Iraq: One Nation Under Allah

Why it's unrealistic to expect a separation of mosque and state

By Robert J. Barro

The insurgency in Iraq goes on, but attention has shifted toward the structure of an independent government. One key issue is whether the new regime will have a state religion, as envisioned by the U.S.-imposed transitional administrative law. Research by Rachel M. McCleary and me at Harvard University's Project on Religion, Political Economy & Society suggests that the Iraqi government will have an official religion. Although separation of religion and state has long been a Western ideal, it seems politically unrealistic for Iraq.

The consequences of state religion are complicated in practice. We find that the presence of state religion encourages participation in formal religious services (likely because of governmental subsidies to organized religion) and provides a smaller boost to religious beliefs. There is a weak negative impact on economic growth. And although some rich countries such as Britain and most of Scandinavia maintain religious liberties despite having official religions, state religions tend to coincide with curbs on religious freedoms.

In our research, we used international data to isolate the demographic, social, and economic factors that lead to the establishment of a state religion. Most important is whether a country has a diversity of religions or is concentrated in a single faith. For example, in 2000, the largest U.S. group was Protestants, with 44%, whereas in Morocco it was Sunni Muslims, with 98%. We estimate that this difference made the probability of a state religion in Morocco (which had one) higher by 69 percentage points than in the U.S. (which did not).

LESS CLEAR IS WHETHER the identity of a country's main religion influences the likelihood of state religion. In 2000, among the 40 countries where more than 50% of the people were Muslim, 28 (or 70%) had a state religion. In contrast, among the 91 countries with more than 50% in a single non-Muslim religion, 42 (or only 46%) had a state religion. We found that half of the Muslim/non-Muslim difference arose because the share of population that adhered to the most popular religion was higher in Muslim countries. The rest of the difference involves a greater tendency by government to curb religious freedoms in Muslim countries.

Another important factor is that communist countries rarely have a state religion (if communism does not count as a religion). However, this influence lacks staying power: Since 1990, 15 ex-communist countries, including six Muslim ones, set up state religions.

The size of a country matters; neither very small nor very large countries tend to have state religions. For small nations, the cost of maintaining an official religious administration is simply too great. For large countries, even a small proportion of the population in minority religions constitutes a large number of people and, therefore, creates resistance to a monopoly religion supported by the government.

For Iraq, a big question is whether the stronger force toward state religion is the 96% of the population that is Muslim or the 61% that is Shiite Muslim. We analyzed this issue using estimated breakdowns of Muslim populations into Sunni, Shiite, and other sects. Sunni is by far the largest worldwide, and only Iran is mainly Shiite (86% of the people). Only a few Muslim countries have substantial representation in more than one category: Azerbaijan, Bahrain, Iraq, Kuwait, Oman, and Yemen. Some other Muslim countries -- Afghanistan, Pakistan, Syria, Turkey, and the United Arab Emirates -- are 10% to 20% Shiite. But our results suggest that overall Muslim share, not size of the most popular sect, is what influences the creation of state religion.

No model predicts perfectly, but ours gets the right answer more than 80% of the time. Thus, it is instructive that the model's probability for a state religion in Iraq is 96%. True, our method also gives neighboring Turkey an 88% probability for state religion, even though it has been officially secular for decades. One can view this mistaken prediction two ways. One is that, despite forces that favor state religion, Turkey can be a model for Iraq on how to separate church and state. The other is that Turkey's secular status represents hard-to-duplicate political influence by its strong President, Mustafa Kemal Atatürk, in the 1920s and '30s. Conceivably, the factors that favor a state religion will eventually generate an Islamic state in Turkey. That reversal seems more likely than Iraq's becoming a secular state.

Robert J. Barro is a professor of economics at Harvard University and a senior fellow of the Hoover Institution (

    Before it's here, it's on the Bloomberg Terminal.