Our target article distinguishes between policy approaches that seek to address societal problems through intervention at the level of the individual (adopting the “i-frame”) and those that seek to change the system within which those individuals live (adopting the “s-frame”). We stress also that a long-standing tactic of corporations opposing systemic change is to promote the i-frame perspective, presumably hoping that i-frame interventions will be largely ineffective and more importantly will be seen by the public and some policy makers as a genuine alternative to systemic change. We worry that the i-frame focus of much of behavioral science has inadvertently reinforced this unhelpful focus on the individual. In this response to commentators, we identify common themes, build on the many constructive suggestions to extend our approach, and reply to concerns. We argue, along with several commentators, that a key role of behavioral public policy is to clarify how to build support for systemic reforms for which there is a broad consensus in the policy community, but which are opposed by powerful special interests.

Behavior genetics is a controversial science. For decades, scholars have sought to understand the role of heredity in human behavior and life-course outcomes. Recently, technological advances and the rapid expansion of genomic databases have facilitated the discovery of genes associated with human phenotypes such as educational attainment and substance use disorders. To maximize the potential of this flourishing science, and to minimize potential harms, careful analysis of what it would mean for genes to be causes of human behavior is needed. In this paper, we advance a framework for identifying instances of genetic causes, interpreting those causal relationships, and applying them to advance causal knowledge more generally in the social sciences. Central to thinking about genes as causes is counterfactual reasoning, the cornerstone of causal thinking in statistics, medicine, and philosophy. We argue that within-family genetic effects represent the product of a counterfactual comparison in the same way as average treatment effects (ATEs) from randomized controlled trials (RCTs). Both ATEs from RCTs and within-family genetic effects are shallow causes: They operate within intricate causal systems (non-unitary), produce heterogeneous effects across individuals (non-uniform), and are not mechanistically informative (non-explanatory). Despite these limitations, shallow causal knowledge can be used to improve understanding of the etiology of human behavior and to explore sources of heterogeneity and fade-out in treatment effects.

I argue that advancing “second-generation” or mechanistic causal knowledge of individual outcomes requires a comprehensive research programme that uses a variety of different methods in addition to the ones described in the paper under discussion. I also highlight that environment-focused approaches can be as instrumental in identifying potential phenotypic causes as gene-focused approaches.

Madole & Harden argue that the Mendelian reshuffling of genes and genomes is analogous to randomised controlled trials. We are not convinced by their arguments. First, their recipe for meeting the demands on randomised experiments is inherently inconsistent. Second, disequilibrium across chromosomes conflicts with their assumption of statistical independence. Third, the genome-wide association study (GWAS) method has many pitfalls, including low repeatability.

Madole & Harden plead for better integration of causal knowledge of different depths to understand complex human traits. Classically, local causes – a particular type of shallow causes – are considered less useful than more generalisable causes, giving a false impression that the latter causes are more useful and desirable. Using a simple example, I show that sometimes the contrary is true.

In their target article, Madole & Harden offer an account of “what it means for genes to be causes” of social outcomes to bolster their claim that genetics should be incorporated into social science with practical implications. Here I object to several key features of their arguments, their representation of the state of science, and claims about the utility of genetics for social science and society.

Evidence of a causal role for genes in human behavior underpins genetic essentialism, the scientifically flawed and socially hazardous idea that heritable characteristics are immutable. Behavior geneticists can challenge this idea by designing research that brings the contextual dependence of heritability estimates into sharper focus, and by incorporating a relevant statement into research reports and public outreach.

Causal dispositionalism developed in metaphysics of science offers a useful tool to conceptualize shallow causes in behaviour genetics, in a way such that (a) it accounts for complex aetiology and heterogeneity of effects, and (b) genetic causal contribution can be considered to be explanatory. Genes are thus causal powers that make a difference.

Any empirical claim about the role of genes in socioeconomic outcomes involves successfully addressing the identification problem. This commentary argues that socioeconomic outcomes such as education are sufficiently complex, involving so many mechanisms, that understanding the role genes requires the use of formal theoretical structures.

Behavioral genetics typically finds that the so-called shared environment contributes little or nothing to explaining within-population variation on most traits. If true, this has important implications for where not to look for good targets of interventions: Namely all things that are within the normal range of variation from one rearing environment to the next in that population.

Madole & Harden describe how genetics can be used in a causal framework. We agree with many of their opinions but argue that comparing within-family designs to experiments is unnecessary and that the proposed influence of genetics on behavior can be better described as inus conditions.

In our commentary we ask whether we should ultimately endeavour to find the deep causes of behaviours? Then we discuss two extensions of the proposed framework: (1) Mendelian randomisation and (2) hypothesis-free gene–environment interaction (leveraging heterogeneity in genetic associations). These complementary methods help move us towards second-generation causal knowledge, ultimately understanding mechanistic pathways and identifying more effective intervention targets.

Madole & Harden argue that just as the results of randomized controlled trials (RCTs) represent gains in causal knowledge and are useful, despite their limitations, so too are the findings of human behavior genetics. We argue that this analogy is misleading. Unlike RCTs, the results of human behavior genetics research cannot suggest efficacious interventions, nor point toward future research.

We are less optimistic than Madole & Harden that family-based genome-wide association studies (GWASs) will lead to significant second-generation causal knowledge. Despite bearing some similarities, family-based GWASs and randomised controlled trials (RCTs) are not identical. Most RCTs assess a relatively homogenous causal stimulus as a treatment, whereas GWASs assess highly heterogeneous causal stimuli. Thus, GWAS results will not translate so easily into second-generation causal knowledge.

Counterfactual causal interpretations of family genetic effects are appropriate, but neglect an important feature: Provision of unique information about expected outcomes following an independent decision, such as a decision to intervene. Counterfactual causality criteria are unlikely to resolve controversies about behavioral genetic findings; such controversies are likely to continue until counterfactual inferences are translated into interventional hypotheses and designs.

Madole & Harden develop some good ideas about how to understand genetic causality more clearly, but they frame the benefits of behavior genetics research at a largely collective level, focused on the pros and cons of different ways to engineer the gene pool or social behavior. This neglects the individual benefits of hereditarian insights for mate choice and parenting.

When building causal knowledge in behavioural genetics, the natural randomisation of genotypes at conception (approximately analogous to the artificial randomisation occurring in randomised controlled trials) facilitates the discovery of genetic causes. More importantly, the randomisation of genetic material within families also enables a better identification of (environmental) risk factors and aetiological pathways to diseases and behaviours.

We explore Madole & Harden's (2022) suggestion that single-nucleotide polymorphism (SNP)/trait correlations are analogous to randomized experiments and thus can be given a causal interpretation.

It has been known for decades that inference concerning genetic causes of human behavioral phenotypes cannot be legitimately made from correlations among relatives. We claim that these inferential difficulties cannot be overcome by assigning different names to causes inferred from within-family and population-level genome-wide association studies (GWASs). For educational attainment, for example, unraveling gene–environment interactions requires more than new names for causes.

Madole & Harden's assertion that the effects derived from within-family genome-wide association studies (GWASs) and from randomized controlled trials (RCTs) are equivalent is misleading. GWASs are substantially more “non-unitary, non-uniform, and non-explanatory” than RCTs. While the within-family GWAS bring us closer to identifying genetic causes, whether it will change behavioral genetics into a causal science is an open question.