When an Ideology Governs, It Grades Itself

For nearly fifty years, the Middle East has looked like the part of the world least interested in liberalism. Iran's 1979 revolution replaced a secular order with absolutist clerical rule. The region became known for theocratic insurgencies, repression of women, and the suppression of free speech. The easy conclusion: these societies simply want something other than individual liberty — a different "regime of truth," as the philosopher Michel Foucault approvingly said of Iran in 1978.

That conclusion confuses the behaviour of regimes with the preferences of the people living under them. Beneath the authoritarian surface, the data point the other way.

The situation: values moving against the government

Between 2000 and 2020, sociologist Mansoor Moaddel ran repeated face-to-face surveys of representative samples in five pivotal countries: Egypt, Iran, Iraq, Saudi Arabia, and Turkey. The surveys measured attitudes across three dimensions:

• Expressive individualism — marrying for love rather than by parental approval; prizing independence over obedience in children; a woman's right to dress as she chooses.
• Gender equality — attitudes toward polygamy, wifely obedience, and male priority in jobs, politics, and education.
• Secular politics — separating religion from the state; views on Islamic government and sharia.

Responses were combined into a "liberalism index" from 1 to 4, higher being more liberal. Across all five countries the scores rose — even as rulers worked to crush demands for democratic change.

The numbers are not subtle. In Egypt, support for separating religion from politics climbed from 44% in 2007 to 81% in 2020; support for Western-style government rose from 30% to 82% over the same span. In Saudi Arabia, those saying it was "very important" to implement sharia fell from 73% in 2003 to 35% in 2011. In Iraq, strong support for Islamic government dropped from 30% in 2004 to 15% by 2011. Belief that university matters more for boys than girls collapsed almost everywhere.

The model: preference falsification, and ideology graded by reality

Two established ideas explain this better than "these peoples wanted theocracy."

The first is preference falsification, named by economist Timur Kuran in Private Truths, Public Lies (1995). Under coercion, people publicly profess support for the ruling order while privately disbelieving it. Because everyone sees only the public professions, the regime looks more popular than it is — and the hidden gap can be enormous. Kuran used it to explain why almost no one, including the CIA, predicted how fast Eastern European communism would collapse in 1989: private disaffection had been building invisibly for years, then cascaded the moment expressing it became safe.

The second is simpler and older in the sociology of belief: values are updated by evidence, and the most powerful evidence is what happens when an ideology actually governs. Abstract support for an idea and support for that idea in power are different things, because governing produces results people can grade.

Put together, they invert the Foucault reading. Values are not a fixed cultural essence. They are revised by lived experience — and in a repressive state, much of that revision stays private until it surfaces, which is exactly what a slow climb in survey scores under hostile regimes would look like.

Ask why Egyptians changed, and the survey's follow-up interviews answer plainly. People who shifted toward secular politics overwhelmingly cited disillusionment with the Muslim Brotherhood after its candidate won the 2012 presidency and governed. Those who warmed to Western-style government pointed to Western prosperity and wanted institutions that might deliver it.

Both are learning responses. One is disappointment with the ideology in power: you backed the idea in the abstract, watched it run the country, and changed your mind. The other is comparison: you look at outcomes elsewhere and want what works. Neither requires anyone to abandon their faith. It requires only that they stop wanting the state to enforce it.

That is the counterintuitive turn. The experience of religious domination may feed the demand for a responsive, secular order rather than foreclose it. Repression does not prove consent; often it manufactures dissent, and preference falsification keeps that dissent off the books until it breaks.

The other view

Several caveats deserve a fair hearing.

Survey honesty under authoritarian rule. People in repressive states may tell interviewers what feels safe. This is preference falsification pointing at the data itself. But it cuts in one direction: fear would more likely suppress liberal answers than inflate them — meaning the true shift is probably larger, not smaller.

Values are not institutions. Rising liberal sentiment has not produced liberal governments. The Arab Spring was crushed; Iran's security forces have met protest waves, including the recent unrest, with lethal force on a scale that is contested and hard to verify independently. A gap between what people want and how they are ruled can persist a very long time.

The honest unknown: whether these shifts eventually reshape institutions or stay bottled up. Both are consistent with the data so far — and Kuran's own point is that cascades are intrinsically unpredictable: the fact that a dam is under pressure tells you little about when it breaks.

But history argues against dismissing the trend. Values tend to move before institutions. Liberal-nationalist movements in early-twentieth-century Egypt, Syria, and Iran were preceded by decades of quiet idea-spread; the Islamic fundamentalism that triumphed in 1979 was itself incubated in the 1960s and 1970s. Ideas lead; regimes follow, or fall.

Where else this runs

The pattern — an ideology discredits itself by governing, while coercion hides the fallout until it cascades — is not Middle Eastern. It is general.

• Eastern Europe, 1989. Decades of lived communism, not Western propaganda, hollowed out belief; preference falsification then masked how total the collapse of belief already was.
• Prohibition-era America. Enforcing a moral code by law generated organised resistance and quietly turned opinion against the law itself, until repeal became thinkable fast.
• Ordinary life. The surest way to lose a true believer is to put their belief in charge and let reality grade it — and the surest way to be blindsided is a culture where dissent is costly, so everyone assumes the consensus is real until it isn't.

The transferable lesson: support measured before a belief holds power tells you little; support measured after is the real test — and in any environment where dissent is punished, the support you can see is not the support that exists.

What to watch

Two signals will show which story is winning.

First, generational data. If the young score consistently more liberal than their elders — and stay that way as they age — the shift is structural, not a mood.

Second, regime behaviour under stress. Governments that sense they are losing the argument tend to lean harder on coercion than persuasion. Escalating repression is, paradoxically, evidence that the underlying values are moving against them — and, per Kuran, that the visible consensus may be thinner than it looks.

Expectation: ~65% that published value surveys across these five countries, where access permits, continue to show rising liberalism-index scores through 2030. Translation of those values into institutional change remains far less certain — well under even odds within the same window, and, if it comes, more likely to arrive suddenly than gradually.

Distilled from Foreign Affairs