improving agricultural productivity in a country’s rural communities. The extent to which small-scale irrigation has improved household livelihoods in Ethiopia’s rural areas is not widely recognized. As a result, research on the influence of small-scale irrigation on farmers’ livelihoods in the Legehida district will be sought. The study took a “with” and “without” strategy, comparing farmers who used irrigation against those who did not. For analysis, both quantitative and qualitative data were employed. The survey’s respondents were chosen using a random sample approach from both irrigation users and nonuser households. Quantitative data for the study were collected from randomly selected 241 farm households, of which 113 were users and 128 were nonusers, using a semistructured questionnaire. Accordingly, the propensity score matching model was employed to examine the impacts of small-scale irrigation on farmers’ livelihoods. The logit model result indicates that cultivated land size, off-farm income, education level, family size, dependency ratio, total livestock unit, and distance to the nearest agricultural extension office/FTC are determinant factors in determining whether to practice irrigation when other factors remain constant. The impact of irrigation on a household’s income and food security (in terms of daily calorie intake) was evaluated using a propensity score matching model. The result shows that a positive and significant impact on farmers who use small-scale irrigation has increased the daily calorie intake and annual income of households by 244.162 kilocalories and 5234.258 ETB, respectively, as compared to nonirrigation users. This shows that households that participate in small-scale irrigation activities have a higher annual income and food security status than comparable groups. In general, the study recommends that to reduce food insecurity and the socioeconomic problems of rural households, irrigation farming is one of the viable solutions; therefore, the government and nongovernmental organizations should extensively focus on the enhancement of small-scale irrigation infrastructure, policies, strategies, and extension services to increase productivity, income, and livelihood improvement in rural households.
... PSM allows us to solve selection bias issues, get non-biased estimates of the treatment effect and compare the factual and counterfactual to estimate the outcome of a programme. Rosenbaum and Rubin [22] first introduced the PSM approach, with Becker and Ichino [3] and Caliendo and Kopeinig [5] later on providing some improvements and more applications. ...
... who. int/ en/ health-topics/ disea se-preve ntion/ nutri tion/a-healt hy-lifes tyle.3 Provided by INRAB. ...
Background
This study aims to assess the effect of a personalised nutrition intervention on nutrient intake in rural Benin as a tool to tackle the double burden of malnutrition. The personalised recommender system uses information from the eating behaviour of all household members combined to provide tailored advice on adjusting the consumption of up to three food items to tackle malnutrition and obesity in the household. Many developing countries nowadays experience the double burden of malnutrition, the coexistence of undernutrition alongside overweight and obesity, as well as other related non-communicable diseases. Personalised nutrition was very effective in European studies in improving nutrition and tackling obesity, which is why this study aims to translate personalised nutrition to a developing country context.
Results
A study was conducted in rural areas of Benin where 720 households were randomly selected. Due to high attrition, we used propensity score matching and looked into average treatment effects. We found that the recommendation to eat less carbohydrates resulted in a reduction in carbohydrate consumption with a significant effect for both the average treatment effect (ATE) of the whole population as well as the average treatment effect of the treated (ATET). We found that households that received the treatment to consume less food items with a high carbohydrate intake have followed this advice and have consumed on average lower levels of carbohydrates than their control counterparts.
Conclusions
Rising obesity is a worldwide problem that poses a severe challenge for policymakers. Especially in developing countries the change from too little, to too much is seamless. For the increasing obese population, the recommender system could be a useful tool. The idea of personalised nutrition has the potential to be one of the necessary steps in the ongoing battle against obesity and unhealthy diets. The personalised application-based recommender system used in this study has the ability to be a strong and effective tool for policymakers in the ongoing battle of food security vs. obesity in Benin and other countries. We propose that future research focuses more on personalised nutrition in the context of a developing country.
... To determine the propensity score (E) for the likelihood of belonging to a cooperative, the researchers utilised a probit model that incorporates a number of conditioning factors (X) that are likely to explain membership behaviour and the not-random distribution of participation in the sampled population. The average treatment effect on the treated (ATT) for the treated (members or participants) population is the primary parameter of interest in PSM (Becker & Ichino 2002). Thus, the following is provided as the PSM nonparametric model (Equation 1): ...
... The ATT are calculated using the findings of the paired comparison. According to Becker and Ichino (2002), the income disparity between the group of households engaging in the cooperative's activities and the group of households not participating is the ATT. The analytical findings for three matching techniques are favourable and significant at 5%. ...
It is commonly acknowledged that cooperatives play important social functions.
that raise the standard of living for their members, particularly those who originate from rural, and low-income backgrounds.This article aims to measure the impacts of cooperatives membership on household income taking Zanzibar as a case study.
The data used were directly collected from 217 cooperative members and 83 noncooperative members. Descriptive statistics were used to analyze the demographic characteristics of the respondents. The probit model and propensity score matching (PSM) was used to analyze the impacts of cooperative membership on household income. The probit model findings show that there are four statistically significant factors affecting cooperatives membership, including gender, educational level, land ownership, and access to credit. In addition, PSM findings reveal that there is a disparity in income level between cooperative members and non-members. On average, cooperative members are able to generate more income than non-cooperative members by 28% per year. The study concludes that, in order to expand the observed benefits to the population, cooperative growth needs proper backing. Because poverty has many different dimensions, it’s crucial to expand the organizations that help the poor while also utilising other support services to reduce it.The article serves as first empirical evidence to be conducted in Zanzibar,
Tanzania. The findings will facilitate the amendment of the cooperative context, including tax reduction, extending loans and grants, and other favourable working conditions necessary for supporting the development of cooperative society.
Keywords: cooperatives; household income; probit model; propensity score matching; Zanzibar.
... Does mobile-app adoption improve paddy yield, or is the positive correlation between the two because households with better yield are richer enough to use a smartphone, internet connectivity, and adopt mobile-app? In other words, we are interested in identifying the correlation between mobile-app adoption and paddy yield and the underlying causation (Becker & Ichino, 2002). ...
... The results obtained in Table 7 depend on the postulation of conditional independence and confoundedness. If any unobserved independent variable is present that can affect both mobileapp adoption and outcome variables, then the chance of unobserved heterogeneity appears, which can alter the influence significance (Becker & Ichino, 2002;Rosenbaum & Rubin, 1983). In non-experimental studies, it is difficult to determine the magnitude of such hidden bias due to the unavailability of a relevant measurement tool. ...
The study assesses the resource use efficiency of smallholder paddy farmers with/without considering undesirable outputs through the mobile-based application. Further, the study performs an impact assessment of digital recommendations on farmers' paddy yield improvement. A mobile app-based questionnaire was used to collect data from 153 paddy farmers in eastern India. The study employed Data Envelopment Analysis (DEA) to identify the farmers' resource use efficiency with/without undesirable output. We found lower farm eco-efficiency scores with undesirable output in the model compared to the case of not considering the undesirable output analysis. Results also showed that farmers are over-utilizing fertilizers, farming machinery, and labor in farming, which needs to be reduced to the recommended optimal level. Finally, using the Propensity Score Matching (PSM), we observed that the farmers achieved better paddy yield, i.e., an additional 0.6t/ha paddy, due to the adaptation of mobile-based recommendations. Subsequently, we used probit modeling to estimate the critical factors for adopting mobile-based services. Results show that farmers’ education level, farm experience, social capital, and market information play a significant role in mobile-app-based recommendation adoption. This study supports that farmers need to be suggested to use digital advisory services, and state/central policies may be aligned towards strengthening farmers' capacities for applying digital services in the farming system.
... Firstly, we converted the raw loneliness scores into binary values and calculated the odds ratios (ORs) to evaluate the association between baseline loneliness and the onset of postpartum depression and MIBD six months later. To balance the differences in background information, including sociodemographics, health, pregnancy, childbirth, and COVID-19-related information, we employed propensity score matching (Becker & Ichino 2002). Secondly, we analyzed the raw loneliness scores as a continuous variable and graphically illustrated its potential nonlinear association with postpartum depression and MIBD employing restricted cubic spline logistic regression (Harrell 2015). ...
... Our aim was to explore the relationship between baseline loneliness and postpartum depression and MIBD at T2, quantified through ORs. For this, we employed two methodologies: stratification matching and nearest neighbor matching, as outlined by Becker & Ichino (2002). ...
Purpose
The COVID-19 pandemic has intensified feelings of loneliness, especially among postpartum women. This nationwide Japanese longitudinal study assessed the impact of such feelings on depressive symptoms and mother-to-infant bonding difficulties (MIBD), two pivotal determinants of maternal and infant well-being.
Methods
Starting with a baseline survey conducted between July and August 2021, we tracked 1254 postpartum Japanese women who initially reported minimal depressive symptoms (i.e., Edinburgh Postnatal Depression Scale < 9) and MIBD (i.e., Mother-to-Infant Bonding Scale < 5), over a follow-up period of approximately 6 months. Baseline loneliness was evaluated with the UCLA Loneliness Scale Short-Form (UCLA-LS3-SF3).
Results
Forty-nine percent of the sample reported the presence of baseline feelings of loneliness. After propensity score matching on sociodemographics and various pregnancy, childbirth, and COVID-19–related aspects, baseline loneliness was associated with increased risks of later depressive symptoms but not MIBD. Using restricted cubic spline logistic regression and considering loneliness as a continuous variable, we found a positive increasing quadratic relationship with depressive symptoms. As loneliness increased, so did the risk of later depressive symptoms. However, there was no significant association between loneliness and MIBD. These results were confirmed through a sensitivity analysis using inverse probability weighting to address attrition bias.
Conclusion
Feelings of postpartum loneliness are associated with future risks of depressive symptoms. The data suggests that addressing loneliness in postpartum women early is crucial to safeguarding their well-being and that of their infants.
... The most common treatment effects in the evaluation literature include the average treatment effect (ATE) which captures the treatment effect for the whole sample, the average treatment effect on the treated (ATT) or the participation effect, and the average treatment effect on the untreated (ATU). Notably, Becker and Ichino (2002) indicate that the parameter of interest in the estimation of the propensity score is the average treatment effect on the treated (ATT) since it suffices in providing a specific estimate of the impact of participation on non-farm work. ...
In African countries like Tanzania, mobile financial services (MFS) is one of the empowerment strategies to enhance economic well-being through the provision of necessary tools and resources to enhance entrepreneurship and business development. However, existing studies have paid less attention to MFS agents’ welfare impact, mostly on youth in this sector. Using a survey of 310 youths from four selected Local Government Authorities in Tanzania, this study explores the factors that drive the decision of youths to participate in livelihood-based activities such as MFS and the effect of their participation on their welfare. The findings from the propensity score matching (PSM) modeling technique revealed that socio-economic-based factors (education, marital status, and experience) and institutional-based factors (business ownership status, deployment of risk management initiatives, and perception regarding the cost of running a business) influence youth participation in MFS-based activities and the magnitude of impact in their welfare. The study’s results highlight the influence of both socio-economic and institutional variables on the participation of youths in MFS-based activities. The findings suggest that it would be necessary to promote development programs that are geared towards enhancing the capacities of youth with regard to MFS businesses to enhance their welfare. These include business-related training, awareness, and provision of financial resources via enhancing access to training on savings and credit.
... To estimate the difference in sales between two groups of farmers, two measurement algorithms of PSM are used: the Robust and Kernel Matching. The reason for using multiple methods is to confirm robustness of the results [31]. Each treated individual is paired with a control group individual who has the closest propensity score when utilizing the nearest neighbor approach [18]. ...
This study investigated the nexus among the profitability, sales, and willingness to pay (WTP) more for weather index-based agricultural insurance premiums of flower farmers. In addition, the effect of sociodemographic and farm characteristics of flower farmers on their WTP more for insurance premiums was also estimated. A total of 160 flower farmers were selected from Bangladesh using the purposive random sampling technique. Propensity score matching technique was employed to identify the sales difference depending on WTP for insurance premiums while the profitability differences of flower farmers were assessed from different points of view. A binary logistic regression model was used to estimate the effect of sociodemographic and farm characteristics of flower farmers on their WTP more for insurance premiums, while a Likert scale was used to identify the major problems faced by flower farmers. Flower growers are willing to pay a higher premium for insurance when their sales decline. Farmers with lower profitability show a greater WTP higher insurance premiums, whereas those with relatively higher profitability are less inclined to do so. Farmers’ WTP more for insurance premiums decreases with age, education, and farm area, while farmers’ WTP more for insurance premiums increases with experience, training, earning members, marigold farming, and consciousness about natural calamities, pests, and diseases. The most significant problems faced in flower production are high input costs, demand fluctuation, pest and disease attacks, price fluctuation, and loss of production. Thus, the introduction of crop insurance in flower farming may increase profitability and reduce the exposure to risks involved in flower farming. The involvement of younger and more trained farmers in flower farming will increase their WTP more for insurance premiums.
... We used different methods to match similar contract and non-contract farmers. We apply nearest neighbor matching (NNM), radius matching (RM), and kernel-based matching (KBM) as the main ATT estimation methods (Becker and Ichino, 2002;Caliendo and Kopeinig, 2008). In the NNM, each treated farmer is matched with a comparable farmer that has the closest propensity score. ...
Foreign investment can facilitate modernization of domestic food chains in emerging economies through increased use of vertical coordination. The paper sheds lights on how beer multinational investments in African food chains affect smallholder market participation and value chain development. In particular, we examine the impact of contract farming arrangements among malt barley producers in Ethiopia. On the basis of cross-sectional survey data, we employ OLS regression and propensity score matching techniques to analyze the impact of contracting on a number of performance indicators. We find that contracting has positive impact on malt barley production, intensification, commercialization, quality improvement and farm-gate prices, ultimately resulting in increased net income and spillover into the productivity of other crops.
... The study also incorporates a radius matching (RM) approach to retrieve the risk associated with NN matching. In such a matching approach, treated subjects are matched only with untreated subjects whose PS falls within the predefined radius-the smaller the radius, the better the matching quality (Becker and Ichino, 2002). Finally, the kernel matching (KM) method (non-parametric) somehow differs from these approaches. ...
Purpose
The present study aims to identify the crucial determinants of the adoption of zero-tillage (ZT) technology in maize production in peninsular India. The study also measures the impact of ZT adoption on maize yield, income generation, and the expenses associated with different agricultural operations.
Methodology
The study used multi-stage stratified random sampling and conducted a face-to-face questionnaire survey to collect primary data from 1,189 maize farmers. Initially, the study employed probit regression analysis to identify the ZT adoption determinants. Subsequently, using the Propensity Score Matching (PSM) approach, the study measures the impact of ZT adoption over conventional tillage in terms of yield, income, and cost management. Finally, the Endogenous Switch Regression (ESR) method was implemented to mitigate unobserved heterogeneity and sample selection bias. Additionally, ESR assessed the robustness of PSM results.
Findings
The probit model identifies that variables like education, institutional credit adoption, crop insurance, visit of extension agent, landholding size, and prior experience of new technology adoption positively influence ZT adoption. The PSM and ESR approach results suggest that ZT adoption positively impacts farmers’ yield and net income while reducing cultivation costs and labor use. Results show that ZT adoption decreases the cost of land preparation, weed, pest management, and harvesting by INR 2708 acre⁻¹, INR 167 acre⁻¹, and INR 649 acre⁻¹, respectively, thereby decreasing the overall cultivation cost by INR 8376 acre⁻¹. However, seed and seed treatment costs and irrigation costs improve by INR 108 acre⁻¹ and 176 acre⁻¹ due to the adoption of ZT in maize cultivation. Moreover, ZT improves maize yield by 2.53 quintal acre⁻¹ and minimises 9.56 person-days acre⁻¹. ESR results suggest that the net return from maize cultivation is 26.1% higher for ZT adopters than conventional farmers. Additionally, ZT adopters can save 8.23 man-days acre⁻¹, providing additional monetary benefits of INR 3259 acre⁻¹ compared to ZT non-adopters.
Practical implications
The study findings may support policymakers in designing suitable agricultural policies to improve technology adoption and motivate small and marginal maize farmers for sustainable production.
... 29 In all cases, we impose the common support condition. This improves the quality of the matches used to estimate the ATTs (Becker & Ichino, 2002). Only a small number of observations do not satisfy this condition. ...
We are the first to study how the resources freed up when a child, child-in-law, or grandchild moves out of a household are reallocated, taking into account the age of the leaver. Using the 2011 and 2013 waves of the China Health and Retirement Longitudinal Study, we document that, on average, the remaining household members save part of the resources freed up by the leaver and consume another part. Differentiating the leavers by age, we find that after the departure of a member of the younger generation aged 0–24, the remaining household members save the resources freed up by the leaver. However, if the leaver is above 24, they spend the freed-up resources. Our results are robust to the use of different specifications, estimation methods, and consumption aggregates. Finally, we observe that remittances directed toward non-resident offspring do not increase after the departure of a member of the younger generation.
... 1} is an indicator of exposure to the intervention, and X is a multivariate vector of observable characteristics (Becker and Ichino 2002). A value of propensity score should be estimated by various techniques; in this study, we employed the logistic regression. ...
Unemployment presents a significant challenge requiring attention not only in developing countries but also in economically developed ones. Active labour market policies offer a potential solution to address this issue. This study focuses on assessing the impact of the intervention called Contribution for Self-employment provided under the Act on Employment Services No. 5/2004 Coll in the Slovak Republic. This financial support is extended voluntarily to unemployed individuals seeking jobs and aims to partially defray the expenses associated with launching business ventures. The primary objective of this research is to quantify the effectiveness of the Contribution for Self-employment in enhancing the employment of its recipients, thereby gauging its efficacy in reducing unemployment. The evaluation employs a counterfactual impact assessment methodology, utilising propensity score matching for analysis, with propensity score estimated by the logistic regression. Data from the registry of jobseekers maintained by the Central Labour Office of Social Affairs and Family in Slovakia are utilised in this analysis. This study’s findings indicate a favourable impact of the contribution on the employment of its participants compared to the comparable non-participants. Consequently, this intervention emerges as a viable mechanism for supporting entrepreneurship and mitigating unemployment in Slovakia.
... Propensity score matching (PSM) served as the method for the "Salary Support for Export-Oriented Industru Workers" and "Revitalizing the Rural Economy and Job Creation" packages, addressing selection bias in the estimation of treated effects. PSM is recognized for mitigating bias in assessing the impact of economic policy interventions (Becker and Ichino 2002). The "Working Capital Loan for Industries and Service Sectors" package employed two techniques, namely, difference-in-difference (DID) and the purchasing managers' index (PMI), to gauge differences over time between receiver and non-receiver groups. ...
With the unexpected onset of COVID-19, governments across the world responded with a range of preventive measures, including the imposition of lockdowns. To mitigate the adverse effects of lockdowns arising from supply chain shocks and employment loss, governments worldwide chose to implement policies to stimulate their economies and keep them working. This study assesses the impact and effectiveness of four of these packages in Bangladesh, employing a mixed-method approach. These packages include "salary support for workers in export-oriented RMG industries", "working capital loans for affected industries and service sectors", "working capital loans for cottage, micro, small, and medium enterprises", and initiatives for "revitalizing the rural economy and job creation". Each package was examined individually because of their differences in beneficiary groups, implementation methods, and individual objectives. Quantitative analysis involved propensity score matching (PSM), the difference in difference model (DID), and structural equation modelling (SEM). Stakeholders, including policy implementers, Bangladesh Bank officials, policy analysts, academics, workers, and beneficiaries, contributed to the qualitative analysis through extensive key-informant interviews, providing a comprehensive assessment of intervention outcomes. Ultimately, the results show that the packages achieved their socioeconomic relief objectives for beneficiaries. The research examined both positive impacts and challenges in their implementation. It suggests that all four packages successfully achieved their goals, such as providing social and economic support, sustaining livelihoods, addressing marginalized groups' needs, ensuring survival for large industries and small businesses, and promoting employment. In order to better address future shocks, establishing a beneficiary database integrated with the national system is recommended for smoother policy rollout. Despite acknowledged limitations, including challenges in beneficiary identification, data availability, and time constraints, the study's unbiased estimations provide valuable insights to guide future policy directions in similar situations.
... In selecting the set of variables included in the logit models, researchers must take care to satisfy the balancing criteria. We follow the recommendation of Becker and Inchino (2002), whereby we split the sample into intervals before matching, then use the average propensity scores for the within-interval treated and control groups for comparison of significant differences across covariates. For the logit models, some models have more or less variables depending on which set of variables satisfied the balancing criteria while also maximizing the hit-or-miss and pseudo-R 2 guidelines that help assess the reliability of the propensity scores (Heckman et al. 1997). ...
In 2020, the COVID-19 pandemic changed the way many businesses conducted business. Notably, regulations imposed by states impacted how green-industry firms sold their plants and landscape products. However, not all states implemented the same stringency of regulations. Using an online consumer survey implemented in Jan 2021, we examine the impact of varying regulation stringencies across five treatment groups (Michigan, and New York, and low, medium, and high stringency). We estimate the difference between 2020 and 2019 self-reported expenditures, in conjunction with propensity score matching to compare each treatment with the other treatments. Results indicate that, for the most part, states with greater stringency associated with their COVID regulations did not impact plant and landscape expenditures negatively between 2019 and 2020. However, Michigan consumers did spend significantly less than medium- and high-stringency states for landscape products. Michigan was one of only two states that put qualifications on green-industry firms, and it was the only state to list green-industry firms as nonessential. Also, New York consumers spent more than low-stringency states, and low-stringency states spent less than high-stringency states for plants. Furthermore, there were no differences in online expenditures between state treatment groups. From a policy perspective, regulation type (i.e., shutting down green-industry sectors as Michigan did) had varying impacts across product categories within the green industry.
... The objective is to ensure that the influence of self-selection bias stemming from observables is minimized in the results. The comparison involves adopters and non-adopters based on shared assistance (Becker and Ichino 2002). In the initial stage of the PSM, a logit model is utilized to regress the adoption status of each farmer against factors that potentially influence the choice to adopt CSA practices. ...
Climate-induced increase in temperature and rainfall variability severely threaten the agricultural sector and food security in the Indian state of Odisha. Climate-smart agricultural (CSA) practices, such as crop rotation and integrated soil management, help farmers adapt to climate risk and contribute to a reduction in greenhouse gas (GHG) emissions. Therefore, this paper examines the impact of CSA practices on yield and income in vulnerable semi-arid districts of Odisha—Balangir, Kendrapara, and Mayurbhanj. We use primary survey data from 494 households collected in 2019–2020, using a multi-stage stratified sampling approach and structured questionnaire. Propensity score matching (PSM) and the two-stage least square method (2SLS) have been used to analyze the impact of CSA on income and productivity. Two instrument variables, namely distance to the extension office and percentage of adopters in a village, are used to control self-selection bias and endogeneity in our model. Both models show a positive and significant impact of the adoption of CSA on farmers’ productivity and income. The study sheds light on the significant contribution of CSA practices in fostering sustainable income growth amid environmental challenges. Overall, our results suggest that small and marginal farmers of Eastern India, a highly environmentally vulnerable area, can significantly improve their income and productivity by adopting CSA technology. Hence, policymakers should scale the adoption of CSA technology through effective extension programs.
... This study uses the Nearest Neighbour (NN), Kernel and Radius matching methods. No approach is without flaws, but combining several methodologies allows the robustness of impact estimations to be assessed [33]. ...
Women’s empowerment is an important policy agenda that is critical for developing countries like Bangladesh to achieve sustainable development goals (SDGs). The prime objective of this paper was to examine whether community savings groups can truly improve the economic conditions of women which turns into women’s empowerment in fishing communities or not. The
propensity score matching (PSM) and logistic regression technique were incorporated, and required data were collected from Community Savings Groups (CSG) interventions and non-CSG villages of coastal Bangladesh. Quantitative data were collected from 615 women comprising 306 CSG participants (treatment group) and 309 non-participants (control group). The results affirm CSG group members were economically more solvent and less dependent on borrowed money than non-CSG group members. Improved economic indicators (savings, income and expenditure) of CSG households make the foundation of attaining women’s empowerment for the intervened group. The findings revealed that CSG women performed better in various dimensions of leadership capacity than non-CSG women. Econometric analysis confirmed positive impacts of CSG interventions on savings, gross household income, earning from catching fish, alternative income generating activities (AIGAs), expenditure, and women’s empowerment. The initiatives of CSG
not only generate economic well-being but also contribute to women’s empowerment. Financial access, improved literacy and an enabling environment for the productive engagement of women reduce gender inequality in fishing communities. To sustain the benefits of CSG, establishing institutional linkages (advisory and financial), legality/registration of CSGs from the government authority, and facilitation of alternative IGAs are crucial.
... Once the propensity score is evaluated for each unit, a matching algorithm must be chosen to match the treatment group with the control group from their propensity score. The most widely used matching algorithms after propensity score evaluation are Nearest-Neighbour (NN) Matching, Radius or Caliper Matching, Kernel Matching and Stratification Matching [34]. In this study, we will use nearest neighbour matching within caliper (NNC), a combination of NN and Caliper matching. ...
Veterinarians are a pivotal force in addressing animal health and welfare surveillance, with a critical role in improving public health security and increasing the profits of livestock farmers. Yet, the veterinary profession is adversely affected by personnel shortages, particularly in rural areas. Since the health of people, animals and their shared environment are interconnected in a One Health perspective, a set of policies are required to ensure public health by attraction and retention of veterinarians in rural areas. In France, a tutored internship programme, financially subsiding students and mentors to execute a training period in remote rural areas, was promoted to better integrate and retain veterinary students ending their veterinary training. This paper aims to evaluate how veterinarians’ tutored internships influences students’ choices for rural practice, using three different statistical methods derived from causal inference theory. Using survey data for the period 2016–2020, we show that: (i) the average effect of the tutored internship on veterinarians’ work in food animal sector is not significant; and that (ii) the tutored internship leads veterinarians with a low share of work in the food animal sector to have a rural practise after they graduated between 13 and 20% greater than those who did not participate in the tutored internship.
... Propensity score weighting was used to minimize potential biases introduced by baseline differences in the treatment and comparison group and to approximate findings obtained from randomized control trials (Becker & Ichino, 2002;Moons, 2020). This method attempts to equalize the mean values of potentially confounding observed covariates in the treatment and comparison groups. ...
This study examines whether [Reading App], an adaptive, personalized game-based program designed to help young learners build a foundation for reading comprehension and literacy, can improve early reading skills for pre-kindergarteners and kindergarteners (N = 402 treatment, 690 comparison). This study measured the feasibility of implementing [Reading App] in classrooms, teacher and student experiences of using the application, and growth in literacy outcomes between students in classrooms using [Reading App] and students in classrooms not using the application, controlling for baseline performance. Results of hierarchical linear models showed that: (a) kindergarteners who used [Reading App] made significantly greater gains on end-of-year literacy assessment than the comparison group, especially in alphabet knowledge and (b) pre-kindergarteners who mastered at least 16 alphabet skills in [Reading App] experienced greater gains in the skill than comparison group peers. Teacher surveys and interviews suggested that [Reading App] is an easy-to-use, effective, engaging learning resource that empowers them to provide personalized instruction and foster a more equitable classroom environment. The study provides initial evidence of [Reading App]’s effectiveness as a program that can enhance educators’ capacity to address learner variability and provide personalized instruction for all learners.
... According to Becker and Ichino (2002), the PSM method should satisfy the following two assumptions: a. Conditional independence assumption (CIA) or confoundedness where outcome variables are independent of participation given X. This is expressed as b. ...
Different social protection programs are designed in low‐income countries to eradicate poverty and improve the food security of poor people. Currently, the Urban Productive Safety Net Program (UPSNP) is designed to support those who are living in poverty and face food insecurity with predictable and reliable support through food, cash, or vouchers. However, limited empirical evidence has been presented about the significant impact of the program on the well‐being of households who participated in the program and the factors that affect the households' decision to participate in the program. Hence, this study aims to evaluate the impact of participation in UPSNP on well‐being using household survey data gathered from three main cities of Ethiopia: Dire Dawa, Harar, and Jigjiga in 2022. This study employed both propensity score matching (PSM) and endogenous switching regression (ESR) models to assess the impact of the UPSNP. The result shows that the probability of a household's participation decision is determined by the age of the household head, number of children, savings, house ownership, employment status of the household head, and shock. Furthermore, we found a consistently positive impact across models, indicating that participation in UPSNP reduces poverty and increases food security of households.
... Ensuring that the PSM estimators identify a consistently estimate the treatment effects of interest leads to the following assumption (Becker & Ichino, 2002;Caliendo & Kopeinig, 2008): ...
Using six sweeps of data from the 1958 British National Child Development Study (NCDS), we employ a quasi-parametric approach of propensity score matching to estimate the impacts of higher education attainment on a wide range of health-related outcomes for cohorts at ages 33, 42, and 50. The non-pecuniary benefits of higher education on health are substantial. Cohorts with higher levels of education are more likely to report better health, maintain a healthy weight, refrain from smoking, exhibit a lower frequency of alcohol consumption, and are less likely to be obese. The effects on self-reported health, body mass index (BMI), drinking alcohol increase with age, but continuously decrease with smoking frequency. When considering gender heterogeneity, higher education has a more significant effect on BMI and the likelihood of obesity for males, while it has a greater impact on self-reported health, drinking alcohol, and smoking frequencies for females. Furthermore, we find no significant evidence that higher education reduces the likelihood of depression. The results of the Rosenbaum bounds sensitivity analysis suggest that, although our overall results demonstrate robustness, there may still be unobserved hidden bias in the relationship between higher education and self-reported health.
... Based on the previous analysis, whether hospitals provide VIP services may not be completely random, which may cause self-selection bias. To alleviate this endogeneity, the non-alternative one-to-one nearest neighbor matching method (60, 61) was used to match the control group for each year's treatment group based on propensity score matching (62). Combined with the observable matching variables (regional GDP per capita (GDP deflated), ownership, profit or not, and support health insurance or not), the predicted probability of each hospital offering VIP medical care was calculated, and then the only control group that did not provide VIP medical care was found for each hospital providing VIP medical services. ...
This study examines the causal impact of very important person (VIP) medical services on hospital total factor productivity in Deyang, a prefectural-level city in western China, spanning the years 2015–2020. This aims to offer empirical evidence and policy recommendations for the implementation of VIP practices in the medical field. A secondary unbalanced panel dataset of 416 observations was collected from the annual reports of the Health Commission and 92 eligible medical institutions were included. This study utilized a two-stage strategy. First, the Global Malmquist index was used to calculate the total factor productivity and its decomposition terms for hospitals from 2015 to 2020. In the second stage, two-way fixed effects models and Tobit models were used to identify the relationship between VIP medical services and hospital efficiency; instrumental variables were used to solve potential endogeneity problems in the model. The results showed that VIP medical services had a significantly negative impact on medical institutions’ efficiency. The technological advances and pure technical efficiency related to VIP medical care may help explain these negative impacts, which were heterogeneous across groups divided by the nature of the hospital and the outside environment. It is imperative to prioritize the standardized provision of VIP medical services for medical institutions, optimize management and service process, enhance the training of clinical and scientific research capabilities of medical personnel, and scientifically allocate resources for both VIP and general medical services. This will help mitigate health inequality while improving the overall quality of medical services.
... We report the area under the receiver operating characteristic curve (AUC) to examine the accuracy of our propensity score models in predicting PSI cases. To ensure that the matched samples would have sufficient statistical power even though PSI-related adverse events are rare, we used a nearest neighbor 1:3 matching (instead of the more traditional 1:1 matching) [29]. However, since 1:1 matching has been used more frequently in previous studies, we also provide results of a 1:1 matching in the supplementary material (S1-S4 Tables in S2 Appendix) to demonstrate similar findings under different matching conditions. ...
There currently exists no comprehensive and up-to date overview on the financial impact of the different adverse events covered by the Patient Safety Indicators (PSIs) from the Agency for Healthcare Research and Quality. We conducted a retrospective case-control study using propensity score matching on a national administrative data set of 1 million inpatients in Switzerland to compare excess costs associated with 16 different adverse events both individually and on a nationally aggregated level. After matching 8,986 cases with adverse events across the investigated PSIs to 26,931 controls, we used regression analyses to determine the excess costs associated with the adverse events and to control for other cost-related influences. The average excess costs associated with the PSI-related adverse events ranged from CHF 1,211 (PSI 18, obstetric trauma with instrument) to CHF 137,967 (PSI 10, postoperative acute kidney injuries) with an average of CHF 27,409 across all PSIs. In addition, adverse events were associated with 7.8-day longer stays, 2.5 times more early readmissions (within 18 days), and 4.1 times higher mortality rates on average. At a national level, the PSIs were associated with CHF 347 million higher inpatient costs in 2019, which corresponds to about 2.2% of the annual inpatient costs in Switzerland. By comparing the excess costs of different PSIs on a nationally aggregated level, we offer a financial perspective on the implications of in-hospital adverse events and provide recommendations for policymakers regarding specific investments in patient safety to reduce costs and suffering.
... The term "palliatives" here refers to relief measures or provisions intended to alleviate the immediate suffering or distress of individuals facing economic hardship due to the pandemic. 2. The Stata "pscore" command, developed by Becker and Ichino (2002), is used to estimate the propensity score and test for the balancing property. The results are available from the author upon request. ...
This study examines the impact of government and humanitarian interventions during COVID-19 on food security among vulnerable populations in Nigeria. Data from a survey conducted by the United Nations High Commissioner for Refugees (UNHCR) in July 2020 was used, with a sample of 4,833 households drawn from various vulnerable groups such as refugees
and internally displaced persons (IDPs). Using matching methods, the results indicate that targeted support initiatives helped alleviate immediate food insecurity challenges during the pandemic. This underscores the significance of focused interventions in mitigating the adverse impact of economic shocks on food security for vulnerable populations.
... Only firms with comparable characteristics are on support, and those off support are dropped off. This is one significant disadvantage of propensity score matching because it is possible to lose many observations (Becker & Ichino, 2002). Nevertheless, since this procedure imposes the region of common support, constrained firms are matched with unconstrained firms with comparable characteristics (Bravo-Ureta et al., 2012;Caliendo & Kopeinig, 2008;Rosenbaum & Rubin, 1983). ...
Small and medium-sized enterprises (SMEs) are fundamental to national economic development through poverty alleviation and employment creation. However, credit constraints severely challenge small and medium-sized enterprises’ (SMEs) performance. Therefore, this study aims to establish the link between credit constraints and the financial performance of SMEs. The paper further examines whether the differences in performance between constrained and unconstrained firms can contribute to poverty reduction. To this end, the endogenous switching regression was applied to the cross-sectional data from a firm-level survey involving 520 SMEs from Cameroon. The paper considered four credit constraint categories: risk, quantity, price, and transaction cost. Results reveal that 74% of SMEs are credit constrained, and the majority (43%) are quantity constrained. Interestingly, risk and price-constrained SMEs are mainly more profitable than unconstrained SMEs, while quantity and transaction-cost-constrained SMEs are less profitable than their counterparts. Consequently, the impact on poverty reduction is equally varied. The scholarly and policy implications of the study are that alleviating credit constraints among SMEs does not always augment the profitability of the SMEs and contributes towards poverty eradication.
... After calculating the propensity score, both matched samples need to satisfy the balancing property (Becker and Ichino, 2002). Covariates which are the main predictors of the treatment status need to be balanced. ...
Research has shown that the Indonesian Family Hope CCT Program aimed at improving children’s health and education of poor households, has had significant impacts. Using different data, we assess whether it changed recipients’ behaviour along other metrics. Despite checks and constraints on how transfers can be spent, low-income families can still spend some of their extra cash on frivolous goods, rather than health and education as intended. Our results show that the program leads recipients to mildly decrease their levels of frivolous consumption and increase their share of spending on education (not for health) when compared to non-participants.
... Through this process, all the pixels from the treatment and the control groups that have similar characteristics were selected as samples for the estimation of the effect of the GECP on the NDVI of grassland. Third, to estimate the effect of the GECP on the NDVI of grassland, we calculated the average treatment effect for the treated (ATT) using the method [52] that is shown in the following equation: ...
The Payments for Ecosystem Services (PES) program is an innovative approach that provides economic incentives directly to natural resource exploiters in order to shape their behavior. Although the implementation of PES programs and the context in which these programs are implemented are often different across space, the spatial heterogeneities in the impacts of PES programs are often neglected in studies. In this study, we demonstrate the spatial and temporal dynamics of the Normalized Differential Vegetation Index (NDVI) in the grassland where China’s Grassland Ecological Compensation Program (GECP) has been implemented, and we evaluate the impacts of the GECP on the NDVI. We found that most of the grassland in the GECP area showed small changes in NDVI between 2000 and 2010. On average, the GECP only had a marginal positive effect on the NDVI of grassland. Although the magnitude of impacts of the GECP was relatively small in most places where the program is implemented, we detected substantial heterogeneities in the impacts of the GECP on the NDVI. The impact of the GECP on the NDVI differed substantially, particularly between Inner Mongolia, Sichuan, and other provinces. Our findings suggest that there can be substantial heterogeneities in the impacts of PES programs across space, which can be leveraged to promote the efficacy of the GECP and many other PES programs around the world.
... the nearest neighbor matching (NNM) and kernel based matching (KBM) algorithms were used. After matching for NNM and KBM, several balancing tests were employed to assess the matching quality, such as checking a reduction in the median absolute bias, the value of R 2 , and the p-value of joint significance of covariates before and after matchings (Ali & erenstein, 2017;Becker & ichino, 2002;caliendo & Kopeinig, 2008;Gebre et al., 2023a;rahut & Ali, 2018). ...
Linking farmers to high-value markets continue to be a viable option for breaking the food insecurity. Many studies have emphasized on factors determining smallholders’ participation in high-value markets. Previous studies have methodological limitations since they neglected the associated effects on food security. Therefore, the purpose this study was to fill a knowledge gap by investigating the factors influencing smallholder avocado farmers’ participation in high-value markets and how it impacts households’ food security in the Aleta chuko district, Ethiopia. The primary data was collected from randomly selected 389 avocado producers using a semi-structured questionnaire. Descriptive statistics, inferential statistics, and propensity score matching model were used to analyze the data. The result of the binary logit model revealed that the participation of avocado producers in a high-value market was influenced by age, educational status, the quantity of avocados sold, and price of avocado and market information. The ATT estimation of PSM model indicated that Avocado producers who participated in high-value market channels have higher food security status (by 4.3–4.8%) compared to those who were not. Thus, this study suggested that concerned bodies in Ethiopia should encourage more households to participate in the high-value markets.
... Propensity score matching was performed with a 1:1 matching ratio for obese and non-obese patients. Propensity scores were computed by modeling a logistic regression with the dependent variable being the odds of experiencing the exposure of interest (i.e., surgery) and the independent variables as age, sex, year of treatment, emergent surgery, type of inflammatory bowel disease, Charlson Comorbidity Index, operative approach, type of operation, income quartile, hospital bed size, and hospital region [19]. Patients were matched with nearest neighbor matching without replacement [20]. ...
Purpose
Up to 40% of patients with inflammatory bowel disease (IBD) are obese. Obesity is a well-known risk factor for increased perioperative morbidity, but this risk has never been quantified in IBD patients undergoing abdominal surgery using the United States National Inpatient Sample (NIS) database. This study aims to compare postoperative morbidity between obese and non-obese patients undergoing bowel resection for IBD using recent NIS data.
Methods
Adult patients who underwent bowel resection for IBD from 2015 to 2019 were identified in the NIS using ICD-10-CM coding. Patients were stratified into obese (BMI > 30 kg/m²) and non-obese groups, then propensity score matched (PSM) for demographic, operative, and hospital characteristics. The primary outcome was postoperative in-hospital morbidity. Secondary outcomes included postoperative in-hospital mortality, system-specific postoperative complications, total admission healthcare costs, and length of stay (LOS). Univariable and multivariable regressions were utilized.
Results
Overall, 6601 non-obese patients and 671 obese patients were identified. The PSM cohort included 659 patients per group. Obese patients had significantly increased odds of experiencing postoperative in-hospital morbidity (aOR 1.50, 95% CI 1.10–2.03, p = 0.010) compared to non-obese patients. Specifically, obese patients experienced increased gastrointestinal complications (aOR 1.49, 95% CI 1.00–2.24, p = 0.050), and genitourinary complications (aOR 1.71, 95% CI 1.12–2.61, p = 0.013). There were no differences in total admission healthcare costs (MD − $2256.32, 95% CI − 19,144.54–14,631.9, p = 0.79) or LOS (MD 0.16 days, 95% CI − 0.93–1.27, p = 0.77).
Conclusions
Obese IBD patients are at greater risk of postoperative in-hospital morbidity than non-obese IBD patients. This supports targeted preoperative weight loss protocols for IBD patients to optimize surgical outcomes.
The study examines the links between India’s outward foreign direct investment (OFDI) and possible income-shifting activities by the parent firms. The exercise is undertaken by examining the impact of OFDI on parent firms’ tax payments, profitability, debt, and intangible assets. The study is driven by the observation that nearly 68% of India’s OFDI flows between 2008 and 2020 were directed to offshore financial centers (OFC). The study relies on the Reserve Bank of India’s (RBI) firm-level overseas direct investment data and the Prowess database. We employed the propensity score matching (PSM) technique in combination with the difference-in-difference method to investigate the post-investment effects. Results suggest that overseas investments have resulted in lower payment of corporate taxes, as well as indirect and direct taxes at home. Moderate negative effects were observed in the case of the profitability of the parent firm. On the contrary, OFDI resulted in higher debt levels, particularly for firms investing in OFC destinations. A positive impact on the firm’s intangible assets suggests that income shifting via relocation of intangible assets is not evident. The analysis calls for policies to counter the possible tax leakage at home due to firms investing overseas, especially in OFCs.
JEL Classification F23, C14
Worldwide, more than 130 million infants are born each year and a considerable number of 13.5 million of these children have inbred parents. The present study aimed to investigate the association between parents’ consanguinity and chronic illness among their children and grandchildren in India. The nationally representative data, Longitudinal Aging Study in India, 2017–2018, Wave 1 was used for the present study. Bivariate analysis, a probit model, and propensity score estimation were employed to conduct the study. The study observed the highest prevalence of consanguinity marriage in the state of Andhra Pradesh (28%) and the lowest in Kerala (5%) among the south Indian States. People who lived in rural areas, belonged to the richer wealth quintile and Hindu religion were the significant predictors of consanguinity marriage in India. For individuals who were in consanguineous marriages, there was 0.85%, 0.84%, 1.57% 0.43%, 0.34%, and 0.14% chances of their children and grandchildren developing psychotic disorders, heart disease, hypertension stroke, cancer, and diabetes, respectively. Moreover, around 4.55% of the individuals have a history of birth defects or congenital disorders. To address the risk of complicated illnesses due to the consanguinity of marriage, medical, genetic, and social counselling services are required.
Purpose
Section 177 of the Company Act 2013 and Regulation 18 of the Listing Obligations and Disclosure Requirements 2015 allow the audit committee to invite firm executives to participate in the audit committee meetings. In this study, we investigate the negative impact of the presence of invitees in the audit committee on firm value.
Design/methodology/approach
The study uses the Propensity Score Matching and Difference-In-Difference methodology (henceforth, PSM-DID) to establish a causal relationship between the presence of invitees and firm value. The final sample consists of 24,232 firm-year observations representing 4,493 distinct firms from 2016 to 2021. We also address the endogeneity and autocorrelation issues using the system-generalized method of moments (henceforth, GMM) as a robustness test.
Findings
We find that the presence of invitees in the audit committee decreases the firm value because investors consider this an alarming signal. We further find that the firms, audited by the Big 4, do not experience a decrease in firm value due to higher audit quality, whereas the firms with high promoter ownership experience a decrease due to the presence of agency cost.
Originality/value
We contribute to the literature on firm value and strengthen the literature on the importance of good governance in a developing nation using the signalling theory. This study adds to the understanding of firm value. The findings have implications for management literature and are valuable for policymakers and standard setters in evaluating the impact of disclosures in the capital market. The managerial implications emphasize the need for careful consideration of invitees in audit committees, considering industry, regulatory environment, and firm goals. Firms are advised to assess the benefits and costs, monitor the impact regularly, and strengthen internal controls.
Firm clusters are considered as a contributing factor to local economic development. However, there are limited studies on the effect of firm clusters on the well‐being of rural communities, particularly in terms of income improvement, poverty reduction, and migration. Our research aims to shed light on these relationships at both the household and commune levels. For empirical analysis, we employ the propensity score matching method to mitigate endogeneity bias. Our results reveal the role of firm clusters in increasing income and reducing poverty. Firm clusters also contribute to decreasing labor emigration and attracting immigrants. However, the magnitude of these impacts is relatively small, with moderate effects on income and modest effects on poverty and migration. In particular, firm clusters reduce the commune poverty rate by around 2.36%–2.51% and enhance household annual income by approximately 16.46–17.08 million VND (725–752 USD). Furthermore, analyses at the household level highlight the significance of larger clusters in improving household income. Our research underscores policy implications for rural development with a specific emphasis on firm clusters.
Durante las últimas décadas, el término nini —referido a los jóvenes que no estudian ni trabajan— se ha instalado en las discusiones académicas y en los debates acerca de qué políticas conviene aplicar con el fin de atender las necesidades de los jóvenes vulnerables ante la exclusión social. Sin embargo, en un contexto como el peruano, en el cual la informalidad alcanza al 70% de la PEA y la precariedad laboral resulta tan importante como la inactividad, el énfasis en los jóvenes nini resulta limitado.
Sobre la base de un diseño de métodos mixtos, este estudio analiza, por un lado, la pertinencia de la definición predominante de nini para comprender a los jóvenes urbanos en situación de vulnerabilidad en el Perú. Por otro lado, plantea una definición alternativa, que incluye a los jóvenes involucrados en trabajos informales, inestables y precarios –quienes corren un alto riesgo de caer en la exclusión social–; asimismo, analiza los factores asociados a su situación.
La conclusión principal es que el concepto tradicional de nini no es adecuado para analizar la vulnerabilidad juvenil en contextos como el peruano. Así, es necesario ir más allá de los nini e incluir en el análisis a los jóvenes que se encuentran insertos en empleos altamente precarios e inestables del sector informal.
Non-alcoholic fatty liver disease (NAFLD) is the commonest cause of chronic liver disease in patients with diabetes; limited data suggested that statins may reduce the risk of NAFLD progression. This study aimed to examine the association between statins and the development or progression of NAFLD in veterans with diabetes. In a new-user negative control design, we conducted a retrospective propensity-score (PS) matched cohort study of patients with diabetes between 2003 and 2015. After excluding patients with other causes of liver disease, we formed PS using 85 characteristics. The primary outcome was a composite NAFLD progression outcome. Primary analysis examined odds of outcome in PS matched cohort. Post-hoc analysis included PS-matched cohort of statin users with intensive lowering of LDL-cholesterol versus low-intensity lowering. We matched 34,102 pairs from 300,739 statin users and 38,038 non-users. The composite outcome occurred in 8.8% of statin users and 8.6% of non-users (odds ratio [OR]: 1.02; 95% confidence interval [95%CI]: 0.97-1.08). In the post-hoc analysis, intensive lowering of LDL-cholesterol compared to low-intensity showed increased NAFLD progression (OR: 1.21, 95%CI: 1.13-1.30). This study showed that statin use in patients with diabetes was not associated with decreased or increased risk of NAFLD progression. Intensive LDL-cholesterol lowering compared to low-intensity LDL-cholesterol lowering was associated with an increased risk of NAFLD progression.
Objectives
This study aimed to explore the effects of short birth spacing (SBS), which is defined as a period of less than 33 months between two successive births, on multiple concurrent forms of child malnutrition (MCFCM) and at least one form of child malnutrition (ALOFCM) using propensity score matching (PSM).
Methods
This study used data extracted from the 2017-18 Bangladesh Demographic and Health Survey. PSM with four different distance functions, including logistic regression, classification and regression tree, single hidden layer neural network and random forest, were performed to evaluate the effects of SBS on MCFCM and ALOFCM. We also explored how the effects were modified in different subsamples, including women’s empowerment, education and economic status (women’s 3E index)–constructed based on women’s decision-making autonomy, education level, and wealth index, and age at marriage, and place of residence.
Results
The prevalence of SBS was 22.16% among the 4652 complete cases. The matched samples of size 2062 generated by PSM showed higher odds of MCFCM (adjusted OR (AOR)=1.25, 95% CI=1.02 to 1.56, p=0.038) and ALOFCM (AOR=1.20, 95% CI=1.01 to 1.42, p=0.045) for the SBS children compared with their counterparts. In the subsample of women with 3E index≥50% coverage, the SBS children showed higher odds of MCFCM (AOR: 1.43, 95% CI=1.03 to 2.00, p=0.041] and ALOFCM (AOR: 1.33, 95% CI=1.02 to 1.74, p=0.036). Higher odds of MCFCM (AOR=1.27, 95% CI=1.02 to 1.58, p=0.036) and ALOFCM (AOR=1.23, 95% CI=1.02 to 1.51, p=0.032) for SBS children than normal children were also evident for the subsample of mothers married at age≤18 years.
Conclusion
SBS was significantly associated with child malnutrition, and the effect was modified by factors such as women’s autonomy and age at marriage.
O objetivo deste estudo é mensurar o diferencial de desempenho acadêmico entre estudantes da educação de jovens e adultos (EJA) e do ensino médio em escolas regulares no Ceará. Utilizando dados do Sistema Permanente de Avaliação da Educação Básica do Ceará (Spaece) e dos Censos Escolares de 2012 e 2014, os alunos foram pareados pelo método de escore de propensão, e suas características, ponderadas por entropia. Os resultados apontam que os integrantes da EJA possuem desempenho em língua portuguesa (-0,82 desvio-padrão) e matemática (-0,58 desvio-padrão) inferior aos do ensino regular. Dessa forma, espera-se que a gestão escolar tente equalizar as disparidades existentes entre estas modalidades de ensino, proporcionando um desempenho com maior equidade.
Shoot grafting (grafting) on plants is a cultural technology that reduces production time compared to new planting methods. This has changed the habits of local coffee farmers, who previously carried out plantation activities only when entering the harvest season, to become more intentional because, generally, the nature of grafting coffee requires intensive attention and care to continue to bear fruit throughout the season. This research aims to analyze the impact of applying grafting to local varieties of coffee plants, especially in smallholder plantations, on the resulting production. Multi-stage sampling was used to choose326 coffee farmers divided into two farming groups, namely 120 farmers from the group of farmers who have implemented shoot grafting (as the treatment group) and 206 farmers who have not implemented shoot grafting or as the control group. The probity model estimates the probability of grafting and then chooses a matching algorithm. In the matching process between covariates, the nearest neighbor without replacement (NN) technique is a matching process for each covariate with only one chance. The results show that grafting technology has not significantly impacted the production of locally cultivated coffee. However, applying top grafting technology influences the number of plants and farmer experience. Keywords: family farms; shoot grafting; changes in agricultural systems; Impacto da enxertia na produção local de café baseada em plantações populares na Província de Bengkulu, Indonésia RESUMO: A enxertia de brotos em plantas pode ser considerada como uma técnica de manejo cultural para reduzir o tempo de produção quando comparado aos novos métodos de plantio. A adoção dessa técnica, provocou uma mudança nos hábitos dos cafeicultores locais na Indonésia; antes, esses produtores realizavam atividades de manejo apenas no início da época de colheita; agora, precisaram se tornar mais intensivistas, visto que a enxertia do café exige atenção e cuidados frequentes para continuidade da produção de frutos ao longo do ano. temporada. Esta pesquisa tem como objetivo analisar o impacto da aplicação de enxertia em variedades locais de café na Província de Bengkulu (Indonésia), especialmente em plantações de pequenos agricultores e nas suas produções resultantes. A amostragem ocorreu em vários estágios e foi usada para escolher 326 cafeicultores, divididos em 2 grupos agrícolas, sendo: i) 120 agricultores que implementaram a enxertia de brotos (como grupo de tratamento); ii) 206 agricultores que não implementaram a enxertia de brotos. Um modelo probabilidade foi usado para estimar a probabilidade de enxerto e, em seguida, escolher um algoritmo de correspondência. No processo de emparelhamento entre covariáveis, foi empregada a técnica do vizinho mais próximo sem substituição (NN), que indica que o emparelhamento para cada covariável possui apenas uma chance. Os resultados mostram que a aplicação da tecnologia de enxertia provou não ter impacto significativo na produção do café cultivado localmente na Indonésia; porém, a aplicação da tecnologia de enxertia de topo, mostra influências no número de plantas e na experiência do agricultor. Palavras-chave: propriedades rurais familiares; enxertia; mudanças nos sistemas agrícolas.
Food security package loan has been found to be a critical instrument in order to improve the income of food insecure households. The main purpose of the program was to enhance the food insecure livelihood status through accessing of micro credit. Therefore, the objective of this study was to analyze the impact of Food Security Package Loan (FSPL) of micro credit service on the income and livelihood of food insecure households residing in West Belesa District. The study applied an econometric model of propensity score matching (PSM) to analyze the impact of FSPL on the income and livelihood of households based on data collected from a sample of 254 rural households (157 were food insecure and 97 food secure). The results of the econometric analysis display that FSPL participation significantly affects positively household's on-farm and off-farm income, employment, animal hold, saving and children participation in formal school. However, the food consumption level and types of house owned show no difference. This suggests that the stakeholders (government authorities, NGOs, aid agencies, etc) that deemed micro finance as a means to poverty reduction should take into account the implications of these indicator variables for better promotion of micro finance specifically FSPL and devise an intervention mechanism to further expand its impact towards improving food consumption and household asset building.
Although studies observe heterogeneity in the effects of adolescent childbearing on schooling, we currently know little about when this pattern emerged or how it changed across cohorts of women who lived in distinct periods of U.S. history. This article identifies the potential origins of effect heterogeneity in the educational costs of adolescent childbearing and extends recent advances in causal inference to detect group differences in heterogeneity. The analysis applies this approach to four cohorts of women from the National Longitudinal Surveys (NLS) who entered adolescence before, during, and after expansive economic, demographic, and cultural change in the twentieth century. Results suggest that the educational costs of adolescent childbearing, as well as heterogeneity in those costs, increased for women in the latter half of the twentieth century, especially for Millennial women born 1980 to 1984. We conclude that midcentury social changes fundamentally altered the educational costs of adolescent childbearing for women.
Importance
Ischemic stroke is a serious complication of cardiac intervention, including surgery and percutaneous procedures. Endovascular thrombectomy (EVT) is an effective treatment for ischemic stroke and may be particularly important for cardiac intervention patients who often cannot receive intravenous thrombolysis.
Objective
To examine trends in EVT for ischemic stroke during hospitalization of patients with cardiac interventions vs those without in the United States.
Design, Setting, and Participants
This cohort study involved a retrospective analysis using data for 4888 US hospitals from the 2016-2020 National Inpatient Sample database. Participants included adults (age ≥18 years) with ischemic stroke (per codes from the International Statistical Classification of Diseases, Tenth Revision, Clinical Modification ), who were organized into study groups of hospitalized patients with cardiac interventions vs without. Individuals were excluded from the study if they had either procedure prior to admission, EVT prior to cardiac intervention, EVT more than 3 days after admission or cardiac intervention, or endocarditis. Data were analyzed from April 2023 to October 2023.
Exposures
Cardiac intervention during admission.
Main Outcomes and Measures
The odds of undergoing EVT by cardiac intervention status were calculated using multivariable logistic regression. Adjustments were made for stroke severity in the subgroup of patients who had a National Institutes of Health Stroke Scale (NIHSS) score documented. As a secondary outcome, the odds of discharge home by EVT status after cardiac intervention were modeled.
Results
Among 634 407 hospitalizations, the mean (SD) age of the patients was 69.8 (14.1) years, 318 363 patients (50.2%) were male, and 316 044 (49.8%) were female. A total of 12 093 had a cardiac intervention. An NIHSS score was reported in 218 576 admissions, 216 035 (34.7%) without cardiac intervention and 2541 (21.0%) with cardiac intervention ( P < .001). EVT was performed in 23 660 patients (3.8%) without cardiac intervention vs 194 (1.6%) of those with cardiac intervention ( P < .001). After adjustment for potential confounders, EVT was less likely to be performed in stroke patients with cardiac intervention vs those without (adjusted odds ratio [aOR], 0.27; 95% CI, 0.23-0.31), which remained consistent after adjusting for NIHSS score (aOR, 0.28; 95% CI, 0.22-0.35). Among individuals with a cardiac intervention, receiving EVT was associated with a 2-fold higher chance of discharge home (aOR, 2.21; 95% CI, 1.14-4.29).
Conclusions and Relevance
In this study, patients hospitalized with ischemic stroke and cardiac intervention may be less than half as likely to receive EVT as those without cardiac intervention. Given the known benefit of EVT, there is a need to better understand the reasons for lower rates of EVT in this patient population.
Irrigation is one pillar of the Green Revolution that drove dramatic agricultural productivity gains across Asia. In Bangladesh, irrigation uptake has been so significant that 97% of dry‐season rice is now irrigated. While most Bangladesh monsoon rice is completely rainfed, supplementary irrigation is sometimes employed where late monsoon onset is potentially yield‐limiting. Station‐controlled experiments provide a narrative of positive yield benefits from supplementary irrigation. In contrast, statistical evaluations of actual farm experience mostly show no yield benefit and lower profitability for supplementary irrigation adopters. To add evidence on this controversial practice, we evaluated data from 2012 and 2015 Bangladesh farm household surveys with causality econometric approaches that control for differences between supplementary irrigation adopter and non‐adopter groups. After controlling for self‐selection and endogeneity, we found no statistically significant yield benefit for supplementary irrigation. Our results support scepticism about the profitability of supplementary irrigation. As such, we recommend careful consideration of the mixed evidence on effectiveness in future supplementary irrigation project benefit cost analyses. Further evidence over a longer time and accounting for a broader range of crops is also important moving forward.
The propensity score is the conditional probability of assignment to a particular treatment given a vector of observed covariates. Both large and small sample theory show that adjustment for the scalar propensity score is sufficient to remove bias due to all observed covariates. Applications include: (i) matched sampling on the univariate propensity score, which is a generalization of discriminant matching, (ii) multivariate adjustment by subclassification on the propensity score where the same subclasses are used to estimate treatment effects for all outcome variables and in all subpopulations, and (iii) visual representation of multivariate covariance adjustment by a two- dimensional plot.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org. This article uses propensity score methods to estimate the treatment impact o f the National Supported Work (N S W) Demonstration, a labor training program, on postintervention earnings. W e use data from Lalonde's evaluation o f nonexperimental methods that combine the treated units from a randomized evaluation o f the NSW with nonexperirnental comparison units drawn from survey datasets. W e apply propensity score methods to this composite dataset and demonstrate that, relative to the estimators that Lalonde evaluates, propensity score estimates o f the treatment impact are much closer to the experimental benchmark estimate. Propensity score methods assume that the variables associated with assignment to treatment are observed (referred to as ignorable treatment assignment, or selection on observables). Even under this assumption, it is difficult to control for differences between the treatment and comparison groups when they are dissimilar and when there are many preintervention variables. The estimated propensity score (the probability o f assignment to treatment, conditional on preintervention variables) summarizes the preintervention variables. This offers a diagnostic on the comparability o f the treatment and comparison groups, because one has only to compare the estimated propensity score across the two groups. W e discuss several methods (such as stratification and matching) that use the propensity score to estimate the treatment impact. When the range o f estimated propensity scores o f the treatment and comparison groups overlap, these methods can estimate the treatment impact for the treatment group. A sensitivity analysis shows that our estimates are not sensitive to the specification o f the estimated propensity score, but are sensitive to the assumption o f selection on observables. W e conclude that when the treatment and comparison groups overlap, and when the variables determining assignment to treatment are observed, these methods provide a means to estimate the treatment impact. Even though propensity score methods are not always applicable, they offer a diagnostic on the quality o f nonexperimental comparison groups in terms o f observable preintervention variables.
Policy makers view public sector-sponsored employment and training programs and other active labor market policies as tools for integrating the unemployed and economically disadvantaged into the work force. Few public sector programs have received such intensive scrutiny, and been subjected to so many different evaluation strategies. This chapter examines the impacts of active labor market policies, such as job training, job search assistance, and job subsidies, and the methods used to evaluate their effectiveness. Previous evaluations of policies in OECD countries indicate that these programs usually have at best a modest impact on participants' labor market prospects. But at the same time, they also indicate that there is considerable heterogeneity in the impact of these programs. For some groups, a compelling case can be made that these policies generate high rates of return, while for other groups these policies have had no impact and may have been harmful. Our discussion of the methods used to evaluate these policies has more general interest. We believe that the same issues arise generally in the social sciences and are no easier to address elsewhere. As a result, a major focus of this chapter is on the methodological lessons learned from evaluating these programs. One of the most important of these lessons is that there is no inherent method of choice for conducting program evaluations. The choice between experimental and non-experimental methods or among alternative econometric estimators should be guided by the underlying economic models, the available data, and the questions being addressed. Too much emphasis has been placed on formulating alternative econometric methods for correcting for selection bias and too little given to the quality of the underlying data. Although it is expensive, obtaining better data is the only way to solve the evaluation problem in a convincing way. However, better data are not synonymous with social experiments.
The propensity score is the conditional probability of assignment to a particular treatment given a vector of observed covariates. Previous theoretical arguments have shown that subclassification on the propensity score will balance all observed covariates. Subclassification on an estimated propensity score is illustrated, using observational data on treatments for coronary artery disease. Five subclasses defined by the estimated propensity score are constructed that balance 74 covariates, and thereby provide estimates of treatment effects using direct adjustment. These subclasses are applied within subpopulations, and model-based adjustments are then used to provide estimates of treatment effects within these subpopulations. Two appendixes address theoretical issues related to the application: the effectiveness of subclassification on the propensity score in removing bias, and balancing properties of propensity scores with incomplete data. Statistics Version of Record
This paper considers causal inference and sample selection bias in nonexperimental settings in which (i) few units in the nonexperimental comparison group are comparable to the treatment units, and (ii) selecting a subset of comparison units similar to the treatment units is difficult because units must be compared across a high-dimensional set of pretreatment characteristics. We discuss the use of propensity score-matching methods, and implement them using data from the National Supported Work experiment. Following LaLonde (1986), we pair the experimental treated units with nonexperimental comparison units from the CPS and PSID, and compare the estimates of the treatment effect obtained using our methods to the benchmark results from the experiment. For both comparison groups, we show that the methods succeed in focusing attention on the small subset of the comparison units comparable to the treated units and, hence, in alleviating the bias due to systematic differences between the treated and comparison units. © 2001 by the President and Fellows of Harvard College and the Massachusetts Institute of Technolog
This paper uses propensity score methods to address the question: how well can an observational study estimate the treatment impact of a program? Using data from Lalonde's (1986) influential evaluation of non-experimental methods, we demonstrate that propensity score methods succeed in estimating the treatment impact of the National Supported Work Demonstration. Propensity score methods reduce the task of controlling for differences in pre-intervention variables between the treatment and the non-experimental comparison groups to controlling for differences in the estimated propensity score (the probability of assignment to treatment, conditional on covariates). It is difficult to control for differences in pre-intervention variables when they are numerous and when the treatment and comparison groups are dissimilar, whereas controlling for the estimated propensity score, a single variable on the unit interval, is a straightforward task. We apply several methods, such as stratification on the propensity score and matching on the propensity score, and show that they result in accurate estimates of the treatment impact.
This paper considers causal inference and sample selection bias in non-experimental settings in which: (i) few units in the non-experimental comparison group are comparable to the treatment units; (ii) selecting a subset of comparison units similar to the treatment units is difficult because units must be compared across a high-dimentional set of pretreatment characteristics. We propose the use of propensity score matching methods, and implement them using data from the NSW experiment.
This paper takes the results of an employment and training program thatwas run as a field experiment, in which the participants were randomlyassigned into a treatment or a control group, and compares these results to the estimates that might have been produced by an econometrician who evaluated the program using the same econometric procedures that have been used in the program evaluation literature. This comparison shows that many of these econometric procedures fail toreplicate the experimentally determined results, and suggests that researchers should be aware of the potential for specification errorsin other nonexperimental evaluations. Copyright 1986 by American Economic Association.
Semiparametric methods are developed to estimate the bias that arises from using nonexperimental comparison groups to evaluate social programs and to test the identifying assumptions that justify matching, selection models and the method of difference in differences. Using data from an experiment on a prototypical social program and data from nonexperimental comparison groups, the authors reject the assumptions justifying matching and their extensions of it. The evidence supports the selection bias model and the assumptions that justify a semiparametric version of the method of difference-in-differences. The authors extend their analysis to consider applications of the methods to ordinary observational data. Journal: Econometrica
This paper considers whether it is possible to devise a nonexperimental procedure for evaluating a prototypical job training programme. Using rich nonexperimental data, we examine the performance of a two-stage evaluation methodology that (a) estimates the probability that a person participates in a programme and (b) uses the estimated probability in extensions of the classical method of matching. We decompose the conventional measure of programme evaluation bias into several components and find that bias due to selection on unobservables, commonly called selection bias in econometrics, is empirically less important than other components, although it is still a sizeable fraction of the estimated programme impact. Matching methods applied to comparison groups located in the same labour markets as participants and administered the same questionnaire eliminate much of the bias as conventionally measured, but the remaining bias is a considerable fraction of experimentally-determined programme impact estimates. We test and reject the identifying assumptions that justify the classical method of matching. We present a nonparametric conditional difference-in-differences extension of the method of matching that is consistent with the classical index-sufficient sample selection model and is not rejected by our tests of identifying assumptions. This estimator is effective in eliminating bias, especially when it is due to temporally-invariant omitted variables.
This paper reviews work on the effectiveness of different methods of matched sampling and statistical adjustment, alone and in combination, in reducing bias due to confounding x-variables when comparing two populations. The adjustment methods were linear regression adjustment for x continuous and direct standardization for x categorical. With x continuous, the range of situations examined included linear relations between y and x, parallel and non-parallel, monotonic non-linear parallel relations, equal and unequal variances of x, and the presence of errors of measurement in x. The percent of initial bias that was removed was used as the criterion. Overall, linear regression adjustment on random samples appeared superior to the matching methods, with linear regression adjustment on matched samples the most robust method. Several different approaches were suggested for the case of multivariate x, on which little or no work has been done.
There has been major progress in recent years in data-based bandwidth selection for kernel density estimation. Some “second generation” methods, including plug-in and smoothed bootstrap techniques, have been developed that are far superior to well-known “first generation” methods, such as rules of thumb, least squares cross-validation, and biased cross-validation. We recommend a “solve-the-equation” plug-in bandwidth selector as being most reliable in terms of overall performance. This article is intended to provide easy accessibility to the main ideas for nonexperts.
The propensity score is the conditional probability of assignment to a particular treatment given a vector of observed covariates. Both large and small sample theory show that adjustment for the scalar propensity score is sufficient to remove bias due to all observed covariates. Applications include: (i) matched sampling on the univariate propensity score, which is a generalization of discriminant matching, (ii) multivariate adjustment by subclassification on the propensity score where the same subclasses are used to estimate treatment effects for all outcome variables and in all subpopulations, and (iii) visual representation of multivariate covariance adjustment by a two-dimensional plot.
This paper advocates the use a nonparametric bound analysis to check the robustness of the results of applied evaluation studies to the problem of a lack of common support. The typical responses by researchers of either ignoring it, or obtaining estimates only for the subpopulation within the common support, can both be misleading: Ignoring the problem may result in biases because the comparison group may not be comparable. Deleting observations at best yields an estimator that is consistent for the common support. When treatment effects are hetereogenous inside and outside the common support this estimator is inconsistent. Furthermore, useful information is ignored, because the response of the treated to the treatment can be estimated even outside the common support. This information can be used to derive bounds with width depending on the configuration of the data. The application to an evaluation study of Swiss active labour market policies shows that the relevance of the bounds for changing the interpretation of the results depends very much on the particular data configuration.
Incluye bibliografía e índice
Estimation of average treatment effects in observational, or non-experimental in pre-treatment variables. If the number of pre-treatment variables is large, and their distribution varies substantially with treatment status, standard adjustment methods such as covariance adjustment are often inadequate. Rosenbaum and Rubin (1983) propose an alternative method for adjusting for pre-treatment variables based on the propensity score conditional probability of receiving the treatment given pre-treatment variables. They demonstrate that adjusting solely for the propensity score removes all the bias associated with differences in pre-treatment variables between treatment and control groups. The Rosenbaum-Rubin proposals deal exclusively with the case where treatment takes on only two values. In this paper an extension of this methodology is proposed that allows for estimation of average causal effects with multi-valued treatments while maintaining the advantages of the propensity score approach.
The assumption that the assignment to treatments is ignorable conditional on attributes plays an important role in the applied statistic and econometric evaluation literature. Another term for it is conditional independence assumption. This paper discusses identification when there are more than two types of mutually exclusive treatments. It turns out that low dimensional balancing scores, similar to the ones valid in the case of only two treatments, exist and be used for identification of various causal effects. Therefore, a comparable reduction of the dimension of the estimation problem is achieved and the approach retains its basic simplicity. The paper also outlines a matching estimator potentially suitable in that framework.
This paper develops the method of matching as an econometric evaluation estimator. A rigorous distribution theory for kernel-based
matching is presented. The method of matching is extended to more general conditions than the ones assumed in the statistical
literature on the topic. We focus on the method of propensity score matching and show that it is not necessarily better, in
the sense of reducing the variance of the resulting estimator, to use the propensity score method even if propensity score
is known. We extend the statistical literature on the propensity score by considering the case when it is estimated both parametrically
and nonparametrically. We examine the benefits of separability and exclusion restrictions in improving the efficiency of the
estimator. Our methods also apply to the econometric selection bias estimator.
Controlling Bias in Observational Studies
- W Cochran
- D B Rubin
Cochran, W. and Rubin, D.B. (1973), " Controlling Bias in Observational Studies ",
Sankhya 35, 417-446.
Becker is assistant professor of Economics at the University of Munich, Germany. He is also affiliated with CESifo and IZA
- O Sascha
Sascha O. Becker is assistant professor of Economics at the University of Munich, Germany.
He is also affiliated with CESifo and IZA.
Vuri for very helpful discussions and for testing our programs on their data. The comments of an anonymous referee helped to considerably improve the paper and the programs
- Barbara Sianesi
Barbara Sianesi and Daniela Vuri for very helpful discussions and for testing our programs on their data. The comments of an anonymous referee helped to considerably
improve the paper and the programs.