Order out of Chaos analyzes the consequences of state collapse in Basra, Iraq’s main southern city, following the collapse of the Ba’athist regime in April 2003. In its wake, the city’s residents, long acclimated to Saddam’s highly regimented regime, immediately faced an unaccustomed reality of full-fledged social disorder. The occupation authorities did not (or could not) step in, producing massive looting and crime. Trash was uncollected, and people dumped it wherever it was most convenient. Streets and canals were overcome with raw sewage. Polluted water pooled throughout the city, and the electrical grid and overall infrastructure cratered. Because there were no municipal workers—even bribable ones—to address these issues, all these problems persisted. This is the situation that David Patel found himself in when he arrived in Basra the following September to conduct an ethnographic study.

The problem of social order is one of the enduring issues in social science, dating at least to the publication of Thomas Hobbes’s Leviathan in 1651. Hobbes attempted to explain how atomistic individuals in a hypothetical state of nature desiring to avoid injury and death could create a social order that would guarantee their security. That social order, according to Hobbes, consisted in the establishment of an authoritative state that could provide security and other public goods.

Whereas the dismantling of the Ba’athist state clearly led to social disorder, Basra in 2003 hardly resembled a Hobbesian state of nature; that is, an institution-free zone. Even if there were no cell phones and social media, newspapers and radio were readily available. In addition, some preexisting institutions—familial, religious, and tribal ones, among others—survived Saddam’s demise. To what extent could they fill in for the absent state? And if they could, which ones would be most effective in producing social order? These are the principal questions that the book seeks to answer.

Patel finds that these preexisting institutions could only provide a modicum of social order. Media were no help. Religious groups were more successful in providing public goods than extended families or tribes. And among religious groups, Shi’ites, who make up the majority in Basra but had been mostly quietist until the collapse of the Baathist regime, were more successful than Sunnis.

The bulk of the book attempts to explain these findings. To begin, Patel argues that attaining social order requires the solution of two different collective action problems: coordination and contribution. Coordination occurs when individuals participate in a group endeavor only if they believe that a sufficiently large number of others will also participate. Contribution, in contrast, requires costly participation, which raises the possibility of free riding. Each of these collective action problems requires different solutions. Whereas cooperation often entails coercion or the provision of selective incentives, coordination requires that people do things in a coherent way and know what others are doing. Despite its centrality, Patel argues that little is known about the resolution of coordination dilemmas.

One reason that coordination problems are overlooked is because their solutions are often taken for granted. Once they have been solved, solutions tend to persist because they are self-enforcing and costly to change. In contrast, many solutions to free-rider problems require costly external enforcement that can end, leading to suboptimal outcomes or outright failure. A central challenge of coordination problems is that there are often multiple ways to solve them. Sometimes individuals observe, learn, and adjust their behavior until solutions are selected in an evolutionary process. Sometimes there is something focal that makes individuals expectations converge. And sometimes coordination can occur through explicit communication that creates common knowledge.

Common knowledge is knowledge of others’ knowledge, knowledge of others’ knowledge of others’ knowledge, and so on. Common knowledge implies meta-knowledge, knowing that others know something: this distinguishes it from merely widespread knowledge. Common knowledge alone is insufficient for a group to coordinate together, however. Patel goes beyond the usual game-theoretic definition of common knowledge by insisting that individuals must also believe that enough others will act on that knowledge. He claims that there is little theory about where such beliefs come from.

Order out of Chaos sheds empirical light on the issue. In brief, the Shi’ias only managed to help solve coordination, rather than contribution, problems. They did so by establishing a new institution—Friday sermons—in which imams addressed a wide variety of issues relevant to local communities. Patel attended 10 Friday congregational prayers and sermons from 2003 to 2004. He discovered that whereas the first sermon generally dealt with religious instruction, the second focused on social and political commentary. In their first post-invasion sermons, many preachers condemned pervasive looting and called on followers to return stolen goods. They suggested that looted property could be returned anonymously to designated locations. What the sermon message really did was coordinate listeners on the way to punish looters by shunning them. Whereas there is a cost to shunning, it is relatively low, especially if everyone else is also shunning for the same reason. Looters who heard this message knew they faced a choice: hand over their looted goods or keep them and risk being shunned. People did heed the preachers’ message to return the loot, but they only tended to return looted goods of low value. Those who took valuable items kept them, even at the risk of being shunned by some of their neighbors. Looting in Basra therefore continued, despite the response to the imams’ decree.

Societal trust in Iraq was low in 2003. Clusters of strong ties were small and connected by only a few weak ties. This kind of network structure inhibits the generation of widespread common knowledge. But there were other potential leaders in Iraqi society, such as tribal sheikhs and local notables. Why did these alternative forms of leadership fail when the mosque sermons succeeded? Amid anarchy, the imams of Friday mosques in cities had critical advantages over rival claimants to authority: they not only had a preexisting religious legitimacy that got people to attend sermons and a pious public receptive to their messages but also an unparalleled ability to render messages as common knowledge to the people living in a specific geographic area, on a routine and predictable schedule.

Finally, why were the Shi’ia so successful? Because the Shi’ia are more hierarchically organized than the Sunni, their grand ayatollahs had great influence over local imams. In consequence, their messages were internally consistent, and this consistency is key to the resolution of coordination dilemmas. Yet, common knowledge was insufficient to address the contribution problem. This would have required coercion or ample resources, neither of which the Shi’ia clergy possessed.

The bottom line is that information control plays a key role in accounting for variations in Iraqi social order: the Shi’ia had more of it, the Sunni less. Whereas we are all familiar with the malign effects of information control in authoritarian countries like China, this book reminds us that it can also have benign effects for collective action. This aligns with the claim that the rise of print media helped make national communities imaginable (Benedict Anderson, Imagined Communities: Reflections on the Origin and Spread of Nationalism, 1983). Likewise, information control accounts for the variable success of eighteenth-century mass mutinies in the Royal Navy (Steven Pfaff and Michael Hechter, The Genesis of Rebellion: Governance, Grievance and Mutiny in the Age of Sail, 2020).

After spending several months in Iraq, it became clear to Patel that foreign scholars and reporters were at high risk of being kidnapped for ransom and, in many cases, killed. This forced him to suspend his fieldwork prematurely and leave the country. As a result, the study suffers from some notable evidentiary limitations. Among other things, it has nothing at all to say about gender, and it lacks systematic data on intergroup stratification and social networks. Despite this, we should be grateful that Patel has managed to produce such an insightful account.