Empirical Constitutional Studies: Future Directions

Introduction

Our new book—How Constitutional Rights Matter—tries to answer a difficult empirical question: do constitutional rights actually change government behavior? We theorize that constitutional rights that protect individuals often fail to constrain governments, but that constitutional rights that protect organizations can be powerful tools to push back against repression. The reason is that organizations have both the incentives and means to strategically use the constitution to protect their rights. To test this theory, we leveraged a range of empirical methods: large-N statistical analyses, case studies from around the world, survey experiments conducted in Turkey and the United States, and insights from an expert survey of constitutional law experts from over 100 countries. The results from these different analyses consistently suggest that enumerating individual rights in constitutions is not associated with better rights protections, but that constitutionalizing organizational rights does appear to improve actual rights practices.

Our Book is part of a growing body of research in empirical constitutional studies. This line of research has sought to use methods from empirical social science to assess the impact of constitutional law on a range of outcomes of interest. In the fall of 2020, we convened a virtual Symposium to discuss our Book and the future of scholarship in this growing field. This Symposium brought together leading researchers that study constitutions around the world. Many of those researchers use empirical methods in their own work, but some are skeptical of empirical approaches to constitutional law.

Many of those participants generously contributed written essays on the future of empirical constitutional studies to this online Symposium. These contributions can be divided into three categories: (1) identifying new methods that should be used (more) in empirical constitutional research, (2) raising new questions for future exploration, and (3) proposing new avenues for data collection.

New Methods

In his 2014 Book laying out conceptual and methodological challenges faced by the growing field of comparative constitutional law, Ran Hirschl argued that “there is no magic bullet or one-size-fits-all research design ‘formula’ for a field as rich and diverse as comparative constitutional studies.” Our Book embraces this view and therefore adopts a range of different methods to explore the effectiveness of constitutional rights. But the methods used in our Book by no means exhaust the available methods to study constitutional law’s impact. Indeed, some contributors to this special issue propose new methods that could be incorporated into the study of empirical constitutional law questions.

Professors Rosalind Dixon and Richard Holden propose that future research in empirical constitutional law should consider the use of the “synthetic control” method. This method, which has become popular in political science and economics, uses “a data-driven procedure to create a comparison unit in comparative case studies.” As Dixon and Holden point out, the advantage of the method is that it allows researchers to explore the impact of a constitutional phenomenon—such as a new court decision or the adoption of a new constitutional provision—when this phenomenon occurs in one country only. Moreover, this method does not require relying on a single comparison case for each country of interest; rather, information from many different countries is used to construct a “synthetic” comparison case. As a result, the method allows one to treat “every constitutional moment” as “sui generis,” which aligns with how many constitutional lawyers think about constitutional law. We agree with Professors Dixon and Holden that this method offers a promising avenue for future research.

Professors Adi Leibovitch and Alexander Stremitzer argue that empirical constitutional law scholarship could benefit from the increased use of abstract laboratory experiments. They observe that there is an experimental turn in constitutional law, but that most experimental studies are comprised of survey experiments, which allow researchers to explore how publics respond to constitutional law. They argue that empiricists interested in constitutional questions should also consider the use of “abstract experiments.” That is, “experiments that confront experimental subjects with stripped-down scenarios that let them make decisions in an artificial and abstract world in which their choices have payoff consequences.” Leibovitch and Stremitzer argue that, because the insights from abstract experiments are general, they can play an important role in theory-building in constitutional law, especially when used as a supplement to observational studies or vignette studies. We agree that this seems to be a particularly appropriate way to leverage abstract experiments to study constitutional questions.

New Questions

Other Symposium contributions focus on our theoretical account of how constitutional rights can matter. A starting premise of our theory is that enforcing constitutional law is difficult. The reason is that constitutions lack an external enforcer that can force governments to comply with their constitutional obligations. In the absence of an external enforcer, the task for imposing constitutional constraints often falls on ordinary citizens. But citizens trying to hold governments to their constitutional promises face two important obstacles: (1) they need to agree that the constitution is indeed violated (in our Book, we call this a “coordination problem”), and (2) they need to orchestrate actions that punish the government for transgressing the boundaries set by the constitution (in our Book, we dub this a “collective action problem”).

One of the main arguments in our Book is that the best way to overcome these problems is through formal organizations. Simply put, organizations persuade people that certain government actions amount to a constitutional violation, and they can rally their members to mobilize against constitutional violations. We further argue that the organizations that are best equipped to enforce constitutional rights are those that themselves enjoy constitutional protections: religious groups protected by the freedom of religion, political parties protected by the right to form political parties, and trade unions protected by the right to unionize. These organizations are not established specifically to protect rights, but they have important incentives to do so when their rights are encroached upon.

Two of the contributions build on these ideas by focusing on the coordination problem inherent in constitutional enforcement. Professors Jerg Gutmann, Mahdi Khesali, and Stefan Voigt share our starting assumption that enforcing constitutional law is difficult in the absence of an external enforcer. But instead of focusing on formal organizations, they theorize that coordination problems can be overcome when constitutional texts are clearly written. When a constitution contains clear rules, there may not be a need for organizations to present people with arguments that certain government actions transgress the boundaries set by the constitution. Gutmann, Khesali, and Voigt present an empirical test of this idea by calculating a comprehensibility score for each constitutional text and correlating this score with a measure that captures the degree to which citizens seek to enforce the constitution. Their initial analysis does not reveal conclusive evidence showing that constitutions that are more clearly written are better upheld. But this interesting insight merits further exploration, perhaps through the use of an experimental research design; experiments are well suited to explore whether people reach similar understandings of a constitution when it is more clearly written.

Professors Kevin Cope and Charles Crabtree also start from the premise that enforcing the constitution is difficult. But instead of focusing on clarity of the text, they probe whether people have “common knowledge” about when government actions cross constitutional lines. When such common knowledge is present, then presumably the comprehensibility of the text of the constitution itself does not necessarily matter and organizations might not be needed to overcome the coordination problems. Cope and Crabtree test these ideas through a nationally representative survey conducted in the United States which asked respondents over a dozen of questions about the constitution (and international law) as well as how confident they are about their answers. They observe that whether people actually know the right answers is less important than the presence of shared understandings of what the constitution requires. But they find that neither are particularly high: people have different understandings of what the constitution requires, although there is more shared knowledge of the constitution than of international law. The findings of both the Gutmann, Khesali, and Voigt study and the Cope and Crabtree study are thus easy to square with our argument that formal organizations might be necessary to overcome the problems associated with enforcing the constitution.

Two other contributions invite us to further theorize about what type of organizations can do the hard work of enforcing constitutional rights. Professor Zachary Elkins shares our premise that organizations may be the key to overcoming the challenges associated with constitutional rights enforcement. But he argues that we need to further probe what kind of groups are able to organize to enforce constitutional rights. Our Book argues that the organizations best equipped to do so are not those specifically established to enforce rights, but those that are necessary for the enjoyment of rights, such as religious groups or trade unions. These organizations have both the incentives and means to mobilize against the government when their rights are encroached upon. Elkins, by contrast, draws attention to strong interest groups that are formed specifically to enforce constitutional rights, such as the National Rifle Association in the United States. He also provides interesting and powerful examples of interest groups named after specific provisions of the constitution, which invites further consideration of whether the constitution itself can lead the formation of new groups dedicated to enforcing the constitution. We believe that studying whether and under which conditions such groups form is an important avenue for future research.

Professors Ran Hirschl and Alexander Hudson draw our attention to two different types of groups that often have powerful incentives to enforce rights: (1) property holders and corporations who seek to enforce constitutional property protections and (2) indigenous groups that seek to enforce minority rights. They theorize that, if rights enforcements indeed happens due to actions taken by constituencies invested in enforcing a right, property rights and minority rights may be particularly effective in constraining the state since they have dedicated constituencies. But Hirschl and Hudson also observe that these rights challenge the individual-organization rights distinction we make in the Book, as property holders and minority groups have both individual interests and organizational interests. We agree that these groups invite further theory development and merit empirical exploration in future work.

Professors Hirschl and Hudson also push us to further probe the origins of constitutional rights provisions. Specifically, they hypothesize that it is possible that powerful groups not only push for the enforcement of rights but also for their constitutionalization in the first place. If so, it may be the case that the constitution is merely epiphenomenal: it is organized groups that are doing all the work of making the constitution become real. We agree with Professors Hirschl and Hudson that this is a possibility, in which case the relationship between certain constitutional rights provisions and better de facto rights protections would represent a correlation and not a causal relationship. While some organizations may have been formed because the constitution protects a certain right (recall Elkins list of organizations named after specific constitutional provisions), we agree that many organizations predate the constitutionalization of rights and may have in fact pushed for their constitutionalization. But importantly, our case studies reveal that organized groups frequently rely on the constitution. We therefore believe that, at minimum, constitutional protections strengthen preexisting groups. But we acknowledge that further research is needed to unpack how organizations and constitutional rights protections can be mutually reinforcing.

New Data

A final contribution draws attention to data challenges in the empirical study of constitutional law. Most of our empirical analysis relies on the coding of large-C constitutions—that is, the constitutional rights explicitly enumerated in the text of the constitution. This approach is common in the recent wave of studies that seek to explore constitutions’ impact. Yet, it is subject to the criticism that the text of the constitution alone gives an incomplete picture of the protections present in any given constitutional system. Rights can also be constitutionally protected through court decisions, constitutionally-incorporated treaties, or even long-standing conventions. In our Book, we make an initial attempt to address this critique by presenting novel data from an expert survey that enlisted the knowledge of experts from over 100 countries on how rights are protected outside of the constitution in their country. But, of course, expert evaluations raise a different set of methodological challenges, such as whether and to what extent experts push their own normative priors when interpreting the constitution. Creating better data on constitutions that goes beyond the text of the constitution alone remains an important challenge for the emerging field of comparative constitutional studies.

Professor Mark Tushnet focuses on a different data challenge. He worries that our approach to coding constitutional rights, which is the basis for our quantitative analysis, is overly formalistic because our coding does not take account of limitation clauses. Without limitation clauses, Tushnet contends, our coding merely captures “magic words,” such as “the right to unionize” or “religious freedom.” More precisely, Tushnet ridicules the notion that we treat provisions with broad limitation clauses as unqualified, which offends “the lawyer” in him. To illustrate, he claims that we “treat the provision in the Constitution of Chile, that everyone has ‘the right to unionize in the cases and manner provided by the law’ as unqualified, which seems odd without some discussion of how the term ‘provided by the law’ has been interpreted.”

Tushnet’s critique is somewhat puzzling. We do not, in fact, treat all rights as unqualified. Instead, our approach relies on a different set of assumptions: (1) that all rights can be limited under some circumstances, but that (2) the scope of permitted limitations tends to be similar across countries and that (3) limitation clauses do a poor job of capturing the extent to which rights restrictions are permitted.

The notion that rights are never unlimited is fairly uncontroversial. But the second assumption deserves some elaborating. As we observe in the Book, we believe that there exist well-developed transnational understandings on whether and to what extent rights can be limited. Indeed, a growing literature in comparative constitutional law shows that there is a “global model” of constitutional rights that transcends national boundaries. As a result, the contours of each right, and the scope of permitted limitations, are similar across countries. When activists make constitutional arguments that certain government actions amount to constitutional violations, they frequently rely on these well-accepted interpretations. It is for this reason that each of the substantive chapters in our Book starts out by drawing on international human rights law and influential constitutional cases from different national systems to sketch well-accepted doctrinal contents of each right we study. Of course, we acknowledge that the assumption that the scope of permitted limitations is similar across systems works better for some countries and some rights than for others (we note, for example, that there is more cross-national variation in the understandings of the right to free speech than for the right to unionize).

But while ours is by no means a perfect approach, we believe that assuming that there are well-developed transnational understandings that national activist rely on is a sounder methodological choice than attempting to interpret limitation clauses. In fact, doing so would itself be overly formalistic (arguably more formalistic than merely capturing whether a right is present and assuming that the scope of permitted limitations is similar across systems). The reason is that constitutional limitation clauses themselves do a poor job of capturing the extent to which rights can be limited. Indeed, it is  particularly difficult to interpret constitutional limitation clauses and to determine whether and to what extent the language is restrictive.

To illustrate, let’s return to Tushnet’s own example on the “by the law” clause in the right to unionize in the Chilean Constitution. One take is that allowing restrictions “by the law” is an open-ended invitation for governments to violate rights. Another take is that this is an important safeguard, as it requires legislative involvement for rights to be curtailed and executive actions that do so would by definition be unconstitutional. What work a clause like this is doing is impossible to judge without substantial in-depth knowledge on the Chilean Constitution. Further compounding the problem is that we would have to decide how to treat rights provisions without limitation clauses (as is common in the U.S. Constitution, for example). Tushnet’s approach implies that we would have to treat these as unrestricted rights, which is of course false. The reality is that defining the scope of rights is the realm of courts, and that taking limitation clauses seriously would require the consultation of court cases across different systems.

We suspect that, if we were to closely examine permitted rights limitations across countries, we would find that courts reach rather similar conclusions on the scope of permitted limitations, usually by relying on proportionality analysis, which is a doctrine with important trans-national dimensions. (Indeed, a growing body of research shows that constitutional courts frequently cite other courts when interpreting the constitution.) Yet, attempting to assess the actual contours of each right in each country through the analysis of court cases is unlikely to be a feasible project for most researchers. A more feasible approach to get a handle on the scope of each right might be the consultation of foreign constitutional law experts, but Tushnet himself objects to relying on expert evaluations to better understand constitutional law across legal systems, as he is worried about expert bias. We suspect that Tushnet’s bottom line is that comparison is not possible without a deep understanding of the nuances of different legal systems, which is why he is skeptical of an enterprise like ours. If one does believe that this kind of comparison is worthwhile, however, we think that our approach is a reasonable one.

Conclusion

As these contributions make clear, our book is neither the start nor the end of empirical research on the effect of constitutional rights specifically or constitutional law more generally. Our hope is that these contributions help inspire research using increasingly sophisticated methods, exploring new theories, or spuring the collection of improved data.