Prospero | Johnson: Multilingualism and social attitudes
Social capital in the 21st century
Multilingual countries have less trust between people and in institutions. That is no reason to force countries into a monolingual mold
By R.L.G. | BERLIN
THE ECONOMIST, as a liberal newspaper, is a believer in globalisation and migration. We have also often argued that multi-ethnic countries should make compromises to stay together, rather than break apart into ethno-nationalist states. We hold diversity to be a virtue, not a weakness or failing.
But is this Panglossian? Diversity is fine in the best of all possible worlds, but we do not seem to live in such a world. For example, economists have long known that highly multilingual countries tend to be poorer than those dominated by a single big language. This raises a question: If linguistically homogeneous states tend to be richer, do they get richer because they are monolingual, or does development tend to favour the biggest language at the expense of smaller competitors?
A new paper* by Bodo Steiner and Cong Wang, two economists at the University of Southern Denmark, reach the conclusion that less linguistically diverse countries tend to prosper (rather than the other way round). Messrs Steiner and Wang set out to find the relationship between linguistic fragmentation and social capital. Social capital itself can be hard to define, but a few scholars have nonetheless put a number on it. Three scholars led by Dan Lee of Sungkyunkwan University in South Korea created an index** of 72 countries across criteria like trust (a feeling of societal fairness, confidence institutions like government, political parties and the press); norms (corruption, the rule of law, the prevalence of tax evasion and benefit fraud); and networks (for example how likely people are to join religious groups, arts and sports clubs and the like). Countries with high levels of social capital tend to be richer.
They also tend to be more linguistically homogeneous. Messrs Wang and Steiner found that the number of language spoken in a country is significantly negatively correlated with social capital (see chart above; the country codes used can be found here.)
Some countries have many languages and relatively high social capital. America and Canada are both outliers here, as immigration destinations that are also host to many indigenous languages. But they are also dominated by a big native language: English, with, in Canada’s case, an important role for another big language, French.
So the authors also plot the probability that two citizens will speak the same first language against social capital. These measures are even more closely (negatively) correlated. Despite headline-grabbing levels of immigration, countries like Denmark and the Netherlands remain linguistically highly homogeneous; they also have the highest rates of social capital in the data set. Uganda and India fall neatly along the trendline at the other end.
The authors checked the robustness of their results in various ways. The results held only slightly less strongly when excluding African countries, which are the poorest in the set and also highly linguistically diverse. The effects also held when controlling for other factors that are correlated with linguistic diversity: for example, mountainous countries, and those closest to the equator, tend to be more linguistically diverse. But even controlling for those factors, diversity was bad for a country’s social capital.
No one factor explains all countries. Bangladesh is poor despite being fairly linguistically homogeneous. Belgium and Switzerland are rich despite the lack of a dominant linguistic nationhood. South Africa does better than its multilingualism would predict; Burkina Faso does worse.
But the overall trend remains clear. What practical conclusions can be drawn? In the case of immigration, linguistic assimilation is important: whether or not immigrants and their children keep their old languages, they must master the languages of their new countries. The host countries must create the kind of inclusive society that says that anyone putting the effort in will be welcome.
The lessons for naturally multiethnic states are not as clear. However much homogeneity might serve an Iceland or a Japan, other countries cannot be engineered to look like them in the real world. Neat partitions and velvet divorces are rare. Messy border wars are much more common. Ukraine’s decision (soon reversed) to strip Russian of its role in southern and eastern provinces gave Russia an excuse to fan the flames of a war to protect Russianness itself in Ukraine. This is the very thing that burns up social capital, otherwise known as trust between people, and between citizens and their state.
Far better is the creation of identities beyond tribal solidarity. The European Union was, of course, born of this kind of thinking. Within it are territories once bloodied by ethnic rivals, now dully prosperous border regions. Alsace and Lorraine are stably part of France; neighboring Saarland voted after the second world war to join Germany. French and German are widely spoken on both sides, and Saarland teaches French before English in many schools.
Or take Sondeborg, where the two authors are based, in southern Denmark. Prussia won the second of two wars over the Danish-German border in 1864. But after the first world war, a plebiscite split the province of Schleswig between the two countries. It is especially fitting that Prof Steiner should research multilingual places; hailing from the Saarland, he now commutes from German Schleswig-Holstein across the border to Sonderborg—which voted in 1920 to join Germany but went with the majority in the surrounding zone to Denmark. He says most of his department's faculty and staff are cheerfully bilingual, and commemorations of the 1864 war show no hint of enmity. The old Danish border barracks is now a refugee centre, hosting Denmark’s biggest concentration of Syrians. It is a needed reminder of the enduring peace that western Europe has improbably achieved—and which too much of the world still sorely lacks.
* Cong Wang and Bodo Steiner, "Can Ethno-Linguistic Diversity Explain Cross-Country Differences in Social Capital?: A Global Perspective", Economic Record, 2015. Also available as a working paper.
** Dan Lee, Kap-Young Jeong and Sean Chae, "Measuring Social Capital in East Asia and Other World Regions: Index of Social Capital for 72 Countries", Global Economic Review, 2011.
