1The complexities, boundaries, and definitions of morality policy have long intrigued policy scholars. Within the study of policy diffusion, scholars have considered whether and how morality policies spread differently than those with less morality content. Of course, morality content is a dimension, not a simple dichotomy. Victor David Cruz-Aceves’s book, The Diffusion of Morality Policies, is an important step forward for this strain of research. While this book still has the rawness of a dissertation, it is a valuable read for policy diffusion scholars, especially those studying diffusion within the American federal context. Cruz-Aceves is an emerging voice in policy diffusion studies, having published both substantive and methodological work prior (Beer & Cruz-Aceves, 2018; Cruz-Aceves & Mallinson, 2019). This book is a good platform for additional work on how different policy dimensions affect diffusion patterns.
2A driving consideration for Cruz-Aceves is the idea that morality policy is better understood in a continuum rather than as a dichotomous concept. Policies vary in the strength of their morality dimension, for example with content reaching to the core moral values of a population (e.g., abortion, capital punishment, etc.). This often means the policies have high salience, they are highly polarized and contentious, whereas they are not technically complex. Scholars have long recognized that policies vary in their morality content, but Cruz-Aceves argues that the effect of variation in this dimension on policy diffusion is not well understood.
3The book begins with an introduction to its basic argument, which Cruz-Aceves draws from recent research on policy attributes, political representation, interest group politics, and diffusion mechanisms. Essentially, Cruz-Aceves argues that as the morality content of a policy increases, and thus its concomitant technical complexity decreases and salience increases, the effects of external actors on its adoption decline. This has to do with the author’s expectations about policy learning. If policies have strong moral content, there is less uncertainty regarding their political consequences for legislators, and thus less motivation to learn from previously adopting jurisdictions. Drawing on Berry and Berry’s (2018) unified model of policy diffusion, the book also lays out motivations (representation, public opinion, and issue advocacy) and resources/obstacles (issue opponents, problem severity, and other state characteristics) that are expected to affect policy adoption. Thus, the work is well grounded in the main theoretical constructs of policy diffusion research. That said, the theoretical rationale for some individual hypotheses is not always fully explained. Additional theoretical justification would have been helpful for supporting some of them.
4Following his theoretical argument, Cruz-Aceves lays out the rationale behind focusing on three policies: same-sex marriage bans, medical marijuana, and anti-obesity legislation. These, he argues, are on a spectrum with respect to morality content. Same-sex marriage bans have a strong moral dimension. Medical marijuana had traditionally had a moral dimension, but its economic dimension grew significantly in public debates as the policy spread, rendering it a mixed policy. Finally, anti-obesity policies have a weak moral dimension, but much stronger public health and economic dimensions.
5Cruz-Aceves is innovative in his approach to assessing how diffusion patterns vary across these three policies. The analysis begins with event history modeling, which is prevalent in diffusion studies, to examine the internal and external determinants of adoption for each policy separately. However, the analysis does not stop there. A two-stage seemingly unrelated regression (SUR) is used to compare diffusion effects more directly, which allows a comparison of the magnitude of hazard ratios for key diffusion variables such as neighbor adoption and ideological distance from past adopting states. SUR has been used sparingly in policy diffusion research (e.g., Taylor et al., 2012) and Cruz-Aceves uses it to effect.
6Having established both the theory and analytical approaches, the text then moves to a thorough examination of each individual policy case. This is an admittedly tedious part of the book, as it is quite thorough in its account of descriptive statistics, model building, and results for each of the three cases. The most interesting results, however, come at the end with the cross-policy comparison using SUR. The analysis yields findings that confirm, in part, and disconfirm, in part, the theory regarding variation in morality content. For neighbor effects, medical marijuana and anti-obesity have the expected positive effect, but there is a backlash effect for same-sex marriage bans. In other words, neighbor state adoption reduces the likelihood of ban adoption. This negative neighbor effect certainly occurs in other diffusion studies, but it is not well understood in the literature. Cruz-Aceves concludes that this confirms that the motivation to learn increases as morality content decreases, but I would argue that even the negative effect is potentially learning. Instead, learning shifts from negative to positive lesson drawing as morality content decreases.
7The most perplexing finding is that ideological distance increases the odds of adopting both medical marijuana and anti-obesity legislation. The ideological effect stands in contrast with other research showing the expected effect of ideological distance from past adopters reducing the odds of adopting medical marijuana (Hannah and Mallinson 2018). Using the book’s framework, this means that states are learning from dissimilar jurisdictions on policies with relatively lower moral content. There is face validity to this conclusion, at least for medical marijuana. While largely liberal states adopted the policy early on through the ballot initiative, the policy has transitioned into wider adoption by more conservative states through both the ballot and legislatures. Thus, ideological similarity weakened as the policy diffused and the non-moral dimensions took root.
8One part of the conclusion that the analysis cannot disentangle is whether it is truly the morality content that is having the effect here or if attributes such as complexity and salience are really driving the findings. Granted, the two are tightly tangled as strong morality policies are expected to be technically simple and highly salient. Future work should try to disentangle these different policy attributes as well as further measure other dimensions of policies in ways that move beyond simple typologies. This work by Cruz-Aceves is helpful in furthering that effort for morality policy and offering a strategy for addressing other policy dimensions. It will be of interest to scholars of policy diffusion and policy process theory. It also demonstrates a good example of a quality dissertation for graduate students.