welcoming of strangers.” 18 The other is that America’s recent immi-
grants are largely Hispanics, as in the study discussed above: people from Latin America. Diversity depends not just upon numbers but on cultural distance between immigrants and indigenous populations. The cultural gap between Hispanics and other Americans looks to be smaller than that between immigrants to Europe from poor countries and indigenous Europeans. But would such a judgment of cultural difference merely be prejudice?
An ingenious objective way of measuring cultural distance is
by a language tree. Modern linguistics has constructed a global
THE SOCIAL CONSEQUENCES 77
language tree, showing how many branches separate any two distinct languages. But while this provides an objective measure of the distance between languages, does the resulting measure have any traction as a proxy for the distance between cultures? Montalvo and Reynal-Querol have recently investigated whether language distance proxies cultural distance by using it in an analysis of inter-
group violence within countries. 19 Does the language gap between two ethnic groups in the same country significantly affect proneness to violent conflict between them? They find that the greater the distance between languages, the greater the proneness to intergroup violence. Their analysis is global, but since intergroup violence in the high-income societies is very limited, the important observa-tions are from other societies. Hence, the result should not be misinterpreted as implying that the immigration of linguistically distant groups makes a high-income society significantly more prone to violence. Modern developed societies have built so many defenses against intergroup violence that it is not a significant issue: the specter of “rivers of blood” flowing from violence between immigrants and the indigenous, which was first raised by Enoch Powell and has haunted liberal intellectuals ever since, is deluded melo-drama regardless of the scale of migration.
I am concerned with trust within groups, not with violence
between them. But if, in those societies where intergroup violence is not unthinkable, language distance increases it, there is a reasonable presumption that language distance also proxies the more general difficulties associated with forging mutual regard. Mutual antipathy and mutual regard are the endpoints on a common spectrum. Measured by language steps, the cultural gap between immigrant groups and indigenous populations in Europe is indeed usually wider than between Hispanics and the host population of America. Hence,
although Putnam’s results are for America, Europeans would be
78 HOST SOCIETIES: WELCOME OR RESENTMENT?
cavalier to dismiss their pertinence to Europe simply on the grounds that Europe is different. Here are a couple of recent instances from Britain that may reflect just such a process of the undermining of social capital within the indigenous population that Putnam is analyzing.
Some Illustrative Anecdotes
I have headed this section “Some Illustrative Anecdotes,” and that heading is important. The purpose of the following stories is to help the reader see how the rather academic-sounding discussion of trust and cooperation might actually play out in a real context.
Since the social theory is about how immigration can weaken trust within the indigenous population, the illustrations necessarily illustrate just that. But whereas theories can only be read with what Daniel Kahneman terms “slow” thinking, stories trigger “fast” thinking reactions: in other words, intellectual effort is replaced by visceral emotions. For the writer this poses a problem: without illustrations the ideas remain too dry to have meaning; with illustrations they risk becoming explosive. To lessen that risk, let me be clear that the following stories are not analyses: the interpretations that I place on these stories might well not be correct. But that they could be correct should help you to grasp the more abstract proposition that migration can have social costs and that on a sufficient scale, the social costs of migration could become substantial.
One of the most remarkable achievements of British culture has been the convention of an unarmed police force. Within Britain this seems so natural that it has largely been taken for granted—there is no right to bear arms in Britain; on the contrary, it is a serious criminal offense. By international and historical standards this state of affairs is highly unusual—a triumph of the civilized society.
THE SOCIAL CONSEQUENCES 79
The convention is evidently fragile, depending as it does upon a tacit agreement between police and criminals that guns will not be used.
Given that the police are unarmed, any one individual criminal would gain an advantage from being armed, yet if criminals routinely carried guns, the police would also do so. This creates a coordination problem within the criminal community. Somehow, over the decades, British criminals managed to enforce a code of not carry-ing guns. In the 1960s one criminal spectacularly breached this code, shooting dead three policemen. What happened next was
remarkable: the criminal tried to go to ground within his London social network, but could not do so. Finding himself ostracized, he fled to remote moorland, where he was caught living in a tent.
Recall that game theory tells us that such willingness of other players to punish transgressors of cooperation is essential to preserve good outcomes. Now roll on to 2011: two policemen arrest a known criminal with previous convictions. In the car taking him to the police station the criminal pulls a gun; the police are also armed and shoot him dead. What then ensues is in stark contrast to the 1960s. The social network of the criminal rushes to the police station and mounts a protest, several hundred strong, against the police. The criminal, Mark Duggan, is posthumously turned into a community hero. Of course, the two instances of armed criminality are not identical: in the former the criminal fired his gun, in the second, while the criminal pulled his gun, he did not get a chance to fire it. Further, in the decades between the two incidents trust in the police had considerably eroded. But the opposing responses of the criminals’ social networks are nevertheless striking. That of the 1960s reinforced the convention that it is impermissible for criminals to resort to guns, whereas that of 2011 undermined it. A salient difference is that Duggan was Afro-Caribbean and that the crowd of protesters that assembled outside the police station was also
80 HOST SOCIETIES: WELCOME OR RESENTMENT?
Afro-Caribbean. The bonds between Afro-Caribbean people living in the locality were evidently stronger than any sense that in possessing a gun he had breached a taboo. Over a long period relations between the Afro-Caribbean community and the police have revealed a mutual lack of trust, and there is evidence of racism within the police force. Members of Duggan’s social network
responded to the news by presuming that the police had shot him unnecessarily, rather than, what is perhaps a more likely interpretation of events, that the police officer reacted to an immediate situation of extreme fear. As a result, far from ostracizing him, his network came out in solidarity, aiming to punish the police. This is precisely the role of the “supervillains” whose behavior is ruinous in cooperation games. Such responses clearly threaten to undermine the fragile convention that neither criminals nor the police should carry guns.
The fact that the police were indeed in this instance armed tells us that the convention had already considerably eroded. In part, this erosion was a reflection of a much more generalized acceptance of violence in Western cultures that, as Steven Pinker has shown,
began in the 1960s, reversing a centuries-long gradual reduction. 20
But it may also have been accentuated by a highly specific difference between the culture of Afro-Caribbean immigrants and that of the indigenous population. While there are variations within the Caribbean, Jamaican culture is among the most violent in the world.
For example, murder rates are fifty times higher than in Britain.
Guns are normal, so it is unsurprising that Jamaican immigrants brought their gun culture with them; indeed, the gun culture of the Afro-Caribbe
an community is now a specific concern of British crime policy. That culture is perhaps why Duggan carried a gun: his uncle had been a gun-toting gang leader in Manchester, and he did not recognize it as breaking a taboo. Manchester itself is struggling
THE SOCIAL CONSEQUENCES 81
to live down its description as “Gunchester.” In 2012 it was the scene of a tragedy in which, for the first time in Britain, two police-women were shot dead. The shooting has initiated a serious public discussion as to whether British police should be armed: the convention is revealed as fragile. The perpetrator in the Manchester shooting was indigenous. Evidently, over the years, the norms of indigenous criminals have changed. Quite possibly this would
have happened without any immigration. But it is also possible that the immigration of a substantial group of people whose social convention was to carry guns had destabilized a benevolent social equilibrium.
Recall that the key prediction of Putnam’s work is that the decline in cooperation induced by immigration extends to the internal behavior of the indigenous community. The key damaging effect is not that immigrants and the indigenous population do not trust each other; it is that indigenous people lose trust in each other and so resort to opportunistic behavior. What happened in the aftermath of the Duggan incident may illustrate this breakdown in the restraints on opportunism within the indigenous population.
The Duggan protests metastasized into looting that spread across the country, conducted by many thousands of teenagers from the indigenous population. The behavior was, as far as we can tell, utterly apolitical. Indeed, public buildings were ignored. The targets were shopping centers, where teenagers smashed windows and gleefully stole the standard accoutrements of teenage life. The behavior was also unrelated to ethnicity: essentially, indigenous teenagers looted from indigenous shops. Such behavior by indigenous teenagers was without precedent. In part, it was accounted for not by cultural change but by technological advance: teenagers, being particularly IT-savvy, were able to coordinate looting on their mobile phones, thereby achieving the safety of numbers. The police
82 HOST SOCIETIES: WELCOME OR RESENTMENT?
response also came in for criticism: having been accused of being too aggressive in the Duggan case, in the riots they were accused of being too passive. But police response to criminal behavior is less revealing than the behavior itself. The looting can reasonably be seen as reflecting a decline in social capital within the indigenous population.
Here is a further possible instance of social capital being undermined by “supervillains.” It comes from community responses to the deaths of British soldiers fighting in Afghanistan. Their bodies are flown to an airbase in Britain, and a tradition has developed whereby as the coffins are driven through the local town people line the streets to pay their respects. This is itself a reflection of a much broader socially important convention in which heroism in public service is honored. The British soldiers fighting in Afghanistan reflect the multiethnic composition of British society, and so one of the soldiers killed was a British Muslim. A member of his family was interviewed on television to speak of the soldier’s courage and their pride in his sense of duty. But the speaker was too fearful of reprisals from a small but violent minority of other British Muslims to reveal either his name or face: he was interviewed in silhouette.
This was a fear of “supervillains.” Of course, the fear may well have been misplaced, but one reason that “supervillains” are so destruc-tive of social capital is that there do not need to be many of them to alter behavior.
Anecdotes are not analysis: they merely illustrate what analysis is trying to show. Working merely from anecdotes we can stack up counterexamples where immigrants have clearly contributed to the social capital of the indigenous population. One major such example is the Notting Hill Carnival, which has become the largest annual street party in Europe. The carnival was created by the Afro-Caribbean community, drawing on its pre-immigration traditions,
THE SOCIAL CONSEQUENCES 83
and huge numbers of people from the indigenous population
now also take part in it. Street parties are paradigmatic of the social capital that Putnam sees as so valuable.
So, working from anecdote, we could stack up whatever appar-
ent support for whatever story we find appealing. For this reason, it is not a valid means of analysis: rather, it is the stuff of opinionated advocacy. The anti-immigration lobby will use one set of stories and the pro-immigration lobby a counter-set. The purpose of the above anecdotes in which immigration appears to have undermined social capital is decidedly not to strengthen an argument. Their role is purely to help the reader see what in practice both Putnam and the game-theory analysts of fragile cooperation are getting at.
Mutual Regard and Equity
So far I have focused on mutual regard as a source of trust, in turn supporting cooperation. But mutual regard is also important for an equal society. Without public transfers, the distribution of income is likely to become grossly unequal. Indeed, in recent decades technological pressures toward inequality have probably been com-
pounded by social pressures.
21 The growth of the information
economy has probably increased the returns to exceptional mental abilities. This new elite of the highly educated tend to cluster together not just at work but socially. They intermarry and their offspring have powerful educational advantages. As a result, social mobility is reduced: a trend that has been most marked in the United States and Britain, where countervailing government policies have been least active. It is not necessary to be on the political left to regard rapidly widening social inequality as undesirable. Wide differences in incomes can make a society less livable. Raghuram Rajan, a respected and sophisticated conservative economist, suggests that
84 HOST SOCIETIES: WELCOME OR RESENTMENT?
the political gridlock in America over fiscal policy may reflect the underlying divergence of interests between America’s rich and America’s poor: the population in the middle ground has shrunk.
So a technologically and socially driven process of widening
inequality calls for more active redistribution. The objective need not be the traditional rallying cry of the Left, a more equal society, but the more modest and conservative one of preventing rapidly increasing inequality. But, in fact, despite the growing need for redistributive policies, actual policies have shifted in the opposite direction. There has been not only a drift to lower taxation of incomes, but, more subtly, many goods and services that used to be supplied through the government are now supplied through the
market. Michael Sandel has brilliantly anatomized this process, which since the 1960s has shrunk the role of the state and thereby
contributed to rising inequality. 22 Lower taxation and the expanded role of the market have reflected and contributed to a weakened sense of a shared society.
For redistribution to be politically feasible, sufficient fortunate people must be willing to subsidize the less fortunate. So the regard of the fortunate for the less fortunate would need to be deepened.
We are back to the concept of empathy: high earners need to be able to see low earners as themselves minus good fortune. Empathy comes from a shared sense of identity. An important way of
building common identity is common membership in a network
of reciprocal obligations.
The immigration of culturally distant people who disproportionately occupy low-income slots in the economy weakens this mechanism. Low-income people become less like high-income people. In turn, unless this is offset, it reduces the willingness of high-income people to make transfers to low-income people. Many influences contributed to the policies of reduced taxation and increased
THE SOCIAL CONSEQUENCES 85
reliance on the market, not least that of the economics profession.
But the pronounced increase in cultural diversity brought about by immigration may have been one of th
em. For example, the recent phase of the open door in Britain has coincided with a collapse in willingness to fund redistribution. In 1991 a substantial majority of Britons—58 percent—agreed that government should spend more
on welfare benefits even if it led to higher taxes; by 2012 this had fallen to an inconsequential minority—28 percent. The argument that cultural diversity reduces the willingness to redistribute income has been formalized and investigated by two highly distinguished
Harvard professors, Alberto Alesina and Edward Glaeser. 23 They posed the question of why there has been so much greater willingness to accept redistribution in Europe than in the United States.
Their explanation was that the distinctive attitudes of the typical European country were grounded in its greater cultural homogene-ity. There is also some evidence that what erodes the willingness to redistribute is the rate at which diversity increases rather than simply its level. However, the importance of the level of diversity is
supported by a wide array of evidence. 24 As predicted by the theory, the greater the level of cultural diversity, the worse the provision of redistributive public goods.
As with diversity and cooperation, particular cases merely have the status of illustrative anecdotes: with that caveat, exhibit A is California. Due to the conjunction of geography and opportunity, California has the highest incidence of immigrants of all American states. All these immigrants have arrived in the past fifty years, because until the 1960s America had a closed door. Most of California’s immigrants cluster in the lower range of the income distribution.
So, according to the theory, California has precisely the pre conditions for a growing reluctance on the part of high-income groups to pay for redistribution. California is an immensely rich state: it can
86 HOST SOCIETIES: WELCOME OR RESENTMENT?
certainly afford redistribution. For example, it is home to Silicon Valley. But its most distinctive feature in recent decades has been the collapse of its public services. The schooling system in California has plummeted down the American league tables and is now
Paul Collier Page 8