When religion and politics become entangled with each other, which one has the upper hand? Do politicians simply instrumentalize religion for their own purposes? Or does involvement with religion constrain what politicians can get away with? If it does, is that because of the content of religious doctrine? Or is it simply because religious movements, like any other coordinated movements of citizens, have political weight?

Adam Smith, whose book An Inquiry into the Nature and Causes of the Wealth of Nations was published a quarter millennium ago this year, devoted many pages in Book V of that work to arguing that religious movements were above all constituted by their activities and only secondarily by their doctrines. Those activities were subject to basic economic constraints, since they consumed resources like the time and energy of participants, space to hold meetings, upkeep of buildings, and the means of broadcasting their message. Religious movements compete against each other, against secular movements, against the weight of indifference, and the lure of alternative ways for people to spend their time. Movements facing stronger competition would adapt their doctrines to appeal to members who might be tempted to defect to their rivals. For Smith, the message they broadcast was the result of their activities, and not an independent constraint on those activities.

It also followed that the arrangements religious movements might make with political leaders would be largely transactional, with relatively few constraints imposed by doctrine or tradition. Such arrangements might be to the short-term advantage of politicians, who would instrumentalize religion to borrow its legitimacy for themselves. It could also be to the short-term advantage of religious leaders, who might gain protection from their rivals (and even use the law to persecute those rivals). But Smith was scathing about the long-term consequences for both politics and religion of such opportunistic arrangements, predicting that they would eventually undermine the legitimacy and appeal of both parties.

For much of the next two centuries, Smith’s approach to these questions fell out of fashion. Economists wrote almost nothing about religion until the 1970s, while sociologists and historians tended to define religious movements by their doctrines, which presupposed that the doctrines made a difference to the activities rather than vice versa. But the revival of economic approaches to the activities of religious movements, which has been going strong for half a century now, provides a very useful lens through which to examine political involvement in religion. The conventional wisdom until recently was that the world was irrevocably secularizing, so politicians would be less and less willing to tarnish their brand by association with ideas that were so obviously on their way out.

Times change. The political leaders of the United States, Russia, Hungary, Poland, India, and even China—to say nothing of the Muslim world—have been deploying religious messaging in increasingly forceful ways, and they seem mostly to have been rewarded for it politically. How is this possible, and is religion capable of acting as a constraint on what political leaders can say or do?

In my 2024 book, The Divine Economy: How Religions Compete for Wealth, Power and People, I was mostly skeptical of the idea that religious doctrines imposed significant constraints on political power. I drew mostly on cases that suggested political leaders could instrumentalize religious messages in favor of whatever strategies they had independently decided to pursue. I was not dogmatic on the point, just unpersuaded that many important exceptions existed. I received some pushback—for example, Timur Kuran’s excellent book Freedoms Delayed: Political Legacies of Islamic Law in the Middle East (published in 2023 after my own book went to press) has persuaded me that the Islamic awqaf (Islamic charitable trusts) really did prevent Middle Eastern societies from evolving in ways that would have been possible if Islamic law had been written otherwise. And work by Samuel Bazzi, Gabriel Koehler-Derrick, and Benjamin Marx has persuaded me that this was true outside the Middle East as well, specifically in Indonesia (“The institutional foundations of religious politics: evidence from Indonesia,” Quarterly Journal of Economics 135(1):1–67). So when the editors of this journal asked me to review these three books, the covers of all of which proclaim that religious ideas matter for political outcomes, I relished the challenge and expected to revise my conclusions substantially.

Surprisingly, reading these three impressive books convinces me more than ever that Smith’s view of politics and religion has a lot to say to the modern world in general, and to these three case studies in particular. That is not because their arguments are unconvincing, but because they are more subtle than the blurbs suggest. They also suggest some ways in which Smith’s view needs to be revised.

Let us start with Josef Hien’s study of religion and the welfare state in Germany and Italy, from the nineteenth century to the present day. The subtitle proclaims that it is about “the impact of religion on the German and Italian welfare state,” suggesting a unidirectional causal mechanism. But the reality it documents is more complex. In both countries, the Catholic Church played an important part in shaping outcomes—but did so very differently in the two countries. This was not because of differences of doctrine (Catholicism has the same doctrines in Italy as in Germany), but because in Germany the Catholic Church had to compete with a strong majority Protestant establishment, while in Italy it had a near monopoly on religious adherence and competed instead with a secular liberal minority. The result, according to Hien, was that:

Specifically, the social security plans of the conservative Protestants under Bismarck put pressure on German Catholics “to develop new workable policy solutions, which then allowed them to influence Bismarckian legislation with concrete Catholic-style policy proposals” (p. 49). In Italy, in contrast, “the Catholic Church could hold on to its centuries-old monopoly of poor relief institutions (Operé Pie)” (p. 178). In short, says Hien, “Catholicism changes if it competes with other political and religious competitors” (p. 26). Adam Smith could not have put it more pithily.

The practical consequences in the two countries were profound, and not just in the nineteenth century. The different conditions of religious competition even shaped the responses of the two countries to the large falls in fertility in the last decades of the twentieth century:

As this example suggests, Hien does not imply that Catholicism could be flexible without limit—the intellectual somersaults necessary to abandon the male breadwinner model were too great even for the previously adaptable German Catholic Church, which therefore lost its influence over an important plank of social policy.

The first sentence of the summary of Miriam Lowi’s excellent book Refining the Common Good asks, “How has Islam as a set of beliefs and practices shaped the allocation of oil revenues in Arab Gulf monarchies?,” which also leads one to expect a causal story running from beliefs to outcomes. Again, the rest of the book is more complex than this implies. In fact, her analysis is trenchant: “In the contemporary period, regime behavior is not merely detached from religious principles, but more significantly, norms are either reconfigured or their interpretation revised for the sake of narrow (political) interests. Maintaining dynastic states is the priority; oil and Islam are its principal tools” (p. 5).

Lowi documents how Gulf rulers have reinterpreted Islamic doctrine to suit themselves in fairly radical ways—for instance, by excluding many kinds of beneficiaries, such as immigrants, from the distribution of charitable donations, despite the clear nature of zakat as being directed toward the whole Muslim community, the umma—to the point where “social categories such as umma…have little meaning” (p.169). She shows how the constraints of Islamic finance are routinely circumvented—for example, by the Saudi government, which “enticed Islamic banks to purchase government bonds by allowing their shareholders’ zakat obligations to be reduced by the price of the bond.” The zakat obligation has been renegotiated by several governments, “as when the Kuwaiti state reduced the zakat obligations of businesses from 2.5 percent to 1 percent” (p. 170).

Her book also shows repeatedly how Muslim intellectuals (the ulama) have trimmed their sails to suit the powerful. This is, of course, hardly novel in the history of Islam, Christianity, or any other set of ideas. Her discussion of this would have benefited from some historical perspective; at times (on page 12 ff.), she writes as if this phenomenon is the product of the modern state in the Middle East and North Africa. She cites Tunisia, Libya, Oman, and Saudi Arabia as places where “states promote a particular understanding of Islam that suits their interests.” But it would be interesting to compare with, say, Egypt, where the government has continually attempted to manipulate or silence the Islamic university Al-Azhar, but has not always succeeded in doing so on its own terms. This is not least because the government has needed the legitimacy that only a somewhat independent source of intellectual capital can provide.

Still, Lowi’s book demonstrates that the extent to which moral and religious ideas constrain rulers’ actions can be quite limited at least in the case of the Gulf monarchies. This sits comfortably with a Smithian view of religious doctrine as politically malleable. But it may be a rather special case: in the Gulf, access to oil revenues gave rulers a great deal of scope for independent action, with no need to seek tax revenues from the population under their rule. A rather different perspective on Islam and politics can, therefore, be expected from Pakistan, a country that struggles to assure basic civil order and the provision of public goods, and as recently as 2010 raised only 8% of GDP in tax revenue (while its neighbor India raised over 16%).

Christopher Candland’s The Islamic Welfare State, like the other two books, has a summary that leads one to expect that religious ideas directly impact political action: “Its humanitarian spirit makes Islam a compelling, community-strengthening faith, which motivates people to provide essential services to the needy and foster moral sentiments that build social solidarity. This challenges the legitimacy of government that focuses on ‘protecting Islam’ and ‘national security’ rather than enhancing the lives of ordinary people.” In fact, his account is better described as an analysis of how Pakistani governments have repeatedly tried and failed to mobilize Islamic commitment on the part of the population in support of a nation-building project. He is scathing about their multiple failures: “Government is not incapable of serving the poor; it is largely unwilling. Governments have not provided millions of people with education, healthcare, personal security, family services, legal aid services, or other vital public good(s)” (p. 96). He shows that the instrumentalization of Islamic rhetoric by the government, and the passage of legislation to nationalize the activities of Islamic charities and educational institutions (beginning in the late 1950s) have repeatedly failed to mobilize either legitimacy or resources on the scale envisaged.

Instead, “the private religious welfare sector provides education and health services to millions of people every year. Government hospitals and government schools are the least preferred option…the Islamic social welfare sector probably serves more people than does the government” (p.4). But Government efforts to regulate and control this sector have repeatedly failed. In education, for example:

Why is this? In part, it is because, as Candland puts it, “Islamic social thought creates three spheres, wherein a part of the private sphere (that of the faithful) can create public goods. Madaris are privately provided public goods, a reality that the dichotomy between an authoritative public sphere and a voluntary private sphere cannot accommodate. The notion that the government should regulate madaris is, accordingly, anathema to the faithful inhabitants of the madaris sector” (p. 218). But it is also because government efforts to regulate and control run up against the unavoidable social, religious, and political diversity of Pakistani society.

Though an overwhelmingly (98%) Muslim country, Pakistan is remarkably confessionally diverse within Islam, with around a quarter of the over 200 million population belonging to various Shia movements, and the Sunnis divided between a large variety of traditions. Candland summarizes the latter as falling into three main currents: the Modernists, the Barelvi, and the Deobandi, the distinction between the latter two corresponding broadly to the Sufi-Wahhabi split familiar from elsewhere. While an outsider might suspect that attempts to Islamicize educational and charitable institutions in Pakistan stem from a victory by Deobandi forces over Modernists, Candland has news for you:

More perhaps than in the Gulf states, there has been pushback from at least some influential Muslim scholars: Candland cites the refusal of the leading ulama to participate in the state funeral in 2016 of the leading Islamic philanthropist Abdu Satter Edhi, a funeral which the government had hijacked for propaganda purposes.

Paradoxically, therefore, “Whenever Pakistani governments have ‘Islamicized’ traditional institutions in Pakistan, they have promoted a Deobandi version of Islam with Wahhabi overtones that is anathema to the practices and understanding of Islam of most Pakistanis” (p. 216). Indeed, the only segments of Pakistani society that respond positively to these initiatives are those that are widely perceived as sectarian. As a result, writes Candland:

This third study provides, therefore, more of a challenge to the Adam Smith perspective on religion and politics than the other two. Religious doctrines—in Pakistan at least—cannot straightforwardly determine what politicians do. If they want to bend the rules of zakat, or of Islamic finance, they almost always will. But such doctrines can act as coordinating mechanisms for enabling forces in civil society to resist the more opportunistic manipulations of politicians. When a government whose members try to “keep the citizenry compliant” resorts to language and doctrine that are perceived as manifestly insincere and unsuited to the situation, the citizens and their intellectual leaders may resist, if only passively. It’s not that the doctrines themselves require resistance. It’s rather that if the citizens are minded to resist political manipulation, the doctrines make it easier for them to do so.

Religious doctrine serves more often as a political coordinating mechanism than as a hard political constraint. The lesson seems much more broadly applicable. Religious doctrine clearly has some constraining power, not just on political action but on economic development more generally. But it can also be manipulated in outrageously opportunistic ways. In particular, doctrinal language can be used not only to constrain the actions of the powerful, but to mobilize their supporters in conflicts whose underlying causes are principally political or economic.

Historical instances of violence between religious communities illustrate this dynamic. Economists Sascha Becker and Luigi Pascali have investigated anti-Semitic violence in Europe after the Reformation with this in mind (“Religion, Division of Labor, and Conflict: Anti-Semitism in Germany over 600 Years,” American Economic Review, 109(5): 1764–1804). Unlike Catholics, for whom usury was forbidden (an example of a religious doctrine with tangible economic effects), Protestants were free to move into the moneylending business. This meant that they perceived Jews for the first time as direct competitors rather than providers of services. Becker and Pascali show not only that anti-Semitic persecution became more common in Protestant regions of Germany than in Catholic ones, but also that it was particularly pronounced in cities where Jews had become most established as moneylenders. But of course, the violence was never justified in economic terms—it was promoted under a variety of entirely specious doctrinal pretexts.

To sum up, none of these studies supports a view that holds religious doctrine to be straightforwardly effective at promoting religiously approved political outcomes. Doctrine itself can be instrumentalized, as Adam Smith knew, and it is flexibly adapted by religious movements that are aware that they compete against each other and against secular rivals. But, as in the case of the Islamic trusts (awqaf) studied by Timur Kuran, religious doctrines developed in one context can create constraints in others, often many centuries after the doctrines were first formulated. These constraints sometimes hold back desirable developments, like forms of innovation. They sometimes allow propagandists to mobilize violent forces for their own purposes—a development we have seen recently with the Russian Orthodox Church’s fawning support for Vladimir Putin’s invasion of Ukraine. But sometimes, too, the constraints of religious doctrines can make it harder for autocrats to impose their will on a compliant population in the name of a supposedly higher religious law.