Moral compass leads way to economic development

Whether people trust their neighbours may seem more a question of morality or religion than economics, but a new CEPR paper suggests that social values can be crucial to determining a nation's wealth and political success.

A growing body of research in recent years has studied the influence of history on levels of economic development today. But these studies often leave unanswered the question of precisely what mechanism is at work. CEPR Researcher Guido Tabellini believes that 'culture' - specifically, generalised norms about morality and trust - can provide the missing link in these political-economic explanations; and he sets out to prove it, using a battery of data about social beliefs, political institutions and even trade specialisations, across the world.

Political scientists and economists often try to derive a complete explanation from the way that institutions - such as parliamentary democracy, for example - interact with people's individual motivations. Corruption and conflict may endure despite elections, for example, because some individuals are able to exploit the process to their own advantage. Politics is an ongoing struggle over the distribution of resources. But Tabellini argues that explanations of this kind fail to account for the behaviour of thousands of atomistic individuals at grass-roots level, which can nevertheless affect country-wide political developments.

Political participation is much more intense in some countries than others, for example; and some electorates seem to demand higher ethical standards of their politicians, and punish corruption more harshly at the ballot box. Differences of this kind are difficult to explain on the basis of individual self-interest; but they can nevertheless influence historical and political outcomes.

Tabellini argues that slow-moving, persistent cultural values can help to fill the gap here, explaining, for example, why the abrupt arrival of new political institutions - after independence from colonial rule, say - does not always change things overnight.

In particular, he says there is a crucial difference between generalised and more limited morality. Generalised morality means the acceptance that abstract individuals are entitled to rights, and equal treatment. Limited morality is more like 'honour among thieves' - in a more hierarchical society, individuals may believe that members of their family, or clan, or close associates are entitled to respect; but at the same time it may be deemed acceptable to behave quite selfishly or opportunistically to anyone outside the group.

Tabellini argues that in a society where norms of generalised morality are widespread, free-riding is likely to be less common; it may be easier to get the rule of law off the ground; and voting is more likely to be motivated by concerns about social welfare, rather than individualistic calculations of personal gain.

In this way, generalised morality can help to generate political stability, better governance, and economic success.

In setting out to test this idea, Tabellini exploits the widely-used World Value Surveys, a series of studies carried out four times between 1981 and 2000, and covering a range of up to 70 countries. The surveys asked a large number of questions about beliefs and values, but he focuses on two that capture the notion of generalised morality in which he is interested.

Respondents were asked: 'Generally speaking, would you say that most people can be trusted, or that you can't be too careful in dealing with people?', and Tabellini uses their answers to score whether they had the value of trust, or not. Second, they were asked to pick which of several values it was especially important for parents to pass on to their children, and one possible answer was 'tolerance and respect for other people.'

Taken together, Tabellini interprets the responses to these two questions as an indication of the presence or absence of generalised morality

He then sets out to demonstrate two 'stylised facts', which begin to build up a cultural explanation for economic and political outcomes. First, governance in a particular country tends to be good or bad across a whole range of dimensions - as Tabellini puts it, 'policy distortions and government inefficiencies are often clustered together, as if they had a common cause'.

The second stylised fact is that moral values tend to be very persistent through generations. He shows that levels of 'trust' among third-generation immigrants to the US remain strongly correlated with the average level in their country of origin.

There is also a strong correlation between an immigrant's propensity to hold the value of trust, and the political institutions in their country-of-origin more than a century ago. Immigrants from countries that were democratic in the distant past (Tabellini uses measures from 1850, 1875 and 1900), tend to have higher levels of trust. Thus these generalised moral values shift very slowly over time, and national political history seems to have a strong hold over them.

These stylised facts provide some evidence of a correlation between moral values and institutions; but Tabellini wants to demonstrate causation - and for that he uses a series of instrumental variable regressions.

To avoid the risk that he is picking up reverse causation - in other words, that good institutions cause generalised moral values, instead of vice versa - the author needs to find an alternative variable to stand in for the value-measures themselves: something which is closely correlated with it, but does not itself have causal power over political institutions.

For this purpose, he uses two of the 'deep' rules of a language's grammar, which linguists have connected with how hierarchical a society is. One is the use of pronouns: in Italian, for example, the speaker can choose whether or not to use the pronoun in a sentence; in English, they cannot drop it. Linguists have suggested this is characteristic of cultures that give a greater focus to the individual, relative to his social surroundings - and thus to individual rights.

The second grammatical rule is the division between an informal, tu form of address, and a more formal vous form. Many languages once had this distinction, separating closer acquaintances from society more generally, but many have dropped it over the years. Some, however, including French and German, still keep it, perhaps suggesting the persistence of more hierarchical beliefs.

These putative connections between grammar and values may sound far-fetched, but Tabellini uses data from multilingual countries to show that there is a strong correlation between the more hierarchical grammar-rules, and the prevalence of the value of 'trust' - even where political institutions are exactly the same.

With these linguistic markers as variables, Tabellini then carries out a series of different regressions to illustrate the connections between cultural values and contemporary outcomes. He finds that the linguistic rules - and thus cultural values - explain the 'quality of government' across countries, controlling for education, per capita income and a range of other factors.

Similarly, using data from 69 different regions within European countries, he shows that the cultural values of trust and respect are strongly correlated with per capita output. He also follows other economists who have linked trade with the facility of making binding contracts, and finds that countries with generalised moral norms are more able to succeed in trade specialisations which require contracts - in effect, they have a comparative advantage against countries where trust is harder to find.

Tabellini also examines the way that shared values influence voting outcomes. The political scientist Robert Putnam conducted a well-known study comparing levels of what he called 'social capital' between more affluent Northern Italy, and the poorer South, where corruption is still much more of a problem. Tabellini explores this idea, relying on the fact that if a public prosecutor in Italy wants to investigate an elected representative, they must make a formal request to parliament.

Since these requests are a matter of public knowledge, and until the early 1990s Italian voters could indicate their preference for particular candidates in a party list voting system, he can track how severely politicians are penalised once it is known that prosecutors want to investigate them.

In the regions of Italy where 'trust' and 'respect' are measured to be weakest, incumbents with a request for investigation against them are barely penalised at all - in fact, public support for them at the polls increases slightly. In areas where generalised moral values are stronger, however, the incumbent's vote decreases by 20-35%. Given that political institutions themselves are the same throughout Italy, Tabellini argues that this is evidence of exactly the sort of mechanism he is investigating - moral norms affecting political outcomes.

Both economists and political scientists often regard 'cultural' explanations as woolly and weak, when compared to hard-nosed assessments of individual agents' incentives and the concrete details of institutions. In this paper, however, Tabellini argues that in fact, reducing political and economic developments to incentives and institutions alone fails to explain puzzling and persistent trends. He tries to fill this explanatory gap with the notion of cultural values - specifically, a generalised acceptance that other individuals can be trusted, and should be respected. Norms like these are passed on from generation to generation, and very slow to change; but once embedded, he argues, they can have dramatic causal effects in their own right.

DP 6589 Culture and Institutions

Guido Tabellini