Making Democracy Work by Robert Putnam - Summary and Notes

MAY 2, 2019

Making Democracy Work: Civic Traditions in Modern Italy by Robert Putnam, Robert Leonardi and Raffaella Nanetti was published in 1993. It was a landmark work that launched the field of social capital research. It was also the more hard-core academic precursor to Putnam’s even better known Bowling Alone which focused on the decline of community (and social capital) in the US. It was a seminal work in a variety of ways: from its use of statistical methods (as well as traditional descriptive and historical materials) to its overall thesis about the importance of social capital (or what I would call cultural factors).

It reported the results of a nearly 20 year research program which tracked the impact of a major political reform in Italy that started in 1975. This reform shifted power from the central government to the regions in several key areas such as education and healthcare, which resulted in the regions controlling between one-quarter and one-third of of the national budget by 1977.1

This was a rare natural experiment: how would an identical shift in money and authority play out in the twenty different regions of Italy? In theory, moving power closer to the people should have improved governance. The results were very different and utterly fascinating:

Crudely put, in regions (primarily in the North) which had a pre-existing culture of high trust, political engagement and “horizontal” connections, the reforms resulted in better governance and better outcomes. By contrast, in regions (primarily in the South) which had a culture of low trust, low engagement and “vertical” connection, the reforms had little affect.

From my perspective, the book is so important because it is exceptional evidence of the importance of culture (collective being). Specifically, it supports two major theses of mine:

  • “Culture as dark matter”: that culture is a poorly understood but central explanatory variable for many of the most interesting and important social and political questions.
  • “The primacy of culture”: that culture has primacy over both technology and institutions in determining and explaining societal performance in key areas (ranging from the economy to politics). I’ve written more on this here.

At the very least, the book provides incontrovertible evidence for the key role that culture plays in mediating — and even determining — the impact of political institutions on governance. This means the same formal rules can have highly divergent outcomes in different societies.2

Table of contents

Key points

The formal and informal rules of the game in a society, once established, are self-reinforcing even when socially inefficient.


Using the term “institution” in a broad sense to mean “the rules of the game in a society,” [formal or informal?] North points out that institutional patterns are self-reinforcing, even when they are socially inefficient. First, it is almost always easier for an individual agent to adapt to the existing rules of the game than to seek to change them. Indeed, those rules tend to induce the rise of organizations and groups with a stake in their inefficiencies. Second, once development has been set on a particular course, organizational learning, cultural habits, and mental models of the social world reinforce that trajectory. Cooperation or shirking and exploitation become ingrained [and automatic]. Informal norms and culture change more slowly than formal rules, and tend to remold those formal rules, so that the external imposition of a common set of formal rules will lead to widely divergent outcomes. All of these hypotheses are consistent with the deep continuities traced in Chapter 5. [emphasis added]

The logic of the book


Each chapter in this book has begun with one question and ended with another. Chapter 2 began with “How did the new regional institutions affect the practice of politics?” and ended with “How successful was each institution at governing?” Chapter 3 answered that question, leading us naturally to ask “Why were some so much more successful than others?” Chapter 4 traced differences in performance to differences in civic engagement, which in turn raised the question, “Where did those differences in civic-ness come from?” Chapter 5 traced those differences to distinctive traditions that have endured for nearly a thousand years, posing the puzzle, “How could such differences have proved so stable?” Chapter 6 has explicated the vicious and virtuous circles that have led to contrasting, path-dependent social equilibria.

This explanation, however persuasive, poses starkly yet another question: “Why did the North and South get started on such divergent paths in the eleventh century? [which is not answered…]

Social contexts condition [largely determine?] the effectiveness of formal institutions.


This is one lesson gleaned from our research: Social context and history profoundly condition the effectiveness of institutions. Where the regional soil is fertile, the regions draw sustenance from regional traditions, but where the soil is poor, the new institutions are stunted. [emphasis in original]

Chapter Four: Explaining Performance

Civic participation is necessary to overcome mass poverty — technologies, institutions, policies and investments are not enough.


More recently, an independent line of research has reinforced the view that associationism is a necessary precondition for effective self-government. Summarizing scores of case studies of Third World development, Milton Esman and Norman Uphoff conclude that local associations are a crucial ingredient in successful strategies of rural development:

A vigorous network of membership organizations is essential to any serious effort to overcome mass poverty under the conditions that are likely to prevail in most developing countries for the predictable future…While other components—infrastructure investments, supportive public policies, appropriate technologies, and bureaucratic and market institutions—are necessary, we cannot visualize any strategy of rural development combining growth in productivity with broad distribution of benefits in which participatory local organizations are not prominent.

Unhappily from the point of view of social engineering, Esman and Uphoff find that local organizations “implanted” from the outside have a high failure rate. The most successful local organizations represent indigenous, participatory initiatives in relatively cohesive local communities.

Although Esman and Uphoff do not say so explicitly, their conclusions are quite consistent with Banfield’s interpretation of life in Montegrano, “the extreme poverty and backwardness of which is to be explained largely (but not entirely) by the inability of the villagers to act together for their common good or, indeed, for any end transcending the immediate material interest of the nuclear family.” Banfield’s critics have disagreed with his attribution of this behavior to an “ethos,” but they have not dissented from his description of the absence of collaboration in Montegrano, the striking lack of “deliberate concerted action” to improve community conditions.

Putnam uses four highly correlated factors to assess the levels of civic community in Italy’s regions.


Data from 1950s - 1980s Italy shows the following four factors were highly correlated:

  • Preference voting (inversely correlated with other factors)
  • Referendum turnout
  • Newspaper readership
  • Participation in sports and cultural associations

Figure 4.4 charts the levels of civic community in Italy’s twenty regions according to these factors.

There was a strong correlation between civic participation and institutional performance.


The strength of this relationship appears with stark clarity in Figure 4.5 [below]. Not only does “civic-ness” distinguish the high performance regions in the upper right-hand quadrant from the laggards in the lower left-hand quadrant, but even the more subtle differences in performance within each quadrant are closely tied to our measure of community life. In this respect, the predictive power of the civic community is greater than the power of economic development, as summarized in Figure 4.2. The more civic a region, the more effective its government.

Religiosity and clericalism were inversely correlated with civic participation in Italy in the 1970s and 80s.


Organized religion, at least in Catholic Italy, is an alternative to the civic community, not a part of it. Throughout Italian history, the presence of the Papacy in Rome has had a powerful effect on the Italian Church and its relationship with civic life. For more than thirty years after Unification, the Papal non expedit forbade all Catholics from taking part in national political life, although after World War II the Church became a senior partner of the Christian Democratic party. Despite the reforms of the Second Vatican Council and the flowering of many divergent ideological tendencies among the faithful, the Italian Church retains much of the heritage of the Counter-Reformation, including an emphasis on the ecclesiastical hierarchy and the traditional virtues of obedience and acceptance of one’s station in life. Vertical bonds of authority are more characteristic of the Italian Church than horizontal bonds of fellowship.

At the regional level, all manifestations of religiosity and clericalism—attendance at Mass, religious (as opposed to civil) marriages, rejection of divorce, expressions of religious identity in surveys—are negatively correlated with civic engagement. (Figure 4.12 summarizes this pattern.) At the individual level, too, religious sentiments and civic engagement seem to be mutually incompatible. Of those Italians who attend Mass more than once a week, 52 percent say they rarely read a newspaper and 51 percent say they never discuss politics; among their avowedly irreligious compatriots, the equivalent figures are 13 percent and 17 percent. Churchgoers express greater contentment with life and with the existing political regime than other Italians. They seem more concerned about the city of God than the city of man.

Citizens of civic communities had higher life satisfaction.


…citizens in civic regions are happier with life in general than are their counterparts in less civic regions. In a series of nationwide surveys between 1975 and 1989, roughly twenty-five thousand people were asked whether they were “very satisfied, fairly satisfied, not very satisfied, or not at all satisfied with the life you lead.” Figure 4.14 shows that citizens of civic regions are much more satisfied with life. Happiness is living in a civic community.

The Northern regions were characterized by horizontally organized public life and high levels of civic engagement; the Southern regions by hierarchically organized public life and far less engagement.


We can summarize our discoveries so far in this chapter rather simply.

Some regions of Italy have many choral societies and soccer teams and bird-watching clubs and Rotary clubs. Most citizens in those regions read eagerly about community affairs in the daily press. They are engaged by public issues, but not by personalistic or patron-client politics. Inhabitants trust one another to act fairly and to obey the law. Leaders in these regions are relatively honest. They believe in popular government, and they are predisposed to compromise with their political adversaries. Both citizens and leaders here find equality congenial. Social and political networks are organized horizontally, not hierarchically. The community values solidarity, civic engagement, cooperation, and honesty. Government works. Small wonder that people in these regions are content!

At the other pole are the “uncivic” regions, aptly characterized by the French term incivisme. Public life in these regions is organized hierarchically, rather than horizontally. The very concept of “citizen” here is stunted. From the point of view of the individual inhabitant, public affairs is the business of somebody else—i notabili, “the bosses,” “the politicians”—but not me. Few people aspire to partake in deliberations about the commonweal, and few such opportunities present themselves. Political participation is triggered by personal dependency or private greed, not by collective purpose. Engagement in social and cultural associations is meager. Private piety stands in for public purpose. Corruption is widely regarded as the norm, even by politicians themselves, and they are cynical about democratic principles. “Compromise” has only negative overtones.

Laws (almost everyone agrees) are made to be broken, but fearing others’ lawlessness, people demand sterner discipline. Trapped in these interlocking vicious circles, nearly everyone feels powerless, exploited, and unhappy. All things considered, it is hardly surprising that representative government here is less effective than in more civic communities.

This discovery poses two new and important questions: How did the civic regions get that way? and How do norms and networks of civic engagement undergird good government? We shall address those questions in the two chapters that follow, but first a few words about other potential explanations for the success and failure of the regional governments.

Chapter 5: Tracing the Roots of the Civic Community

The origins of these regional differences in civic participation can be found in early medieval Italy.


Our inquiry into the performance of Italian regional governments in the 1970s and 1980s has pinpointed the unique character of civic life in some regions. Following that thread now draws us deep into the contrasting pasts of Italy’s regions. Our story begins with a momentous time of transition on the Italian peninsula nearly a thousand years ago, as Italians were emerging from that obscure era justly termed the Dark Ages. Early medieval Italy, when our story opens, was closer to ancient Rome than to our own times, not only chronologically but also in everyday ways of life.

Nevertheless, social patterns plainly traceable from early medieval Italy to today turn out to be decisive in explaining why, on the verge of the twenty-first century, some communities are better able than others to manage collective life and sustain effective institutions.

The autocracy or republicanism of Italian regions at the beginning of the 14th century corresponds closely to the strength of civic tradition in the 20th century.


Figure 5.1 shows the various regimes that characterized Italy at the beginning of the fourteenth century. The map clearly reveals four bands across the peninsula, corresponding to differing degrees of republicanism and autocracy.

From south to north, they are as follows:

• The feudal monarchy founded by the Normans in the Mezzogiorno;

• The Papal States with their variegated mixture of feudalism, tyranny, and republicanism;

• The heartland of republicanism, those communes which had retained republican institutions into the fourteenth century; and

• The erstwhile republican areas further north that had, by this time, fallen prey to signorial rule.

The parallel between this pattern and the distribution of civic norms and networks in the 1970s, as displayed in Figure 4.4, is remarkable. The southern territories once ruled by the Norman kings constitute exactly the seven least civic regions in the 1970s. Almost as precisely, the Papal States (minus the communal republics that lay in the northern section of the Pope’s domains) correspond to the next three or four regions up the civic ladder in the 1970s. At the other end of the scale, the heartland of republicanism in 1300 corresponds uncannily to the most civic regions of today, followed closely by the areas still further north in which medieval republican traditions, though real, had proved somewhat weaker.


Although these groups were not overtly political, they often came to have political affinities with one or another of the tendences of French political life. Social interaction and the exercise of organizational skills widened the cultural horizons of the participants and quickened their political awareness and (eventually) their political involvement. “For the lower classes of Provence at this period, to set themselves up as a chambrée was, just as much and perhaps even more than learning to read, to become accessible to whatever was new, to change and to independence.” Agulhon’s painstaking reconstruction of life in several southern French villages of this era has shown how this cultural mobilization in the years after 1830 contributed directly to the great political mobilizations of 1848. [emphasis added]

Civic solidarity was not equally extensive or successful everywhere following national unification in Italy — unification did little to inculcate civic habits in the South.

(p.142 ff.)

Although mutual aid societies, cooperatives, and other manifestations of civic solidarity were established in all sectors of the economy and in all parts of the peninsula, they were not equally extensive or equally successful everywhere. In north-central Italy, mirroring almost precisely that area where the communal republics had longest endured five centuries earlier (and where the most civic regions would be found in the 1970s), the medieval traditions of collaboration persisted, even among poor peasants. “A significant network of social and economic obligations, particularly in the countryside, is formed by the recognition of neighborhoodship. Between vicini [neighbors] there is continuous mutual aid and exchange of services.

Sharecropping families had in fact developed a rich network of exchanges and mutual aid: typical of these was the aiutarella, the exchange of labour between families at crucial moments in the agricultural calendar, such as at threshing time. On a cultural level there was also the important practice of the veglia. During the long winter evenings, families would gather in the stables or kitchens of the farmhouses, to play cards and games, to knit and to mend, to listen to and tell stories. Participation in the veglia was not segregated family by family. Rather…it involved rotating hospitality and a complex system of visiting.

By stark contrast, an 1863 report concluded that in Calabria, a desolate land locked in the southern traditions of authoritarian rule (and destined to rank as the least civic of all the regions in the 1970s), there were “no associations, no mutual aid; everything is isolation. Society is held up by the natural civil and religious bonds alone; but of economic bonds there is nothing, no solidarity between families or between individuals or between them and the government."

In areas of Italy long subjected to autocratic rule, national unification did little to inculcate civic habits:

In all classes the absence of a community sense resulted from a habit of insubordination learned in centuries of despotism. Even the nobles had become accustomed to obstruction, and thought that governments could be fairly cheated without moral obliquity so long as the cheating were successful…Instead of recognizing that taxes had to be paid, the attitude was rather that if one group of people had discovered a profitable evasion, then other groups had better look to their own interests. Each province, each class, each industry thus endeavored to gain at the expense of the community.

Southern agriculture, although complicated by a crazy-quilt patchwork of landholding, was typified by the latifondo, or large estate, worked by impoverished peasants:

The peasants were in constant competition with each other for the best strips of land on the latifondo, and for what meagre resources were available. Vertical relationships between patron and client, and obsequiousness to the landlord, were more important than horizontal solidarities. As Bevilacqua has written for the period 1880-1920: The peasant classes were more at war amongst themselves than with the other sectors of rural society; a war which fed off a terrain of recurring and real contrasts, both economic, psychological and cultural.’ That such attitudes triumphed can only be understood in the context of a society which was dominated by distrust…[T]he weight of the past, when combined with the failures of state authority after 1860 and the disastrous peasant-landlord relations…produced a society where fede pubblica (civic trust) had been reduced to a minimum: ‘chi ara diritto, muore disperato’ (he who behaves honestly comes to a miserable end) was a noted Calabrian proverb.

The primeval mistrust that rent the social fabric in these regions was, in fact, captured in innumerable proverbs:

  • “Damned is he who trusts another.”
  • “Don’t make loans, don’t give gifts, don’t do good, for it will turn out bad for you.”
  • “Everyone thinks of his own good and cheats his companion.”
  • “When you see the house of your neighbor on fire, carry water to your own.”

In the Mezzogiorno, above all, observed Pasquale Villari in 1883, “One feels too much the I and too little the ‘we’.”

The combination of impoverishment and mutual distrust forestalled horizontal solidarity and fostered what Banfield has called “amoral familism.” “In an overcrowded latifundia economy,” recalls Sidney Tarrow, “the village square was an employment bureau where the fortunate few found a day’s labor while their bitter neighbors looked on.“101 “Each became different from the other; he came to find himself ever more involved in a bitter battle of competition to obtain work or to be able to cultivate a little land, and thus participated less in class solidarity and in the life of the collectivity, and appeared exclusively interested in the progress of himself and his family.” Mark the contrast with those landless braccianti of civic Emilia-Romagna who, facing a similar dilemma, formed a voluntary cooperative to seek shared work.

As Tarrow, among other scholars, has emphasized, the South was not (and is not) apolitical or asocial. On the contrary, political cunning and social connections have long been essential to survival in this melancholy land. The relevant distinction is not between the presence and absence of social bonds, but rather between horizontal bonds of mutual solidarity and vertical bonds of dependency and exploitation. The southerner—whether peasant or city-dweller, whether in the old Hapsburg kingdom of the sixteenth century, the new Italian kingdom of the nineteenth century, or (as we saw in the previous chapter) the regional politics of the late twentieth century—has sought refuge in vertical bonds of patronage and clientelism, employed for both economic and political ends:

Clientelism is the product of a disorganic society and tends to preserve social fragmentation and disorganization. . . . Turiello [a close observer of the Mezzogiorno in the 1880s] refers again and again to the ‘excessive isolation (scioltezza) of individuals’ who feel no moral bond outside the family, and views the clientele as the specific remedy for a disjointed society. The clientele, he wrote, are ‘the only associations which actually show real operative energy in a civil society which has been divided within itself for centuries’ and in which people unite not on the basis of mutual trust but only when forced by necessity.

The new institutions of the unified nation-state, far from homogenizing traditional patterns of politics, were themselves pulled ineluctably into conformity with those contrasting traditions, just as the regional governments after 1970 would be remolded by these same social and cultural contexts:

In the 1870s, one can say that the most advanced provinces of Italy already were expressing their preferences through free institutions or associations—agrarian associations, mutual aid societies, chambers of commerce, savings banks—while the southern ones were more inclined to make use of personal

The southern feudal nobility—along with elements of the urban professional classes who had acquired common land and Church properties expropriated by the newly-forged Italian state—used private violence, as well as their privileged access to state resources, to reinforce vertical relations of dominion and personal dependency and to discourage horizontal solidarity. Leopoldo Franchetti, a civic-minded Tuscan landowner who in 1876 authored a remarkable analysis of social conditions in Sicily, concluded:

The landed classes ruled from on high the network of clientelistic structures at various levels and maintained contact for their own advantage with the supreme representative organs of the country…Every local notable in his jurisdiction of power was the head of a network of persons of the most diverse social conditions, who depended on him for their economic survival and social prestige and who furnished him legal support in terms of electoral suffrage and illegal support in the recourse to private violence in defense of his particular interests, in a rigorously hierarchical relationship of para-feudal dependence.

For wretchedly vulnerable peasants, recourse to patron-client ties was a sensible response to an atomized society. One recent account of the “moral economy” of life on a latifondo estate in Calabria in the first half of the nineteenth century recounts that peasants in fact feared exclusion from the patron-client system, for it alone assured their physical subsistence, along with the necessary intermediation with distant state authorities and a primitive kind of private welfare program (pensions for widows and orphans and occasional “gratuities”), so long as the peasant-client remained obedient, “faithful” to the estate, and “available” to perform chores as required by the landlord-patron. In the absence of horizontal solidarity, as exemplified by mutual aid societies, vertical dependence is a rational strategy for survival—even when those who are dependent recognize its drawbacks.

The dispossessed southern peasantry did not always endure their fate in silence. Violent protest movements, including chronic brigandage, flared like heat lightning across the Mezzogiorno landscape throughout the late nineteenth century. However, these anarchic episodes (unlike the contemporary urban and rural strike waves in the center and north of the country) produced no permanent organization and left little residue of collective solidarity. The South remained, as the great Communist intellectual Antonio Granisci lamented, “a great social disaggregation.” Despite the occasional violent revolts, “it is more important to emphasize the more usual passive reaction of resigned submission. For it is this submission that provides the historical background to the acceptance of the arrogation of power by individuals, viz. the mafiosi, by the rest of the population.”

Southern Italy after the risorgimento was stuck in a vertical clientilistic social equilibrium.


The North after risorgimento rapidly returned to its rich network of horizontal ties established 700 years earlier with a dense and diverse set of mutual aid societies, cooperatives as well as more informal practices such as the aiutarella (mutual assistance at harvest). The two new parties of the socialists and catholics reflected and built on these traditions — whilst politically in competition at a socio-cultural level they drew on the same roots of collective being: high trust, mutual aid and solidarity and a rich horizontal network of ties and support.

But in the South, the centuries-old (or even millenial-old) vertical clientilism continued even under these new, more liberal conditions. Here, with no civic tradition to draw on, no civic tradition was created and politics remained a case of feudalistic vertical assistance and dependency. Distrust between all parties remained high and all groups detested and distrusted the state. Day to day culture was a case of amoral familism of each for his own, of brutal competition between impoverished peasants fighting for scraps of work and food on the feudal latifundia.

An important aside: this is real evidence for collective being because these “amoral familial” peasants of the south once they migrated north after the second world war would to a man become active members of the communist party in Emilia romagna and Lombardy (see Norman Lewis in the “Honoured Society”). It would be interested to study the trust level of, say, peasants from Italy when they migrated northwards.

And we have a perfect example of the primacy of “being”, in this case collective being as culture, over structure (institutions and rules of governance):

The new institutions of the unified nation-state, far from homogenizing traditional patterns of politics, were themselves pulled ineluctably into conformity with those contrasting traditions, just as the regional governments after 1970 would be remolded by these same social and cultural contexts.

Careful counting backs up the anecdotal picture of the North / South divide.


The available statistical evidence confirms the stark differences from region to region in associationism and collective solidarity a century ago. By 1904, for example, Piedmont had more than seven times as many mutual aid societies as Puglia, in proportion to population. By 1915, cooperative membership per capita was eighteen times greater in Emilia-Romagna than in Molise. These regional concentrations depended in turn on the pre-existing traditions of collaboration and sociability. Often an ancient guild found reincarnation in a “pious society” in the eighteenth century, which in turn evolved into a mutual aid society, which encouraged cooperatives, which subsequently formed the basis for labor unions and mass-based political parties.

Regional differences in civic involvement changed little between the 1860s and 1970s, despite huge social upheavals.


Even a cursory comparison of Figure 5.2 with Figure 4.4 attests to the astonishing constancy of regional traditions of civic involvement through more than a century of vast social change. A more convenient way of visualizing this continuity is provided in Figure 5.3, which arrays the almost perfect correlation between our Civic Community Index for the 1970s and 1980s and our comparable measure of civic involvement a century earlier. Despite the massive waves of migration, economic change, and social upheaval that have swept along the peninsula in the intervening decades, contemporary civic norms and practices recapitulate regional traditions that were well established long ago.


The success of regional governments in 1980s Italy could have been predicted from civic traditions a century earlier.


Figure 5.4 presents the correlation between institutional performance in the 1980s and civic traditions in 1860-1920. The pattern is stark: One could have predicted the success or failure of regional government in Italy in the 1980s with extraordinary accuracy from patterns of civic engagement nearly a century earlier.

The correlation between civics and economics primarily reflects the impact of civics on economics, not the reverse.


Figure 5.6 synthesizes our findings. Arrow b (the effect of economics on civics) is nonexistent, while arrow c (the effect of civics on economics) is strong—stronger even than arrow d. Moreover, arrow a (civic continuity) is very strong, while arrow d (socioeconomic continuity) is generally weak. A region’s chances of achieving socioeconomic development during this century have depended less on its initial socioeconomic endowments than on its civic endowments. Insofar as we can judge from this simple analysis, the contemporary correlation between civics and economics reflects primarily the impact of civics on economics, not the reverse.

Almost precisely the same regions were characterized by civic involvement in the late 20th century and in the 12th century.


Despite this whirl of change, however, the regions characterized by civic involvement in the late twentieth century are almost precisely the same regions where cooperatives and cultural associations and mutual aid societies were most abundant in the nineteenth century, and where neighborhood associations and religious confraternities and guilds had contributed to the flourishing communal republics of the twelfth century. And although those civic regions were not especially advanced economically a century ago, they have steadily outpaced the less civic regions both in economic performance and (at least since the advent of regional government) in quality of government. The astonishing tensile strength of civic traditions testifies to the power of the past.

But why is the past so powerful? What virtuous circles in the North have preserved these traditions of civic engagement through centuries of radical social, economic, and political change? What vicious circles in the South have reproduced perennial exploitation and dependence? To address such questions we must think not merely in terms of cause and effect, but in terms of social equilibria. To that task we turn in the next chapter.

The presence or absence of many forms of social capital are self-fulfilling and are reinforced through virtuous or vicious circles.


Most forms of social capital, such as trust, are what Albert Hirschman has called “moral resources”—that is, resources whose supply increases rather than decreases through use and which become depleted if not used. The more two people display trust towards one another, the greater their mutual confidence. Conversely:

Deep distrust is very difficult to invalidate through experience, for either it prevents people from engaging in the appropriate kind of social experiment or, worse, it leads to behaviour which bolsters the validity of distrust itself…Once distrust has set in it soon becomes impossible to know if it was ever in fact justified, for it has the capacity to be self-fulfilling.

Other forms of social capital, too, such as social norms and networks, increase with use and diminish with disuse. For all these reasons, we should expect the creation and destruction of social capital to be marked by virtuous and vicious circles.

One special feature of social capital, like trust, norms, and networks, is that it is ordinarily a public good, unlike conventional capital, which is ordinarily a private good. “As an attribute of the social structure in which a person is embedded, social capital is not the private property of any of the persons who benefit from it.” Like all public goods, social capital tends to be undervalued and undersupplied by private agents. For example, my reputation for trustworthiness benefits you as well as me, since it enables us both to engage in mutually rewarding cooperation. But I discount the benefits to you of my being trustworthy (or the costs to you of my being untrustworthy) and thus I underinvest in trust formation. This means that social capital, unlike other forms of capital, must often be produced as a by-product of other social activities.

Trust is an essential component of social capital. As Kenneth Arrow has observed, “Virtually every commercial transaction has within itself an element of trust, certainly any transaction conducted over a period of time. It can be plausibly argued that much of the economic backwardness in the world can be explained by the lack of mutual confidence.”

Chapter 6: Social Capital and Institutional Success

Horizontal networks outperform vertical networks in solving collective action problems.


The fact that vertical networks are less helpful than horizontal networks in solving dilemmas of collective action may be one reason why capitalism turned out to be more efficient than feudalism in the eighteenth century, and why democracy has proven more effective than autocracy in the twentieth century.

This is an absolutely crucial point if it is the case, as I think, that solving collective action problems is the great engine of progress. What I would draw out (or add) to Putnam’s analysis is how horizontal networks, whilst perhaps initially based on some kind of rational calculation, over time become embedded in (semi)-automatic sets of beliefs, norms and values — in culture.

Aside: I feel Putnam et al are sometimes too enamoured of social structure — “networks of civic engagement” — versus the culture associated to that. Of course, structure and culture co-evolve.

Networks of civic engagement, like the neighborhood associations, choral societies, cooperatives, sports clubs, mass-based parties, and the like examined in Chapters 4 and 5, represent intense horizontal interaction. Networks of civic engagement are an essential form of social capital: The denser such networks in a community, the more likely that its citizens will be able to cooperate for mutual benefit. Why, exactly, do networks of civic engagement have this powerfully beneficial side-effect?

Therefore we should see a correlation between good government and horizontally ordered groups.

(p. 175-6)

If horizontal networks of civic engagement help participants solve dilemmas of collective action, then the more horizontally structured an organization, the more it should foster institutional success in the broader community. Membership in horizontally ordered groups (like sports clubs, cooperatives, mutual aid societies, cultural associations, and voluntary unions) should be positively associated with good government. Since the organizational realities of political parties vary from party to party and region to region (vertical in some places, horizontal in others), we should expect party membership as such to be unrelated to good government. Membership rates in hierarchically ordered organizations (like the Mafia or the institutional Catholic Church) should be negatively associated with good government; in Italy, at least, the most devout churchgoers are the least civic-minded. All these expectations are consistent with the evidence of this study, as we saw in Chapters 4 and 5. Good government in Italy is a by-product of singing groups and soccer clubs, not prayer.

In summary, there are two social equilibria which can be established through historical turning points: one is characterized by horizontal ties, high cooperation and high performance and the other by vertical ties, low cooperation and low performance.

  • Two social equilibria are possible.
  • These are: horizontal ties, high cooperation, high performance and vertical ties, low cooperation, low performance.
  • The latter is an example of a local equilibria (once it’s exists it’s difficult to change and it becomes rational to act within it).
  • Which equilibria is established depends on historical accident; “path dependence.”
  • I would emphasize the greater role of culture vs e.g. social structure. Putnam points to this (see emphasis below)


Historical turning points thus can have extremely long-lived consequences. As the “new institutionalists” have emphasized, institutions—and we would add, the social settings that condition their operation [aka culture]—evolve through history, but they do not reliably reach unique and efficient equilibria.

See the excerpt below for more detail on these points.

(pp.177 - 179)

In all societies, to summarize our argument so far, dilemmas of collective action hamper attempts to cooperate for mutual benefit, whether in politics or in economics. Third-party enforcement is an inadequate solution to this problem. Voluntary cooperation (like rotating credit associations) depends on social capital. Norms of generalized reciprocity and networks of civic engagement encourage social trust and cooperation because they reduce incentives to defect, reduce uncertainty, and provide models for future cooperation. Trust itself is an emergent property of the social system, as much as a personal attribute. Individuals are able to be trusting (and not merely gullible) because of the social norms and networks within which their actions are embedded.

Stocks of social capital, such as trust, norms, and networks, tend to be self-reinforcing and cumulative. Virtuous circles result in social equilibria with high levels of cooperation, trust, reciprocity, civic engagement, and collective well-being. These traits define the civic community. Conversely, the absence of these traits in the uncivic community is also selfreinforcing. Defection, distrust, shirking, exploitation, isolation, disorder, and stagnation intensify one another in a suffocating miasma of vicious circles. This argument suggests that there may be at least two broad equilibria toward which all societies that face problems of collective action (that is, all societies) tend to evolve and which, once attained, tend to be self-reinforcing.

The strategy of “never cooperate” is a stable equilibrium, for reasons that are well explicated in standard accounts of the prisoner’s dilemma. Once trapped in this situation, no matter how exploitative and backward, it is irrational for any individual to seek a more collaborative alternative, except perhaps within the immediate family. The “amoral familism” that Banfield observed in the Mezzogiorno is, in fact, not irrational, but the only rational strategy for survival in this social context. Actors in this social equilibrium may well realize that they are worse off than they would be in a more cooperative equilibrium, but getting to that happier equilibrium is beyond the power of any individual.

In this setting, we should expect the Hobbesian, hierarchical solution to dilemmas of collective action—coercion, exploitation, and dependence—to predominate. This oppressive state of affairs is clearly inferior to a cooperative outcome, for it dooms the society to self-perpetuating backwardness. Nevertheless, it is preferable to a purely anarchic “state of nature,” as has also been clear to southern Italians from medieval to modern times. This Hobbesian outcome has at least the virtue that it is attainable by individuals who are unable to trust their neighbors. Minimal security, no matter how exploitative and inefficient, is not a contemptible objective for the powerless.

The difficulty of solving dilemmas of collective action in this Hobbesian equilibrium means that society is worse off than in a cooperative outcome. This shortfall is probably even greater in a complex industrial or postindustrial context, where impersonal cooperation is essential, than a simple agricultural society. As Douglass North, an astute theorist of economic history, has observed, “the returns on opportunism, cheating, and shirking rise in complex societies.” Thus, the importance of social capital (to inhibit opportunism, cheating, and shirking) increases as economic development proceeds. This may help explain why the gap between the civic North and the uncivic South has widened over the last century.

Authoritarian government, patron-clientelism, extralegal “enforcers,” and the like represent a second-best, “default” solution: Through them, individuals can find some refuge from the war of all against all, without pursuing the impossible dream of cooperation. Force and family provide a primitive substitute for the civic community. This equilibrium has been the tragic fate of southern Italy for a millennium.

Given an adequate stock of social capital, however, a happier equilibrium is also attainable. Assuming that prisoner’s dilemmas are iterated or interconnected (as they are in a civic community), “brave reciprocity” is also a stable equilibrium strategy, as the game theorist Robert Sugden has recently shown: “Cooperate with people who cooperate with you (or who cooperate with people like you), and don’t be the first to defect.” Sugden shows, specifically, that in what he calls “the mutual-aid game” (a formalization of the implicit bargaining that underlies mutual aid societies, cooperatives, rotating credit associations, Hume’s game of the two farmers, and so on) cooperation can be sustained indefinitely. To be sure, even in an indefinitely repeated mutual-aid game, “always defect” is also a stable equilibrium, but if a society can somehow move toward the cooperative solution, it will be self-reinforcing. In a society characterized by dense networks of civic engagement, where most people abide by civic norms, it is easier to spot and punish the occasional “bad apple,” so that defection is riskier and less tempting.

Sugden’s analysis leads to the conclusion that both “always defect” and “reciprocate help” are contingent conventions—that is, rules that have evolved in particular communities and, having so evolved, are stable, but that might have evolved otherwise. In other words, reciprocity/trust and dependence/exploitation can each hold society together, though at quite different levels of efficiency and institutional performance. Once in either of these two settings, rational actors have an incentive to act consistently with its rules. History determines which of these two stable outcomes characterizes any given society.

Historical turning points thus can have extremely long-lived consequences. As the “new institutionalists” have emphasized, institutions—and we would add, the social settings that condition their operation [aka culture]—evolve through history, but they do not reliably reach unique and efficient equilibria. History is not always efficient, in the sense of weeding out social practices that impede progress and encourage collective irrationality. Nor is this inertia somehow attributable to individual irrationality. On the contrary, individuals responding rationally to the social context bequeathed to them by history reinforce the social pathologies.

Recent theorists of economic history have dubbed this feature of social systems “path dependence”: where you can get to depends on where you’re coming from, and some destinations you simply cannot get to from here. Path dependence can produce durable differences in performance between two societies, even when the formal institutions, resources, relative prices, and individual preferences in the two are similar. The implications of this point for economic (and political) development are profound:

“If the process by which we arrive at today’s institutions is relevant and constrains future choices, then not only does history matter but persistent poor performance and long-run divergent patterns of development stem from a common source.”

Douglass North has illustrated this point by tracing the post-colonial experiences of North and South America to their respective colonial legacies. After independence, both the United States and the Latin republics shared constitutional forms, abundant resources, and similar international opportunities; but North Americans benefited from their decentralized, parliamentary English patrimony, whereas Latin Americans were cursed with centralized authoritarianism, familism, and clientelism that they inherited from late medieval Spain. In our language, the North Americans inherited civic traditions, whereas the Latin Americans were bequeathed traditions of vertical dependence and exploitation. The point is not that the preferences or predilections of individual North and South Americans differed, but that historically derived social contexts presented them with a different set of opportunities and incentives. The parallel between this North-South contrast and our Italian case is striking.

Using the term “institution” in a broad sense to mean “the rules of the game in a society,” North points out that institutional patterns are self-reinforcing, even when they are socially inefficient. First, it is almost always easier for an individual agent to adapt to the existing rules of the game than to seek to change them. Indeed, those rules tend to induce the rise of organizations and groups with a stake in their inefficiencies. Second, once development has been set on a particular course, organizational learning, cultural habits, and mental models of the social world reinforce that trajectory. Cooperation or shirking and exploitation become ingrained. Informal norms and culture change more slowly than formal rules, and tend to remold those formal rules, so that the external imposition of a common set of formal rules will lead to widely divergent outcomes. All of these hypotheses are consistent with the deep continuities traced in Chapter 5.

The question of what started the North and South on the path to different equilibria merits further research.


This explanation, however persuasive, poses starkly yet another question: “Why did the North and South get started on such divergent paths in the eleventh century?” The hierarchical Norman regime in the South is perhaps readily explained as the consequence of conquest by an unusually effective force of foreign mercenaries. More problematical and potentially more interesting are the origins of the communal republics. How did the inhabitants of north-central Italy first come to seek collaborative solutions to their Hobbesian dilemmas? The response to that question must await further research, not least because historians report that the answer seems lost in the mists of the Dark Ages. Our interpretation highlights the unique importance of trying to pierce those mists.

Culture and structure are interdependent, but culture once established is more influential than structure.


Social scientists have long debated what causes what—culture or structure. In the context of our argument this debate concerns the complicated causal nexus among the cultural norms and attitudes and the social structures and behavioral patterns that make up the civic community. Quite apart from the ambiguity of “culture” and “structure,” however, this debate is somewhat misplaced. Most dispassionate commentators recognize that attitudes and practices constitute a mutually reinforcing equilibrium. Social trust, norms of reciprocity, networks of civic engagement, and successful cooperation are mutually reinforcing. Effective collaborative institutions require interpersonal skills and trust, but those skills and that trust are also inculcated and reinforced by organized collaboration. Norms and networks of civic engagement contribute to economic prosperity and are in turn reinforced by that prosperity.

Ed: of course there is interdependence: culture and structure are an ecosystem, with each feeding the other. Neither, de facto, has precedence: in a tabula rasa state of culture, institutions would play a profound, perhaps determining, role; meanwhile a culture once in existence sets up institutions that support culture. However, it is clear from the book that culture (plus its associated informal structures), once in place, trumps (formal) structure. Take that statistic on newspaper reading in Emilia Romagna. Newspaper reading is not the product of any set of structures—it is a cultural artefact.

Cf. the material in footnote 85:

See, for example, Michael Thompson, Richard Ellis, and Aaron Wildavsky, Cultural Theory (San Francisco: Westview Press, 1990), p. 21: “Values and social relations are mutually interdependent and reinforcing: Institutions generate distinctive sets of preferences, and adherence to certain values legitimizes corresponding institutional arrangements. Asking which comes first or which should be given causal priority is a nonstarter.” See also Ronald Inglehart, “The Renaissance of Political Culture,” American Political Science Review 82 (1988): 1203-1230, who emphasizes the reciprocal linkages among political culture, economic development, and stable democracy. An older idiom traced institutional performance to “civic virtue,” and our emphasis on civic community echoes that approach. Classically, “the republic made the virtuous individual and the virtuous individual made the republic." [emphasis added] (Richard Vetterli and Gary Bryner, In Search of the Republic: Public Virtue and the Roots of American Government [Towata, N.J.: Rowman and Littlefield, 1987], p. 20.) In our terms, the civic community is a self-reinforcing equilibrium.

The social contract is vital in maintaining civic community.


Culture again (emphasis added):

The harmonies of a choral society illustrate how voluntary collaboration can create value that no individual, no matter how wealthy, no matter how wily, could produce alone. In the civic community associations proliferate, memberships overlap, and participation spills into multiple arenas of community life. The social contract that sustains such collaboration in the civic community is not legal but moral. The sanction for violating it is not penal, but exclusion from the network of solidarity and cooperation. Norms and expectations play an important role. As Thompson, Ellis, and Wildavsky put it, “Ways of life [ways of collective being] are made viable by classifying certain behaviors as worthy of praise and others as undesirable, or even unthinkable." A conception of one’s role and obligations as a citizen, coupled with a commitment to political equality, is the cultural cement of the civic community.

Social capital is key to collective action, political stability, government effectiveness and economic and social progress.

How prescient this was (written in 1993)—the future of Moscow was indeed Palermo (but even worse—a totalitarian Palermo).


Where norms and networks of civic engagement are lacking, the outlook for collective action appears bleak. The fate of the Mezzogiorno is an object lesson for the Third World today and the former Communist lands of Eurasia tomorrow, moving uncertainly toward self-government. The “always defect” social equilibrium may represent the future of much of the world where social capital is limited or nonexistent. For political stability, for government effectiveness, and even for economic [ed: and social] progress social capital may be even more important than physical or human capital. Many of the formerly Communist societies had weak civic traditions before the advent of Communism, and totalitarian rule abused even that limited stock of social capital. Without norms of reciprocity and networks of civic engagement, the Hobbesian outcome of the Mezzogiorno—amoral familism, clientelism, lawlessness, ineffective government, and economic stagnation—seems likelier than successful democratization and economic development. Palermo may represent the future of Moscow.

However, institutions do matter … plus they are something we can act on (culture is hard).

Two obvious points:

  • Institutions impact culture over time—even if the impact is more limited than one would hope.
  • Regionalism—making the “collective” smaller—helps with collective action problems, although they are still difficult.

The view that civic engagement is challenging to increase through institutional reform could lead to a minimization of efforts…


The civic community has deep historical roots. This is a depressing observation for those who view institutional reform as a strategy for political change. The president of Basilicata cannot move his government to Emilia, and the prime minister of Azerbaijan cannot move his country to the Baltic. “A theory of change that gives priority to ethos can have unfortunate consequences…It may lead to minimizing efforts at change because people are believed to be hopelessly enmeshed in an ethos.”

But regional reform in Italy shows that reforming formal institutions does induce cultural change, albeit gradually.


The full results of the regional reform, however, are far from an invitation to quietism. On the contrary, a second lesson of the regional experiment is (as Chapter 2 demonstrates) that changing formal institutions can change political practice. The reform had measurable and mostly beneficial consequences for regional political life. As institutionalists would predict, institutional changes were (gradually) reflected in changing identities, changing values, changing power, and changing strategies. These trends transpired in the South no less than the North. In both South and North, the new institution nurtured a more moderate, pragmatic, tolerant elite political culture. In both South and North, the reform altered old patterns of power and produced more genuine subnational autonomy than unified Italy had ever known. In both South and North, the reform itself generated pressures, both inside and outside the government, in support of further decentralization. In both South and North, the regional government is generally regarded by community leaders and ordinary voters as an improvement over the institutions it replaced—certainly more accessible and probably more effective. The regional reform allowed social learning, “learning by doing.” Formal change induced informal change and became self-sustaining.

  1. Making Democracy Work, p.22 ↩︎

  2. cf my comments that transplanting a Finnish school to e.g. the UAE won’t work because it isn’t the structure of the school (in itself) that makes the Finnish school good, it is the culture in which it operates. ↩︎