This is a methodologically exceptional and urgently topical tome. The authors masterfully integrate multiple methods to isolate and quantify the economic penalty for being Muslim in France, and to identify some drivers of discrimination and poor integration. As one would expect from such relevant research, Claire Adida, David Laitin, and Marie-Anne Valfort suggest policies to facilitate Muslim integration (Chapter 10). It is, however, tricky to formulate good policy recommendations, which often deal with messy and complicated realities, on the basis of rigorous research that clinically isolates a few suspected causal mechanisms. The authors do it as well as currently possible and consequently illustrate how even the most brilliant minds in our discipline cannot transcend the limitations of formulating policies based on rigorous research findings.
There may be a topical scholar’s dilemma (TSD): The more rigorous and clinical the research design, the more reliable the findings, but also the less useful these findings are in formulating policy recommendations. Mastery of quantitative research is measured by the ability to identify, isolate, and quantify the effect of a particular causal mechanism—a costly endeavor. The more complex the causal mechanism, and the more the scholar wants to increase confidence in his or her findings, the more costly is the implementation of the research design. The high cost of rigorous inquiry limits the range of what can be tested, and consequently the scope of findings on which policy recommendations could be based. Rigorous scholars therefore find themselves forced to rely on others’ research and on their intuition in order to produce policy recommendations, instead of building on their narrow findings. Let us look at three policy recommendations that the authors make to illustrate the point.
In their first policy proposal, Adida, Laitin, and Valfort recommend broadcasting one of their findings—that a part of anti-Muslim discrimination is nonrational (pp. 150–51). They base their recommendation on research by others that shows that referees in basketball games tend to correct discriminatory behavior when informed that it is nonrational and that it violates institutional norms. Their recommendation relies on reasoning by analogy instead of their own findings—they assume that agents make similar calculations in different institutional and cultural contexts. Their analogy also ignores the fact that only the French government has the authority and ability to broadcast and reinforce a powerful pro-equality narrative in the context that interests them. Past French governments, however, seemed more interested in marginalizing Muslims than helping them integrate. The authors are therefore making a policy recommendation based on wishing away another intractable problem afflicting French society—the government’s lack of interest in addressing the marginalization of Muslims.
The same issue applies to the recommendation that Muslim integration can be improved by increasing Muslim representation in the French parliament (pp. 175–76). While also sensible, this recommendation relies on dissolving the similarly formidable political obstacles that have kept Muslims from being represented in the French parliament. The book does not speak to how this could be done, nor does it refer to works that can speak to how it could be done. When minority advancement in one area (employment) requires breakthroughs in two others (representation and government support), the three are likely to be interlocked, with advancement in any area being held back by discrimination in the other two. The two ways to break the logjam may be a broad civil rights movement that aims to redress discrimination in all areas of life, or incremental improvements in each area that may ultimately unravel the entire discriminatory edifice. The scope of this research only allows the authors to explore measures addressing discrimination in the area of employment, but not in connected areas of governmental agenda setting or political representation.
In their second recommendation, Adida, Laitin, and Valfort advise Muslims to change their names to reduce discrimination against them in the labor market and in other spheres (pp. 151–53). They base this advice on findings from another study that shows through correspondence experiments that, all else being equal, French employers discriminate heavily in calling back applicants with Maghrebin names, but temper their discrimination in the case of Maghrebins with Christian first names. The authors claim that this finding suggests that Muslims applicants with a Christian first name nudge employers to be less discriminatory by signaling commitment to integration. This policy recommendation illustrates the TSD in two ways.
First, and just like in the earlier example, Adida, Laitin, and Valfort do not utilize their own research to make this recommendation—they rely on another study. Their rigorous and costly correspondence experiment is only designed to quantify the substantial amount of discrimination against Muslim job applicants in contrast with Christian applicants who share their ethnicity and white French natives, but could not possibly shed light on whether Muslims’ adoption of Christian first names effectively signals integration or reduces bias. They could have made this same recommendation on the basis of the other scholar’s focused research without conducting a single experiment.
Second, neither the authors’nor the other scholar’s correspondence experiments shed light on post-call-back discrimination. A Muslim may benefit from a Christian name at the correspondence stage of a job application because of a corporate officer’s inattention, but his or her ethnicity may be only too obvious at the interview stage. It is perfectly fine to limit research to the correspondence stage of the hiring process, but such research does not suffice to show that individuals with Muslim last names would reduce future discrimination against their newborns by calling them Georges or Marie.
A third recommendation made by the authors is for the state to use correspondence tests, such as the one they use, to regularly audit the recruiting practices of corporations (p. 157). Unfortunately, the tools used by social scientists are geared toward the identification of trends across large samples, not the enforcement of policy or the auditing of specific firms. Random selection for a correspondence test allows the characteristics of firms in each treatment group to be comparable, which is perfect for social science purposes, but also means that no human resources (HR) officer needs to see two matched applications. Government enforcers, however, will need to produce firm-specific evidence and contend with firm-specific hiring justifications. They would need to submit matched pairs of pretend applicants to the same corporation in large enough numbers to prove that the firm is discriminating against minority applicants. The number of matched applications submitted may have to become prohibitive if the firm has a mix of biased and unbiased recruiters or if, for example, its officers discriminate against Muslim men but not Muslim women (four treatments). Tests may therefore only help audit compliance by very large enterprises with enough positions to fill. Also, a correspondence test works best with unaware firm officers who do not react strategically. An HR officer anticipating potential state auditing could behave fairly at the call-back stage but discriminate during the interview stage, or could completely ignore applications that appear to form matched pairs.
These problems are not unique to this outstanding book—they afflict the work of all scholars trying to inform policy based on rigorous research. I am not arguing that rigorous scholars should not formulate policy recommendations or that researchers should not be rigorous, but there may be a need for a disciplinary conversation about how to convincingly transition from research findings to policy formulation.
To be clear, the recommendations by Adida, Laitin, and Valfort are instructive, sensible, and compelling, and they are likely to be effective if implemented. These recommendations are informed by the intuition of brilliant thinkers and their research, even if they are not directly based on their findings. And finally, the cachet of intellect may very well be what is needed to help societies such as France’s to break out of damaging thought loops on Islam, integration, and discrimination. For that, and for much more, Why Muslim Integration Fails in Christian-Heritage Societies deserves to be read and discussed in graduate seminars for many years to come.
