The Informative Power of Treaty Commitment: Using the Spatial Model to Address Selection Effects

Abstract

The effects of international institutions on state behavior make up a key research agenda in international-relations scholarship. Because states self-select into treaties, we cannot infer that these commitments have causal effects unless we address this selection effect. I explain the significant limitations of the methods used thus far to overcome this problem and argue that a more effective approach must take into account states' treaty preferences. I describe a novel combination of ideal-point estimation and propensity-score matching that can estimate the probabilities of treaty commitment and use them to test hypotheses. I use this procedure to test the effects of three key international human-rights treaties. My results provide significant new findings regarding the effects of these important agreements. I show that the Convention on the Elimination of All Forms of Discrimination against Women has significantly improved respect for women's rights, but that the Convention against Torture and the International Covenant on Civil and Political Rights have not had significant effects on human rights.

Human Rights Texts: Converting Human Rights Primary Source Documents into Data

“…It is our hope that researchers from all different fields will use a variety of statistical analyses and machine learning algorithms to better understand the content of these documents and how this content has evolved over time and in response to political and reporting changes. More generally, these datasets, coupled with the existing human coding schemes, automated coding algorithms, and innovative new research designs [20,22,[64][65][66], will allow scholars to more thoroughly analyze reports of human rights abuse and therefore extend and contribute to the growing human rights literature.…”

“…Unsupervised statistical learning tools also exist, which are useful for revealing other patterns within the human rights document corpus [7,[13][14][15][16] without reference to the existing coded human rights variables, which we describe below. These tools are more generally part of the emergent field of computational social science or "big data" analysis [17][18][19] of which there are several recent examples in the study of human rights [1,10,[20][21][22][23] and many other examples from political science and social science more generally [11,[24][25][26][27][28].…”

Flexible Causal Inference for Political Science

“…It is perhaps not surprising, then, that practitioners have turned to matching methods to estimate causal effects (Simmons and Hopkins, 2005;Gilligan and Sergenti, 2008;Hill, 2010;Lupu, 2013).…”

“…Foremost among them is the assumption that all confounders have been measured and incorporated into the analysis-an assumption known as "selection on observables." To the extent that this assumption has been recognized as being potentially problematic, the solution is typically to measure unmeasured confounders and include them in the analysis (see, e.g., Simmons and Hopkins 2005;Lupu 2013). As Keele (2015, 322) notes in a recent review, however, "selection on observables is a very strong assumption.…”

Testing the Value of the Postwar International Order

“…Second, the international human rights regime engages domestic constituencies, laying the foundation for the development of NGOs dedicated to monitoring and improving human rights practices in countries with poor track records. 21 Third, international organiza-17 See, for instance, Cole, 2005;Chayes and Chayes, 1995;Goodliffe and Hawkins, 2006;and Hathaway, 2007. 18 See Lupu, 2013;and Von Stein, 2005. 19 Simmons, Dobbin, and Garrett, 2006, pp. 783-784.…”