Bowers, Jake, Bruce A. Desmarais, Mark Frederickson, Nahomi Ichino, Hsuan-Wei Lee, and Simi Wang. 2018. “Models, Methods and Network Topology: Experimental Design for the Study of Interference,” Social Networks 54:1 96-208.
How should a network experiment be designed to achieve high statistical power? Experimental treatments on networks may spread. Randomizing assignment of treatment to nodes enhances learning about the counterfactual causal effects of a social network experiment and also requires new methodology (ex. Aronow and Samii 2017a; Bowers et al. 2013; Toulis and Kao 2013). In this paper we show that the way in which a treatment propagates across a social network affects the statistical power of an experimental design. As such, prior information regarding treatment propagation should be incorporated into the experimental design. Our findings justify reconsideration of standard practice in circumstances where units are presumed to be independent even in simple experiments: information about treatment effects is not maximized when we assign half the units to treatment and half to control. We also present an example in which statistical power depends on the extent to which the network degree of nodes is correlated with treatment assignment probability. We recommend that re- searchers think carefully about the underlying treatment propagation model motivating their study in designing an experiment on a network.
Ahlquist, John, Nahomi Ichino, Jason Wittenberg, and Daniel Ziblatt. 2018. “How Do Voters Perceive Changes to the Rules of the Game? Evidence from the 2014 Hungarian Elections,” Journal of Comparative Economics, 46: 906-19.
Voters often rely on partisan attachments as they evaluate new policy proposals, but does partisanship also color their interpretation of incumbent efforts to entrench themselves in power by changing the “basic rules of the political game”? We explore this question by taking advantage of a rare instance where a single party held a supermajority sufficient to unilaterally amend the constitution and overhaul the electoral system. We embedded a randomized experiment in a panel survey around the 2014 Hungarian elections, providing respondents with different information about recent changes to the Hungarian electoral rules. While respondents were largely pessimistic about the reforms, providing information yielded no significant effects on their views on the election’s legitimacy. But when information was presented alongside partisan cues, respondents became more negative in their views. Subgroup analysis shows that this effect is concentrated entirely among those not supporting the incumbent. Partisan differences in opinion dwarf any treatment effects we were able to induce. We provide evidence that these findings are unlikely the result of a well-informed populace. Rather, we provide the first experimental evidence that partisan-motivated reasoning applies not only to public policy under fixed institutions but also to changes to the institutional rules of a political system. Incumbents can exploit strong partisan attachments to reduce political competition.
Glynn, Adam N., and Nahomi Ichino. 2016. “Increasing Inferential Leverage in the Comparative Method: Placebo Tests in Small-n Research,” Sociological Methods and Research 45(3): 598-629.
We delineate the underlying homogeneity assumption, procedural variants, and implications of the comparative method and distinguish this from Mill’s method of difference. We demonstrate that additional units can provide “placebo” tests for the comparative method even if the scope of inference is limited to the two units under comparison. Moreover, such tests may be available even when these units are the most similar pair of units on the control variables with differing values of the independent variable. Small-n analyses using this method should therefore, at a minimum, clearly define the dependent, independent, and control variables so they may be measured for additional units, and specify how the control variables are weighted in defining similarity between units. When these tasks are too difficult, process tracing of a single unit may be a more appropriate method. We illustrate these points with applications to two studies.
Glynn, Adam N., and Nahomi Ichino. 2015. “Using Qualitative Information to Improve Causal Inference” American Journal of Political Science 59(4): 1055-71. (supplementary information) (R package qualCI)
Using the Rosenbaum (2002, 2009) approach to observational studies, we show how qualita- tive information can be incorporated into quantitative analyses to improve causal inference in three ways. First, we can ameliorate the effects of difficult-to-measure outcomes by including qualitative information on outcomes within matched sets, sometimes reducing p-values. Second, additional information across matched sets enables the construction of qualitative confidence in- tervals on effect size. Third, qualitative information on unmeasured confounders within matched sets reduces the conservativeness of Rosenbaum-style sensitivity analysis. This approach accom- modates small to medium sample sizes in a nonparametric framework, and therefore may be particularly useful for analyses of the effects of institutions in a given set of units. We illus- trate these methods by examining the effect of using plurality rules in transitional presidential elections on opposition harassment in 1990s sub-Saharan Africa.
Ichino, Nahomi, and Noah L. Nathan. 2013. “Crossing the Line: Local Ethnic Geography and Voting in Ghana.” American Political Science Review 107(2): 344-61.
Theories of instrumental ethnic voting in new democracies propose that voters support co-ethnic politicians because they expect politicians to favor their co-ethnics once in office. But many goods that politicians deliver to voters are locally nonexcludable in rural areas, so the local presence of an ethnic group associated with a politician should affect a rural voter’s assessment of how likely she is to benefit from that politician’s election. Using geocoded polling-station–level election results alongside survey data from Ghana, we show that otherwise similar voters are less likely to vote for the party of their own ethnic group, and more likely to support a party associated with another group, when the local ethnic geography favors the other group. This result helps account for the imperfect correlation between ethnicity and vote choice in African democracies. More generally, this demonstrates how local community and geographic contexts can modify the information conveyed by ethnicity and influence voter behavior.
Ichino, Nahomi, and Noah L. Nathan. 2013. “Do Primaries Improve Electoral Performance? Clientelism and Intra-Party Politics in Ghana.” American Journal of Political Science 57(2): 428-41.
We consider the effect of legislative primaries on the electoral performance of political parties in a new democracy. While existing literature suggests that primaries may either hurt a party by selecting extremist candidates or improve performance by selecting high valence candidates or improving a party’s image, these mechanisms may not apply where clientelism is prevalent. A theory of primaries built instead on a logic of clientelism with intra-party conflict suggests different effects of legislative primaries for ruling and opposition parties, as well as spillover effects for presidential elections. Using matching with an original dataset on Ghana, we find evidence of a primary bonus for the opposition party and a primary penalty for the ruling party in the legislative election, while legislative primaries improve performance in the presidential election in some constituencies for both parties.
Ichino, Nahomi, and Noah L. Nathan. 2012. “Primaries on Demand? Intra-Party Politics and Nominations in Ghana.” British Journal of Political Science 42(4): 769-91.
In new democracies, why do political party leaders relinquish power over nominations and allow legislative candidates to be selected by primary elections? Where the legislature is weak and politics is clientelistic, democratization of candidate selection is driven by local party members seeking benefits from primary contestants. Analysis of an original dataset on legislative nominations and political interference by party leaders for the 2004 and 2008 elections in Ghana shows that primaries are more common where nominations attract more aspirants and where the party is more likely to win, counter to predictions in the existing literature. Moreover, the analysis shows that party leaders interfere in primaries in a pattern consistent with anticipation of party members’ reactions.
Ichino, Nahomi, and Matthias Schündeln. 2012. “Deterring or Displacing Electoral Irregularities? Spillover Effects of Observers in a Randomized Field Experiment in Ghana.” Journal of Politics 74(1): 292-307.
This article studies the effect of domestic observers deployed to reduce irregularities in voter registration in a new democracy, and in particular, the response of political parties’ agents to these observers. Because political parties operate over large areas and party agents may relocate away from observed registration centers, observers may displace rather than deter irregularities. We design and implement a large-scale two-level randomized field experiment in Ghana in 2008 taking into account these spillovers and find evidence for substantial irregularities: the registration increase is smaller in constituencies with observers; within these constituencies with observers, the increase is about one-sixth smaller on average in electoral areas with observers than in those without; but some of the deterred registrations appear to be displaced to nearby electoral areas. The finding of positive spillovers has implications for the measurement of electoral irregularities or analysis of data collected by observers.
Kashin, Konstantin, Adam N. Glynn, and Nahomi Ichino. 2014. qualCI v0.1. R package available from CRAN.
Ichino, Nahomi and Noah L. Nathan. 2018. “Primary Elections in New Democracies: The Evolution of Candidate Selection Methods in Ghana,” in Routledge Handbook of Primary Elections, Robert Boatright, ed. pp. 369-83.
Primary elections in advanced democracies are usually conceptualized as institutions that generate nominees that have higher valence – better campaigning skills and popularity within the party – at the potential cost of being more ideologically extreme. But spatial models of primary elections are inappro- priate for new democracies, where there is often little ideological competition in elections. We identify key imperatives that shape the decisions of party leaders in new democracies when choosing among dif- ferent candidate selection mechanisms: the need to prevent elite defections and to motivate grassroots activists. We explore this argument in Ghana, a new democracy in which both major parties have grad- ually adopted and adapted primaries to select legislative candidates. Ghana’s experience with primaries highlights the central role that candidate selection institutions play in the development of political parties in new democracies.
Glynn, Adam N. and Nahomi Ichino. “Generalized Nonlinear Difference-in-Difference-in-Differences.” August 2018 version.
Difference-in-difference-in-differences (DiDiD) allow for the correction of unmeasured confounding and function as a robustness check for difference-in-differences (DiD) techniques. However, this technique is not scale invariant and requires that the outcome variable be measured on units for which the treatment could have had no effect in either the pre-treatment or post-treatment periods. Athey and Imbens (2006) provides a scale invariant, nonlinear DiD approach known as Changes-in-Changes (CiC). Sofer et al. (2016) extends CiC by showing that pre-treatment outcome measures are a special case of placebo (negative) outcomes and proposes a generalization of CiC called Negative Outcome Control (NOC). We develop a generalized nonlinear DiDiD approach we call NOCNOC that can be used either in the traditional DiDiD setting or when a placebo outcome is available in the pre and post-treatment data. We show that NOCNOC can correct for bias in DiDiD, CiC, and NOC. We apply this method to a study of whether exposure to candidate debates affected Nepalese citizens’ sense of political efficacy.
Glynn, Adam N. and Nahomi Ichino. “Nonlinear Difference-in-Differences and Difference-in-Difference-in-Differences with Placebo and Surrogate Outcomes.” July 2018 version.
Difference-in-differences (DiD) and difference-in-difference-in-differences (DiDiD) allow for the correction of unmeasured confounding. However, these techniques require that the same variable be measured prior to treatment or on units for which the treatment could have had no effect. Athey and Imbens (2006) provides a quantile-based DiD approach which Sofer et al. (2016) extends to allow the use of placebo (negative) outcomes in lieu of pre-treatment measurements. We extend these approaches in three ways. First, we demonstrate the use of placebo outcomes in the DiD context for the estimation of the average treatment effect on the controls. Second, we show that surrogate outcomes can be used in an analogous but reverse manner to placebo outcomes. Third, we show that these quantile-based techniques can be implemented in the DiDiD context. We apply these methods to a study of whether exposure to candidate debates affected Nepalese citizens’ sense of political efficacy.
Ichino, Nahomi and Noah L. Nathan. “Democratizing the Party: The Effects of Primary Election Reforms in Ghana.” September 2018 version.
Recent reforms to primary election rules by one of Ghana’s major parties offer a rare opportunity to assess the effects of expansions of the franchise in contemporary new democracies, where universal suffrage was usually already established at independence. We propose that democratizing candidate selection by expanding the primary electorate has two consequences in patronage-oriented political systems: the electorate will have more diverse preferences and vote buying will become a less effective strategy. These changes, in turn, affect the types of politicians who seek and win legislative nominations. Using an original dataset on candidate entry and nominations, we show that expanding the primary electorate opened paths to office for politicians from social groups that were previously excluded, including women and members of ethnic groups outside a party’s core national coalition.
Faller, Julie K., Adam N. Glynn, and Nahomi Ichino. “Electoral Systems and Corruption.” May 2014 version.
What is the effect of electoral systems on corruption? Persson, Tabellini and Trebbi (2003) proposed that plurality electoral systems should lead to lower corruption compared to proportional representation (PR) systems because the former creates a direct link between voters and politicians whom voters can hold accountable for corruption. The empirical question remains unresolved, however, in part due to the endogeneity of the electoral institutions and difficulties in measuring corruption. Using nonparametric methods and new data to reduce sensitivity to these problems, we find no evidence for this hypothesis. Instead, we find some evidence in the opposite direction, that PR leads to less corruption.
Glynn, Adam N., and Nahomi Ichino. “Using Post-Treatment Variables to Establish Upper Bounds on Causal Effects.” May 2013 version.
We propose an adjustment based on post-treatment variables for pair matching estimators of the average treatment effect on the treated. Under relatively weak conditions, this adjusted estimator will provide an upper bound for the effect. Additionally, this approach allows for unbounded outcome variables and multiple mechanisms by which the treatment has an effect on the outcome. We also demonstrate that this adjustment will reduce the estimated effect in a wide variety of circumstances, and therefore, when the assumptions for the adjusted estimator are preferable to the assumptions for the unadjusted estimator, the adjustment can be used as a robustness check.