In his new book, “Nation Building: Why Some Countries Come Together While Others Fall Apart,” Andreas Wimmer, a professor at Columbia University, asks why some countries collapse along ethnic lines, while equally diverse nations remain stable. Looking at countries that include Botswana, Belgium, China, Russia, Somalia and Switzerland, Wimmer finds that the most stable nations are the most inclusive.

In a recent interview, edited for length and clarity, Wimmer explained why citizens develop stronger attachments to their nation in some countries than in others, why nation-building is so slow, and what policymakers can do to accelerate it.

Among Western liberal elites, nationalism has an increasingly bad reputation. But you argue that it can have a positive side. Could you explain a bit?

Nationalism comes in different forms and ideological orientations. It comes from the left, it comes from the right, it has historically been allied with very different kinds of social movements. In the 1920s, 30s, and 40s, it was allied with right-wing, even fascist tendencies. But in the developing world, it’s also been allied with left-wing communist movements and anti-imperial, anti-colonial movements. Nationalism doesn’t have a political color on its own.

It’s important to recover some of the positive consequences of national identification. National identities can encourage collective solidarity and lead people to work for a shared common good. Governments supported by a strong national identification can redistribute resources and build welfare states more easily. Citizens in these countries are less resistant to paying taxes.

You argue that national identification depends on what you call ethnopolitical inclusion. How does that work exactly?

The basic idea is that if a citizen sees herself represented in the highest levels of government — let’s say you’re an African-American and you see that there is strong, meaningful representation of African American politicians in the cabinet of the United States — then you’re more likely to identify with the nation as whole.

Political inclusion promotes national identification. I show this empirically using hundreds of surveys from all over the world. Strong national identification is the consequence, not the cause, of inclusionary forms of governing.

What leads to this kind of ethnopolitical inclusion?

Everything depends on whether political alliance networks stretch across the whole country and include ethnic majorities and minorities alike. Three factors promote such wide-reaching, inclusionary political alliances.

First is political economy. Can the government provide public goods — things like infrastructure, protection from arbitrary violence, the rule of law — equally across the country? If so, citizens will support representatives of the national government rather than leaders of their own ethnic community or religious figures or local warlords.

Second is whether or not, in crucial moments of nation-building or early in the process, a country already has a well-established set of civil society or voluntary organizations spread across the country. Leaders can then rely on these voluntary organizations to build political support. Since these voluntary organizations have the tendency to spread horizontally across a country, they can link up easily with each other. The chances that these networks will be multiethnic are much higher compared to, for example, networks built on patron-client relationships, which are more likely to be monoethnic.

Third is a shared medium of communication, so that people can easily talk or write to each other. This makes it easier for people to understand each other, to renegotiate alliances, to guess other people’s intentions, and so on. I discuss and rule out a range of alternative factors, including democracy and economic growth.

You compare pairs of countries in which nation-building has succeeded in one case and failed in the other — Botswana vs. Somalia, for example, or Switzerland vs. Belgium — to show how each factor works. Could you talk about a particularly good such case study?

They’re all good! But the case study that illustrates the importance of having a shared medium of communication is China vs. Russia. In China, even among the Han majority there’s a very high degree of linguistic heterogeneity. But there is also a common script that is equally distant from all the spoken languages. This script has historically allowed people who could not understand each other when they talked to each other to write to each other, to understand each other’s pamphlets, and so on. Since the Middle Ages, this has allowed alliance networks in China to be multi-linguistic. That enabled the recruitment of an elite through the meritocratic exam system that is also multi-linguistic.

To this day, the Chinese governing elite is multiethnic, with each of these major language groups more or less equally represented at the highest level. This explains why there has never been a linguistic nationalist secessionist movement among China’s Han majority throughout its history.

Russia had even more linguistic diversity than in China, but its languages were also written in different scripts, including Cyrillic, Latin, and Arabic. This linguistic and scriptural diversity impeded alliances from forming across linguistic boundaries. Political alliances remain confined to speakers of the same language and no pan-Russian nationalism ever emerged. Instead, a series of nationalisms emerged for each separate linguistic community, which helped lead to the collapse of the Romanov Empire and, later, the Soviet Union, along ethnic fault lines.

Is there anything policymakers can do to promote nation building?

First, support civil society organizations that can connect individuals across ethnic and racial divides.

Second, help states become better at providing public goods, like the World Bank’s emphasis on good governance and the rule of law. But, as I show with data from Afghanistan, public goods need to be provided by the state, not the private sector or foreign NGOs, to foster nation-building.

Third, strengthen public school systems that teach all citizens a uniform language, or that promote the command of multiple languages systematically across the country.

But because I emphasize long-term, slow-moving factors, I’m quite skeptical about the possibility of turning things around quickly. Afghanistan and Iraq show that if domestic conditions are not ripe for cross-ethnic political alliances, then even if outsiders invest considerable resources, they cannot expect quick results.

Sam Winter-Levy is a PhD student in politics at Princeton University.

Nikita Lalwani is a student at Yale Law School.