Figure 2 - uploaded by Luke John Keele
Content may be subject to copyright.
Milwaukee County with Geographic Discontinuity Based on Milwaukee City Limit Note: Areas outside Milwaukee County did not have any initiatives on the ballot.
Source publication
Political scientists often turn to natural experiments to draw causal inferences with observational data. Recently, the regression
discontinuity design (RD) has become a popular type of natural experiment due to its relatively weak assumptions. We study
a special type of regression discontinuity design where the discontinuity in treatment assignmen...
Similar publications
The Merriam-Webster dictionary defines insurgent as “one who acts contrary to the policies and decisions of one’s own political party.” In this paper, an “insurgent presidential campaign” is defined as the campaign of a candidate who did not have the support of the United States of America’s (U.S.) Democratic Party establishment.
A “populist campa...
This article is based on data from the 1980 National Election Study surveys. It reports findings concerning the rates at which voters become familiar with presidential candidates and their policy positions, trends in public opinion during the 1980 presidential campaign, and the dynamics of individual attitudes that underlie those trends. The impact...
This paper evaluates the two most common methods of measuring voter time-of-decision—the recall method and the panel method—and asks whether the two methods are consistent with each other. Using data from the National Annenberg Election Survey collected during the 2008 U.S. presidential election, the findings suggest that these two methods measure...
Growing racial, ideological and cultural polarization within the American electorate contributed to the shocking victory of Donald Trump in the 2016 presidential election. Using data from American National Election Studies surveys, we show that Trump's unusually explicit appeals to racial and ethnic resentment attracted strong support from white wo...
Using 2012 and 2016 candidate disbursements and independent expenditures filing data collected by the Federal Election Commission, this study compares how much money both the presidential campaigns and outside groups spent on digital relative to traditional media. Digital campaign expenditures have increased their share of total media outlays marke...
Citations
... Geographic regression discontinuity design (GRDD) is an extended form of regression discontinuity design (RDD) [33], initially proposed by Hahn et al. (2001) [34], and further developed with a more robust theoretical foundation by Imbens and Lemieux (2008) [35] as well as [36]. This approach leverages discontinuities in geographical boundaries or spatial distributions to estimate causal effects. ...
... Rischard et al. (2021) utilized GRDD to simulate spatial structures and explore the impact of school districts on housing prices in New York City [41]. applied the GRDD to re-examine the effects of political advertisements on voter turnout during a presidential campaign, leveraging exogenous variations in the volume of presidential ads created by media market boundaries [36]. D'Arcangelo and Percoco (2015) employed the GRDD method, controlling for dynamic differences in geographic variables, to overcome some limitations of simple discontinuity designs, and studied the impact of road pricing schemes (Ecopass) on the housing market from the perspective of rent changes [42]. ...
Regional integration plays a pivotal role in the socio-economic advancement of various global regions and is closely linked with the expansion of construction land. This expansion is a major contributor to urban carbon emissions. Utilizing a geographical regression discontinuity design (GRDD), this paper estimates the impact of urban construction land expansion on carbon emissions and explores the underlying mechanisms within the regional integration process of the Yangtze River Delta (YRD), China. The findings reveal that urban construction land expansion significantly influences carbon emissions, displaying an inverted “U”-shaped pattern. Furthermore, this expansion affects carbon emissions through the transformation of industrial structures, shifts in consumption patterns, and enhancements in scientific and technological investments. Our findings span the performance of the Yangtze River Delta from its early development stages to a relatively mature phase. This paper also partially reveals how the Yangtze River Delta, with both megacities and large- to medium-sized cities, manages urban construction land expansion during the integration process and strives for low-carbon emissions reduction. These results can provide green growth recommendations that balance socio-economic development, low-carbon emissions, and social equity not only for other urban agglomerations in China but also for similar regions in other developing countries by altering construction land utilization patterns.
... 10 Since only 10% of homicides (5,470 out of 50,158) are not registered in our dataset, the precision gained in our measures by using this unit of analysis outweighs the costs of dropping observations for these missing localities. 11 These are homicides that exclusively occurred as result of confrontations among DTOs or within a DTO. 12 All distances were calculated with the sf package in R. the earth is an oblate spheroid (see Keele and Titiunik 2015). All distances are coded in kilometers and logged to account for differences in scale. ...
This article helps understand why locations close to strategic infrastructure to transport illegal drugs (seaports, airports, highways, and US ports of entry along the Mexico-US border) or to increase income (pipelines) experience different levels of violence due to DTOs operations. Our theory breaks down the impact of the geographical distance to these facilities on violence into two effects. The first effect is produced by the level of (violent) competition among DTOs, measured by the number of DTOs employing violence. We report that greater proximity to the U.S. ports of entry along the Mexico-US border, ports, and airports furthers the number of competitors, and such increase boosts violence. The second effect shapes the intensity of competition among DTOs. Reductions in the costs of excluding competing DTOs from using the facility could trigger greater confrontation among DTOs. We confirm the importance of this second effect in relation to ports and the U.S. ports of entry along the Mexico-US border.
... If a voter has no network members running for office inside the district boundary, we use distance to the geographically closest network member outside the district. Our identification strategy is related to the geographic regression discontinuity design, where a geographic or administrative boundary splits the units into treatment and control (Keele and Titiunik 2015). Examples include Black (1999), who leveraged school district boundaries to estimate parents' willingness to pay for good schools, and Huber and Arceneaux (2007), who compared same-state voters in different media markets to study the effects of advertising. ...
A vast and growing quantitative literature considers how social networks shape political mobilization but the degree to which turnout decisions are strategic remains ambiguous. Unlike previous studies, we establish personal links between voters and candidates and exploit discontinuous incentives to mobilize across district boundaries to estimate causal effects. Considering three types of networks – families, co-workers, and immigrant communities – we show that a group member's candidacy acts as a mobilizational impulse propagating through the group's network. In family networks, some of this impulse is non-strategic, surviving past district boundaries. However, the bulk of family mobilization is bound by the candidate's district boundary, as is the entirety of the mobilizational effects in the other networks.
... Standard RD estimation methods that include two running variables can be applied to geographic RD designs with minor adjustments. For further information on geographic RD, refer to Keele and Titiunik (2015). ...
Regression discontinuity (RD) designs have gained significant popularity as a quasi-experimental device for evaluating education programs and policies. In this paper, we present a comprehensive review of RD designs, focusing on the continuity-based framework, the most widely adopted RD framework. We first review the fundamental aspects of RD designs, drawing on potential outcomes and causal graphs. We then discuss the validity threats in RD designs, including manipulation, discreteness of the running variable, statistical power, and generalizability. Additionally, we provide an overview of the existing extensions to RD designs. To exemplify the application of RD methods, we analyze the effect of New Jersey’s pre-kindergarten program on children’s vocabulary test scores, using an educational dataset. Finally, we offer practical guidelines in the conclusion to promote the appropriate use of RD methods in educational research.
... The intuition is that households or individuals very close to each other but separated by this threshold are similar in observable and unobservable characteristics. In a particular type of RD study, the discontinuity in treatment is geographic -often an administrative boundary that divides individuals into two groups (Keele & Titiunik, 2015). Our study began by verifying the existence of marked geographic discontinuities in piped water provision in the peri-urban areas of metropolitan Lima. ...
This paper presents the results of a quasi-experimental study using information collected through a survey conducted in peri-urban areas of metropolitan Lima and applied to households residing close to and on both sides of the geographic boundary of piped water supply. We find that access to piped water was associated with a reduction in the probability of contracting COVID-19. Furthermore, the results suggest that a minimum consumption level must be guaranteed to lower infection probability. These results highlight the need for investment in infrastructure to close access gaps and the importance of ensuring quality and affordable services for the population.
... Therefore, it is necessary to construct an experiment in which treatments are assigned to all municipalities according to a completely randomized design (Figure 2), where, in other terms, each municipality has the same marginal probability of receiving treatment. Few papers in the literature have exploited space to construct a random discontinuity except for a few in which there were administrative or geographical boundaries that justified the presence of a border (see, for example, Spatial Regression Discontinuity Designs as in Keele and Titiunik, 2015;Keele et al., 2017). ...
The provision of health services plays a central role in the promotion of public welfare and maintaining a healthy society. However, inequalities in access to health care persist between countries, regions, and communities, reflecting the complex interplay of many social, economic, and cultural factors. This article aims to explore the complex relationship between tax deductions and the spatial correlation between health demand and private supply. By using an original data set at an unprecedented municipal level and employing a spatial counterfactual empirical strategy, we contribute to the existing literature in two ways. First, we show a strong territorial heterogeneity of health tax expenditure at the municipal level, which persists even after controlling for explanatory factors, and essentially rewards northern Italian regions and penalizes southern ones. Second, we investigate whether greater proximity to a private health provider respects a public one produces a different spending behavior in citizens, highlighting once again the specificity of private healthcare provision in the Italian context. This behavior was analyzed with a geographically weighted analysis, which allowed us to assess the strong spatial non-stationarity by including local potential hidden confounders.
... For our purpose, a relevant extension pertains to cases in which eligibility depends upon more than one running variable, or simply scores, the so called Multiple Regression Discontinuity (MRD), in which typically, but not necessarily, the scores are two. Examples are given by borders identified by latitude and longitude (Keele and Titiunik 2015;Dell 2010;Gerber et al. 2011) or test scores at school (Jacob and Lefgren 2004;Papay et al. 2011;Zajonc 2012;Clark and Martorell 2014) and multiple requirements for eligibility in an income support program (Frey 2019). A common practice to deal with such occurrences consists in reducing the MRD problem to a single score problem. ...
This paper investigates the causal effect of co-payment exemption on the number of specialist visits in the Italian National Health System. Exploiting a discontinuity in the multiple eligibility criteria, we apply multiple regression discontinuity in a quasi-experimental setting, considering both age and income requirements. Differently from the standard regression discontinuity, this twofold discontinuity allows to identify the effect of co-payment on a particularly needy sub-population of less wealthy people and how it changes according to the eligibility criteria. We find positive effects of co-payment exemption and the effects are stronger for less wealthy and older individuals. The result may be useful to the policy maker to tailor ad-hoc policies aimed at disadvantaged sub-populations.
... The fixed effect (time-invariant variables) was based on the subdistrict level. Subdistrict boundaries were chosen as being more consistent and reflective of the policy outcome, as noted by Keele and Titiunik (2015). (For details of map source, please find in A ppendix D). ...
... The evidence is suggestive of a possible positive associ-835 ation between the presence of geminophilous norms and the incidence of twinning, but the coarse-grained (i.e., country-level) nature of these data is not optimal. Ideally, both anthropological (i.e., twin-related norms) and epidemiological (i.e., twinning rates) information should 840 be aggregated at the smallest geographic unit possible, and analysed with robust tools (e.g., regression discontinuity designs: Keele and Titiunik, 2015). ...
Natural selection should favor litter sizes that optimize trade-offs between brood-size and offspring viability. Across the primate order, modal litter size is one, suggesting a deep history of selection favoring minimal litters. Humans, however---despite having the longest juvenile period and slowest life-history of all primates---still produce twin-births at appreciable rates, even though such births are costly. This presents an evolutionary puzzle. Why is twinning still expressed in humans despite its cost? More puzzling still, is the discordance between the principal explanations for human twinning and extant empirical data. Such explanations propose that twinning is regulated by phenotypic plasticity in polyovulation, permitting production of larger sib-sets if-and-when resources are abundant. However, comparative data suggest that twinning rates are actually highest in poorer countries and lowest in developed economies. We propose that a historical dynamic of gene-culture coevolution might explain this geographic patterning. Our explanation distinguishes geminophilous and geminophobic cultural contexts, as those celebrating twins (e.g., through material support) and those hostile to twins (e.g., through sanction of twin-infanticide). Geminophilous institutions, in particular, may buffer the fitness cost associated with twinning, potentially reducing selection pressures against polyovulation. We conclude by synthesizing a mathematical and empirical research program that might test our ideas.
... However, incorporating causal modeling in a spatial setting potentially violates the no interference assumption (SUTVA) as discussed at length in Keele and Titiunik (2015) and Noreen (2018). In investigating the effect of convenience voting and voter turnout, Keele and Titiunik (2015) are concerned about interference and spillover effects -the units (individuals) may be influenced due to proximity of geographical regions (or influenced at the workplace or by their social network, etc.). ...
... However, incorporating causal modeling in a spatial setting potentially violates the no interference assumption (SUTVA) as discussed at length in Keele and Titiunik (2015) and Noreen (2018). In investigating the effect of convenience voting and voter turnout, Keele and Titiunik (2015) are concerned about interference and spillover effects -the units (individuals) may be influenced due to proximity of geographical regions (or influenced at the workplace or by their social network, etc.). ...
This research develops a model-based LAtent Causal Socioeconomic Health (LACSH) index at the national level. Motivated by the need for a holistic national well-being index, we build upon the latent health factor index (LHFI) approach that has been used to assess the unobservable ecological/ecosystem health. LHFI integratively models the relationship between metrics, latent health, and covariates that drive the notion of health. In this paper, the LHFI structure is integrated with spatial modeling and statistical causal modeling. Our efforts are focused on developing the integrated framework to facilitate the understanding of how an observational continuous variable might have causally affected a latent trait that exhibits spatial correlation. A novel visualization technique to evaluate covariate balance is also introduced for the case of a continuous policy (treatment) variable. Our resulting LACSH framework and visualization tool are illustrated through two global case studies on national socioeconomic health (latent trait), each with various metrics and covariates pertaining to different aspects of societal health, and the treatment variable being mandatory maternity leave days and government expenditure on healthcare, respectively. We validate our model by two simulation studies. All approaches are structured in a Bayesian hierarchical framework and results are obtained by Markov chain Monte Carlo techniques.
