Why do states abide by the terms of international human rights treaties? In this interesting and informative book, Audrey L. Comstock argues that part of the answer lies in the way states commit to those treaties in the first place. She distinguishes between four commitment mechanisms: signature, ratification, accession, and succession. Signature is nonbinding; although it is only an “initial commitment step” (p. 17), it can nonetheless serve as a “focal point” around which human rights activists and NGOs mobilize and can induce “advocate executives” to promote human rights prior to ratification (p. 80). Ratification, accession, and succession are all legally binding but, in Comstock’s account, involve different groups of states: those that participated in negotiations over the treaty (ratification), those that did not participate in negotiations but wish to bind themselves anyway (accession), and those that previously ratified a treaty before regime change brought a new state into being (succession).

Comstock argues that these divergent commitment pathways produce equally disparate human rights records. Her theory is complex and counterintuitive. She hypothesizes that human rights will improve after a state signs a human rights treaty, but only if the state requires legislative approval for ratification; in states where the executive can ratify unilaterally, Comstock expects signature to yield little or no change. In these latter “executive approval” states, Comstock hypothesizes that human rights will improve after ratification; in contrast, in states that require legislative approval, she expects no change after ratification or even a deterioration in human rights. She expects human rights to improve after accession, but only among states that were not members of the UN at the time the treaty was negotiated; among states that were members but opted out of negotiations, she again expects human rights to plateau or decline. The only commitment mechanism that she expects to have unconditionally positive effects is succession—although, as she notes, this is an “infrequent and unique” pathway that is relevant to only a “small handful of states” (p. 55).

Comstock’s account offers both theoretical and methodological advances beyond existing human rights scholarship, most of which focuses on ratification. Previous studies have typically assumed that ratification constitutes no more or less of a commitment than accession and that signature is no commitment at all. Comstock argues that this assumption is misguided. She provides theoretical reasons to believe that even an act as (apparently) ceremonial as signing a treaty with no legally binding consequences can nonetheless precipitate significant changes in states’ human rights behavior; she also provides empirical evidence to suggest that different commitment pathways do indeed correlate with different human rights outcomes. Treaty law has become the dominant form of international law since the end of World War II, and international human rights treaties are among the most prominent mechanisms for disseminating human rights protections around the globe. Understanding why and under what conditions states adopt these protections is a first-order concern for anyone interested in international law, international organizations, or the rights of historically marginalized groups.

These strengths notwithstanding, the book does have some limitations. Perhaps the most important one is the difficulty of isolating the impact of human rights treaties from the conditions that lead states to commit to those treaties in the first place. Does committing to a treaty cause states to improve their human rights practices? Or do states commit because they are already invested in improving human rights and wish to signal their investment to domestic and international stakeholders? Disentangling these possibilities requires overcoming enormous inferential obstacles. Consider, for example, the United States’ commitment to the Committee on the Elimination of Racial Discrimination (CERD) treaty, which Comstock discusses in chapter 4. The United States is a legislative approval state; it signed CERD in 1966 but did not ratify it until 1994. According to Comstock, this is an example of a state “signing a human rights treaty early, moving toward improved rights, and then ratifying much later” (p. 91). The implication seems to be that there is a causal connection between the decision to sign CERD and the United States’ subsequent successes (limited and halting though they may be) in mitigating racial discrimination.

But is this the most plausible interpretation of the historical record? By the time President Lyndon Johnson signed CERD in 1966, the civil rights movement was already in full swing. The Civil Rights Act passed in 1957 and the Voting Rights Act in 1964. Johnson mandated equal opportunity for minorities in federal contracting in 1965. Indeed, in Comstock’s own telling, Johnson signed CERD to “signal US commitment” to the treaty’s provisions and strengthen the United States’ position in negotiations over other international human rights laws (p. 91). Did Johnson sign CERD because he was already pursuing a civil rights agenda and wanted to signal his commitment to fulfilling it (a selection effect)? Or did signing CERD catalyze improvements in civil rights that would not have materialized otherwise (a treatment effect)? It’s hard to say for certain, and both dynamics could be at play. But the former seems more likely than the latter.

Or take the case of Nigeria and the Convention on the Rights of Persons with Disabilities (CRPD), also discussed in chapter 4. Nigeria—a legislative approval state—signed the CRPD in 2007 and ratified it in 2010. Per Comstock’s theory, we should expect to observe an improvement in Nigeria’s human rights record after 2007, followed by stagnation or deterioration after 2010. In reality, however, Nigeria’s score on the Freedom House civil liberties index (one of Comstock’s key dependent variables) remained constant between 2007 and 2010 (and beyond). Nigeria’s CIRI scores (the other key dependent variable) actually fell between 2007 and 2010: of the 16 CIRI indicators that are available for this period, Nigeria’s score improved on just one (women’s political rights) and by just one point. Comstock notes that Nigeria created a ministerial committee on albinism but only after it signed and ratified the CRPD. Similarly, the state of Lagos passed a law aimed at securing the rights of people with disabilities, but only after CRPD ratification. It is possible, of course, that activists’ efforts take time to bear fruit, but these temporal dynamics only compound the inferential problems. If a state signs a treaty in one year, ratifies it a few years later, and improves its human rights record a few years after that, it is very hard to tell what explains the improvement: signature, ratification, or something else entirely.

Ambiguity between selection and treatment effects is implicit in the hypotheses as well. Comstock argues that signature can “activate human rights mechanisms” as activists, NGOs, and other states pressure, shame, and socialize governments toward ratification (p. 17). This is a treatment effect. But signature also “signifies a state’s willingness to recognize and support international standards of human rights behavior” (p. 17). This is a selection effect. Similarly, Comstock explains that negotiation is a lengthy and sometimes contentious process that leaves states “more invested and likely to adhere to the treaty’s terms” (p. 18). This is a treatment effect. But she also notes that “states that are at the negotiating table for human rights treaties have an interest in promoting international human rights” (pp. 122–23). This is a selection effect. Are treatment or selection effects more important for explaining states’ subsequent human rights records? Again, it’s hard to say. But the latter seem at least as important as the former.

To her credit, Comstock does address the possibility of selection effects in her chapter on accession. But the discussion is brief and a bit too sanguine. Indeed, she devotes parts of chapters 4 and 5 to documenting systematic differences between states that sign human rights treaties and those that do not; between states that sign and ratify and those that do not; and between states that ratify and those that accede. These differences only magnify potential selection concerns. Comstock also uses instrumental variables (IV) to attempt to mitigate these concerns, but the book is surprisingly silent on the logic underlying the instruments and the (quite demanding) assumptions that using IV requires. For example, Comstock uses an indicator for common law states as an instrument for committing to the International Covenant on Civil and Political Rights (ICCPR). But other studies that Comstock cites (e.g., Sara McLaughlin Mitchell, Jonathan J. Ring, and Mary K. Spellman, “Domestic Legal Traditions and States’ Human Rights Practices,” Journal of Peace Research, 50 [2], 2013) have shown that common law states have better human rights records for a variety of reasons that are entirely independent of their decision to sign a particular treaty. This seems like a clear violation of the excludability assumption needed for consistent IV estimation.

These limitations notwithstanding, Comstock’s book offers an important corrective to a literature focused too narrowly on ratification, and it provides valuable insights to understand how even a legally nonbinding action like signature can induce meaningful behavioral change. Although the book will be of particular interest to human rights scholars, these latter insights are potentially applicable to a wide range of questions in international relations and political science more broadly. It is not uncommon for states to comply with international laws, norms, and customs to which they are not legally bound. Comstock’s argument helps explain why.