Regression Discontinuity Methods Research Articles

Association between implanted cardioverter-defibrillator and mortality for patients with left ventricular ejection fraction near 35%

Abstract Background Consensus guidelines support the use of implanted cardioverter-defibrillators (ICD) for primary prevention of sudden cardiac death in patients with either non-ischemic or ischemic cardiomyopathy with left ventricular ejection fraction ≤ 35%. Despite that class 1 recommendation, evidence from trials for efficacy specifically for patients with left ventricular ejection fraction (LVEF) near 35% is weak. Past trials are underpowered for this population and future trials are unlikely to be performed. As such, we aimed to use a large longitudinal echocardiography database linked to mortality to assess the comparative effectiveness of ICD therapy near the conventional primary prevention LVEF threshold. Methods Patients with lowest LVEF between 33-38% without an ICD prior to the lowest-LVEF echo (defined as "time zero") were identified by querying echocardiography data from November 28, 2001 until July 9, 2020 at large general hospital-based echocardiography registry linked to ICD treatment status. These data were then linked to mortality and comorbidity data. After inclusion and exclusion criteria were applied, 6,144 patients remained for the analytic cohort. Given evidence of covariate imbalance around the LVEF = 35% threshold, fuzzy regression discontinuity methods were considered but not used. Instead, to assess the association between ICD and mortality, multiple propensity score methods including matching and inverse probability of treatment weighting (IPTW) were used to create synthetic reference groups. Survival was then assessed with Cox proportional hazards models considering treatment status as a time-dependent covariate. Results In bivariate unadjusted comparisons, patients who received ICDs were substantially more often male (85.6% vs. 64.9%, p < 0.0001), more often white (87.3% vs. 82.9%, p < 0.048), and more often had a history of ventricular tachycardia (41.1% vs. 12.1%, p < 0.0001) and myocardial infarction (42.1% vs. 34.7%, p = 0.008). In the propensity matched sample, after accounting for time-dependence, there was no association between ICD and mortality (HR 1.09, 95% CI 0.86-1.39, p = 0.474). In the IPTW analysis, there also was no association between ICD and mortality (HR 1.09, 95% CI 0.91-1.30, p = 0.3672). Conclusions ICD therapy was not associated with reduced mortality near the conventional LVEF threshold of 35%. Although this treatment design cannot definitively demonstrate lack of efficacy, our results are concordant with available prior trial data. A definitive, well-powered trial is needed to answer the common and important clinical question of primary prevention ICD efficacy near the threshold of LVEF = 35%.

Dividing Lines: Racial Segregation across Local Government Boundaries

We describe the empirical relationship between local government boundaries and residential segregation in the United States. First, we study recent changes in the distribution of segregation within and between local governments in all metropolitan areas, using census block data on residential demographics over the period 1990–2020. We find that segregation across local government boundaries explains a substantial share of racial stratification, which has changed only little over the last thirty years. Next, we use spatial regression discontinuity methods to distinguish between household sorting due to neighborhood amenities and public goods provided by local governments. The prevalence of demographic discontinuities at local government boundaries suggest that between-jurisdiction segregation patterns cannot be explained solely by proximity to neighborhood amenities. We discuss implications for policy, showing that both between-jurisdiction segregation and jurisdictional discontinuities can partly explain the correlation between total segregation and racial gaps in educational outcomes. (JEL H41, H75, I24, I26, J15, R23)

Red Blood Cell Transfusion at a Hemoglobin Threshold of 7 g/dl in Critically Ill Patients: A Regression Discontinuity Study.

Rationale: In critically ill patients, a hemoglobin transfusion threshold of <7.0 g/dl compared with <10.0 g/dl improves organ dysfunction. However, it is unclear if transfusion at a hemoglobin of <7.0 g/dl is superior to no transfusion. Objectives: To compare degrees of organ dysfunction between transfusion and no transfusion at a hemoglobin threshold of <7.0 g/dl among critically ill patients using quasiexperimental regression discontinuity methods. Methods: We performed regression discontinuity analysis using hemoglobin measurements from patients admitted to intensive care units in three cohorts (Medical Information Mart for Intensive Care IV, eICU, and Premier Inc.), estimating the change in organ dysfunction (modified sequential organ failure assessment score) in the 24- to 72-hour window following each hemoglobin measurement. We compared hemoglobin concentrations just above and below 7.0 g/dl using a "fuzzy" discontinuity approach, based on the concept that measurement noise pseudorandomizes similar hemoglobin concentrations on either side of the transfusion threshold. Results: A total of 11,181, 13,664, and 167,142 patients were included in the Medical Information Mart for Intensive Care IV (MIMIC-IV), eICU, and Premier Inc. cohorts, respectively. Patient characteristics below the threshold did not differ from those above the threshold, except that crossing below the threshold resulted in a >20% absolute increase in transfusion rates in all three cohorts. Transfusion was associated with increases in hemoglobin concentration in the subsequent 24-72 hours (MIMIC-IV, 2.4 [95% confidence interval (CI), 1.1 to 3.6] g/dl; eICU, 0.7 [95% CI, 0.3 to 1.2] g/dl; Premier Inc., 1.9 [95% CI, 1.5 to 2.2] g/dl) but not with improvement in organ dysfunction (MIMIC-IV, 4.6 [95% CI, -1.2 to 10] points; eICU, 4.4 [95% CI, 0.9 to 7.8] points; Premier Inc., 1.1 [95% CI, -0.2 to 2.3] points) compared with no transfusion. Conclusions: Transfusion was not associated with improved organ dysfunction compared with no transfusion at a hemoglobin threshold of 7.0 g/dl, suggesting that evaluation of transfusion targets other than a hemoglobin threshold of 7.0 g/dl may be warranted.

Can gender quotas break the glass ceiling? Evidence from Italian municipal elections

Do gender quotas promote the election of female mayors? I exploit the roll-out of three gender quota policies governing Italian municipal elections to answer this question. The quotas institute minimum legal levels of female representation in the lists of councilor candidates and in the municipal governments' executive bodies, but do not directly target mayoral positions. Therefore, their ability to promote female political leadership entirely depends on whether they induce a broader increase in female representation in municipal governments beyond that mechanically required by the law. Using event-study and regression discontinuity methods, I decompose the aggregate effects of the quotas into mechanical effects, driven by compliance with the law, and additional impacts, or “acceleration effects.” The reforms have increased the share of female politicians in the less senior government positions beyond the minimum legal level, but there is no evidence of effects on the mayoral position. I also find no evidence of effects on female mayoral candidacies or on electoral support for female mayoral candidates. I conclude that the acceleration effects produced by the quotas have been too weak to advance female political leadership in municipal governments.

Economic analysis of the environmental sustainability of agriculture: recent studies using quasi-experimental methods

PurposeThe purpose of this review article is to demonstrate how the quasi-experimental approach has been used to study environmental and natural resource issues related to agricultural production.Design/methodology/approachThis review article first provides a short introduction to the quasi-experimental approach using the potential outcomes framework and then uses studies on the environmental sustainability of agricultural production to illustrate how quasi-experimental methods have been applied. Papers reviewed consist of studies that estimate the environmental externalities from agricultural production, evaluate agri-environmental and other related policies and programs, and demonstrate issues related to on-farm resource use and climate adaptation.FindingsDifference-in-differences (DID) and two-way fixed effects methods that utilize the spatial and temporal variation in panel data are widely used to estimate the causal impact of changes in agricultural production and policy on the environment. Utilizing the discontinuities and limits created by agricultural policies and regulations, local treatment effects on land and other input use are estimated using regression discontinuity (RD) or instrumental variable (IV) methods with cross-sectional data.Originality/valueChallenges faced by the food systems have made agricultural sustainability more critical than ever. Over the past three decades, the quasi-experimental approach has become the powerhouse of applied economic research. This review article focuses on quasi-experimental studies on the environmental sustainability of agriculture to provide methodological insights and to highlight gaps in the economics literature of agricultural sustainability.

Breaking Ties: Regression Discontinuity Design Meets Market Design

Many schools in large urban districts have more applicants than seats. Centralized school assignment algorithms ration seats at over‐subscribed schools using randomly assigned lottery numbers, non‐lottery tie‐breakers like test scores, or both. The New York City public high school match illustrates the latter, using test scores and other criteria to rank applicants at the city's screened schools, combined with lottery tie‐breaking at the rest. We show how to identify causal effects of school attendance in such settings. Our approach generalizes regression discontinuity methods to allow for multiple treatments and multiple running variables, some of which are randomly assigned. The key to this generalization is a local propensity score that quantifies the school assignment probabilities induced by lottery and non‐lottery tie‐breakers. The utility of the local propensity score is demonstrated in an assessment of the predictive value of New York City's school report cards. Schools that earn the highest report card grade indeed improve SAT math scores and increase graduation rates, though by much less than OLS estimates suggest. Selection bias in OLS estimates of grade effects is egregious for screened schools.

Child Health and Parental Responses to an Unconditional Cash Transfer at Birth

We estimate the impact on child health of the unanticipated introduction of the Australian Baby Bonus, a $3,000 one-off unconditional cash transfer at birth. Using regression discontinuity methods and linked administrative data from South Australia, we find that treated babies had fewer preventable, acute, and urgent hospital presentations—medical care available without co-payments—in the first two years of life. The payment later increased demand for elective care, which requires planning, medical referrals, and often co-payments. Our effects are strongest for disadvantaged families. Our findings suggest that up to 34% of the payout were recouped within the first year.

Does the party in power affect FDI? First causal evidence from narrow margin US state elections

Does the party of government influence the amount and type of inward foreign investment? Correlational studies provide inconsistent evidence. Moreover no existing study, for any level of government or any jurisdiction, uses methods that allow for causal inference. We apply regression discontinuity methods to a set of narrow margin US gubernatorial elections. Over the course of a 4-year term the election of a Republican governor causes a 17% boost in the growth of manufacturing-oriented FDI stock, compared to a Democrat. However, the same approach provides no evidence that partisanship matters for the overall level of FDI.

The Causal Effect of Polls on Turnout Intention: A Local Randomization Regression Discontinuity Approach

AbstractThis letter reports the results of a study that combined a unique natural experiment and a local randomization regression discontinuity approach to estimate the effect of polls on turnout intention. We found that the release of a poll increases turnout intention by 5%. This effect is robust to a number of falsification tests of predetermined covariates, placebo outcomes, and changes in the time window selected to estimate the effect. The letter discusses the advantages of the local randomization approach over the standard continuity-based design to study important cases in political science where the running variable is discrete; a method that may expand the range of empirical topics that can be analyzed using regression discontinuity methods.

Analysis of the effect of serious illness medical insurance on relieving the economic burden of rural residents in China: a case study in Jinzhai County

BackgroundIn 2003, China established a New Rural Cooperative Medical System (NRCMS) for rural residents to alleviate the burden of medical expenses among rural residents. However, its reimbursement for high medical costs was insufficient. Therefore, China gradually established the Serious Illness Insurance System (SIMIS) based on NRCMS. After receiving payment through NRCMS, patients in rural areas who met the requirements of SIMIS policy would receive a second payment for their high medical expenses. This study aimed to analyze the effect of the implementation of SIMIS on alleviating the economic burden of rural residents in Jinzhai County.MethodsThe study used the inpatient reimbursement data of NRCMS in Jinzhai County, Anhui Province, from 2013 to 2016. We adopted descriptive and regression discontinuity (RD) methods to analyze the payment effect of SIMIS. The RD analysis targeted patients (n = 7353) whose annual serious illness expenses were between CNY 10,000 (1414 USD) and CNY 30,000 (4242 USD), whereas the descriptive analysis was used for data of the patients compensated by SIMIS (n = 2720).ResultsThe results of RD showed that the actual medical insurance payment proportion increased by about 2.5% (lwald = 0.025, P < 0.01), inside medical insurance self-payment proportion increased by about 2% (lwald = 0.020, P < 0.10), and outside medical insurance self-payment proportion decreased by about 1.6% (lwald = − 0.016, P < 0.05). The descriptive results showed that patients with serious illnesses mostly chose to go to a hospital outside the county. The annual average number of hospitalizations was 3.64. The reimbursement mainly came from the NRCMS. The payment amount of SIMIS was relatively small, and the out-of-pocket medical expenses were still high.ConclusionThe medical technology level of Jinzhai County could not meet the needs of patients with seriously illnesses, the number of beneficiaries of SIMIS was small, and the ability to relieve the burden of medical expenses of the rural residents was insufficient. The high out-of-pocket expenses increased the possibility that only people with good economic conditions could benefit from the reimbursement of SIMIS, resulting in inequity.

Mobilize for Our Lives? School Shootings and Democratic Accountability in U.S. Elections

Gun violence is a large and growing problem in the United States. Many reformers look towards elections to spur policy change in this area. In this paper, we explore the effects of school shootings on electoral mobilization and election outcomes. We pair data from several sources that measure validated voter registration; validated voter turnout; and the electoral performance of officials at the local, state, and federal levels with regression discontinuity and panel methods. Our effects show that shootings have little to no effect on electoral outcomes in the United States. Our work demonstrates that even when tragic events occur that are squarely in the realm of elected officials’ responsibility, have high levels of issue salience, are highly-covered by the media, draw citizens’ attention, and (perhaps) shift public opinion, these seemingly favorable conditions may not be enough to elicit democratic accountability.

Diversion in the Criminal Justice System

AbstractThis article provides the first causal estimates on the popular, cost-saving practice of diversion in the criminal justice system, an intervention that provides offenders with a second chance to avoid a criminal record. We exploit two natural experiments in Harris County, Texas where first-time felony defendants faced abrupt changes in the probability of diversion. Using administrative data and regression discontinuity methods, we find robust evidence across both experiments that diversion cuts reoffending rates in half and grows quarterly employment rates by nearly 50% over 10 years. The change in trajectory persists even 20 years out and is concentrated among young black men. An investigation of mechanisms strongly suggests that stigma associated with a felony conviction plays a key role in generating these results. Other possible mechanisms including changes in incarceration, other universal adjustments in policy or practice, and differences in criminal processing are ruled out empirically.

Female Education, Fertility Decisions, and Infant Health: Evidence from China's Compulsory Education Law

This is the first study to analyze the causal effect of female education on fertility and infant health in China. I use Regression Discontinuity and Instrumental Variable methods and study the national Compulsory Education Law, which created a natural entrance threshold date that results in an exogenous variation in the timing of females’ primary school entry. Using data from the China Family Panel Studies, I find that on average, females born right after the threshold date enter school 0.25 years older in age and obtain 0.47 more completed years of schooling compared to their counterparts born right before the threshold. Starting school one year later reduces infant low birth weight by about 8 percent. However, it does not have a significant effect on fertility outcomes, such as the probability of becoming mother and age at first birth.

Student achievement against national minimum standards for reading and numeracy in Years 3, 5, 7 and 9: A regression discontinuity analysis

AbstractLiteracy and numeracy are essential for success in life. In Australia, students’ academic achievement is monitored using the National Assessment Program – Literacy and Numeracy (NAPLAN). One of the aims of NAPLAN is that results will drive student improvement by identifying which students are underperforming against national minimum standards and who require additional learning support. This study investigated the performance trajectories of students who were identified as “below the national minimum standards”, relative to students above the national minimum standards. This study used standardised reading and numeracy test scores for Australian students and regression discontinuity methods to investigate whether students who performed below the national minimum standards showed improved performance in subsequent years compared with similar students above the national minimum standards threshold. On average, achievement gaps between students below the standards and students at or above the standards persisted across year levels. Labelling students below the national minimum standards had no significant effects on later reading or numeracy achievement. The results suggest the NAPLAN national minimum standards labelling is not driving student improvement, highlighting the need for effective early identification and interventions for children who do not meet school year benchmarks for literacy and numeracy.

Political accountability and public service delivery in decentralized Indonesia: Incumbency advantage and the performance of second term mayors

To what extent do voters hold local elected leaders accountable for public service delivery in fiscally and politically decentralized environments, as theory suggests should be the case? We examine political accountability and service delivery in subnational Indonesia, through the lens of mayoral incumbency advantage. We apply regression discontinuity methods to a unique data set on local elections to identify the causal impact of incumbency on election outcomes and relate those effects to changes in citizen access to local public services. We find that voters in Indonesia are, in general, very willing to return incumbents to office compared to their counterparts in other countries. We also determine that the incumbent advantage is conditional on advances in local service provision: as service access expands more quickly, voters are more likely to vote incumbents back into office. Finally, we find that electorally successful incumbents—second term mayors—spend substantially less on education and health and more on infrastructure, manage their budgets less prudently, and deliver public services neither more nor less effectively than their first term equivalents. We conjecture that term limits and the attendant lack of electoral incentives leads to the disappointing second-term mayor performance.

Greek Life, Academics, and Earnings

Using records from a large public university, we examine the impact of Greek life on academic performance and salaries. To isolate the causal effect of Greek life, we exploit a university policy prohibiting students from joining a Greek organization during their first semester and a minimum GPA for subsequent eligibility. Regression discontinuity and panel methods reveal that Greek affiliation reduces student grades by 0.1-0.3 standard deviations. Greek effects are largest during the semester of pledging, semesters of increased social activities, and for males. We find no evidence of a Greek salary premium and rule out even modest positive effects.

Assessing the causal impact of compulsory schooling policy in Indonesia

Having achieved near universal primary education by the late 1980s, Indonesia turned its attention to increasing secondary school enrolments. In this context, we examine the causal impact of Indonesia’s 1994 compulsory junior secondary schooling initiative. We apply regression discontinuity methods to a sample of about 20,000 individuals. We investigate both intent-to-treat and treatment-on-the-treated effects. We subject our empirical examination to an exhaustive battery of robustness tests. We find no evidence that the compulsory schooling program increased educational attainment. The apparent lack of impact is a unique result across studies that have examined such reforms. We discuss some limitations in government implementation that may underlie program ineffectiveness and acknowledge possible shortcomings in our analysis that may drive the estimated null effects.

Teenage Marriage and Motherhood in Vietnam: The Negative Effects of Starting School Early

This paper examines the impact of school starting age on teenage marriage and motherhood in Vietnam. The investigation uses data from the 2009 Vietnam Population and Housing Census and regression discontinuity methods to identify the causal effects of school starting age on teenage marriage—a first in the literature—and early motherhood. Results show that girls who start school earlier are more likely to marry and/or give birth as teenagers than their counterparts who begin school later. School starting age impacts are heterogeneous across girl subgroups. The deleterious effects of starting school early are particularly pronounced for teenage girls who are members of ethnic minorities, whose mothers have relatively less education, and whose households are relatively poor. Girls that fall into these subgroups are more likely to benefit from a delay in school entry. Government may wish to adopt a more flexible approach to its age at entry regulations to allow disadvantaged girls the option of starting school later.

The Microeconomic Impacts of Employee Representatives: Evidence from Membership Thresholds

Employee representatives in firms are a potentially key but not yet studied source of the impact of unions and works councils. Their actions can shape multiple drivers of firm performance, including collective bargaining, strikes, and training. This article examines the impact of union representative mandates by exploiting legal membership thresholds present in many countries. In the case of Portugal, which we examine here, while firms employing up to forty‐nine union members are required to have one union representative; this increases to two (three) union reps for firms with fifty to ninety‐nine (100–199) union members. Drawing on matched employer–employee data on the unionized sector and regression discontinuity methods, we find that a one percentage point increase in the legal union representative/members ratio leads to an increase in firm performance of at least 7 percent. This result generally holds across multiple dimensions of firm performance and appears to be driven by increased training. However, we find no effects of union representatives on firm‐level wages, given the predominance of sectoral collective bargaining.

The Effect of Pension Growth on the Labor Force Participation of Pensioners in Russia

This paper examines the impact of the 34% increase in pensions in Russia at the end of 2009 and the beginning of 2010 on the labor market participation of pensioners. Several particular features of the pension system in Russia allow us to estimate the net effect of income from such a reform. For evaluation, we used a method combining difference-in-difference and regression discontinuity methods. The results showed that real pension growth by a third reduced labor force participation rate by 6–7.1% for men and by 6–6.4% for women. The heterogeneity of the impact of this reform was also investigated. Estimates showed that the effect was lower for more educated people or those living in villages and was completely absent among those who rated their health as poor or very bad.