Following public outcry, Decathlon – a major French sports company – recently cancelled plans to sell an athletic headscarf.

This is the most recent controversy over Islamic clothing, part of a long-running debate about the integration of Muslims in France. For some, veiling (covering the head or face) symbolises Muslims’ rejection of French values, most notably secularism, and a refusal to assimilate into French culture. Muslims’ preservation of cultural or religious practices is particularly viewed with fear against the backdrop of increasing homegrown radicalisation and terrorism.

To emphasise national values and integrate Muslims into them, France has enacted several policies to regulate Islamic dress, including the 2004 ban on headscarves in schools and the 2010 ban of burqas (full-face covering) in public. Today, about one third of European countries restrict face and head covering in some settings, whether in state institutions (such as courts) or the public sphere more broadly.

Despite the increasing ubiquity of such policies, there is little systematic evidence of their impact. What is the effect of religious bans on the economic and social integration of Muslim minorities?

The answer is not obvious. Many countries have used ‘assimilationist’ strategies to cultivate a national identity. A textbook success case is French nationalisation of rural regions through public education and military conscription. But some research suggests that cultural prohibitions can intensify a minority’s sense of identity.

We bring empirical evidence to bear on this debate in a recent study evaluating a landmark cultural ban. In 2004, the French National Assembly banned conspicuous religious symbols in public primary and secondary schools. Though the law did not single out any particular religious symbol, it was aimed at (and mostly affected) veiled Muslim girls. The law also affected non-veiled Muslim girls, whose behaviour and religious identity came under scrutiny to ensure their compliance with the law.

To study the effects of the headscarf law, we focus on two groups of French women: those born before 1986 and who thus completed secondary school before the law was enacted in 2004; and those born in 1986 and later and who were thus in school during the ban’s implementation.

For pre- and post-ban cohorts, we compare Muslim women’s educational and economic outcomes with those of their non-Muslim peers (using the France’s labour market survey). Then, we assess how the difference in outcomes between Muslim and non-Muslim women changes for cohorts in school during the law’s enactment compared with cohorts in school before the ban.

On average, Muslims in France have been worse off than their non-Muslim counterparts. So we observe a gap in educational attainment (and other outcomes) between Muslim and non-Muslim women for all cohorts in our data.

But if the ban had no effect, the difference in outcomes between Muslim and non-Muslim women would remain unchanged between cohorts born before 1986 – who were not exposed to the 2004 ban – and cohorts born from 1986 onwards – who were exposed to the ban.

Our analysis shows that the law did widen the gap between Muslim and non-Muslim women. Specifically, the gap in secondary school attainment between Muslim and non-Muslim girls more than doubled after the ban.

This was partially due to Muslim women leaving the school system. Their rate of dropping out of secondary school increased by 6 percentage points more than that of their non-Muslim counterparts between the 2003/04 school year (before the ban) and the 2004/05 school year (after the ban).

Affected cohorts of Muslim women also took longer to complete secondary education, further depressing their attainment.

This negative educational shock hindered economic and social integration in the long term. After the ban, the employment gap between Muslim and non-Muslim women widened by a third, while the gap in labour force participation widened by a half. Muslim women were also less independent after the ban; on average, they have more children and are more likely to live with their parents.

What explains these negative outcomes?

First, the ban increased Muslim women’s perceptions of discrimination. The law singled out Muslim schoolgirls who chose to veil and subjected them to differential treatment because of their mode of dress. Public debate accompanying the passage of the law moreover reinforced Muslim girls’ difference.

Both the direct changes in schools and the broader anti-Muslim sentiment made Muslim girls feel targeted. In a social attitudes survey (known as Trajectories and Origins), Muslim women in cohorts affected by the 2004 ban are significantly more likely to report experiencing racism in school and to report lower trust in the French school.

As qualitative interviews reveal, this perceived discrimination placed Muslim girls under considerable psychological stress and disrupted their ability or willingness to perform at school – thereby impairing their educational and long-term economic outcomes.

The second explanation for the negative outcomes from the ban is that it cast religion and national identity as incompatible. The law defined the Muslim headscarf as what Joan Wallach Scott calls a ‘violation of French secularism, and by implication, a sign of the inherent non-Frenchness of anyone who practiced Islam’.

Until that point, French Muslim girls could readily identify as members of both their religious community (by wearing the headscarf) and their country of birth (France). After the ban, they received the signal that their two identities were incompatible and that one could not be French without embracing the principle of secularity as enshrined in the law.

Some Muslim women were thus alienated from broader French society and chose to retreat into their religious communities. Indeed, in the Trajectory and Origins survey, Muslim women affected by the ban are more likely to identify with their father’s country of origin than with France.

Consistent with the idea that the ban was accompanied by general scrutiny that reinforced Muslims’ difference, we find that Muslim men born in 1986/87 – who were in school during the ban’s implementation – experienced a small dip in educational attainment relative to prior cohorts.

But the gap in Muslim and non-Muslim men’s education rebounds to the earlier trend for subsequent cohorts born after 1987. This suggests that once the public debate died down, the ban continued to discriminate against Muslim women by subjecting them to differential treatment in schools, but it did not directly affect Muslim men.

With an estimated 26 million Muslims across Europe and a recent influx of Syrian refugees, the challenge of integration is at the forefront of public policy debate in Europe. ‘Multiculturalist’ policies, which embrace and accept cultural differences, have come under fire, most famously when Germany’s Angela Merkel proclaimed that ‘the multicultural concept is a failure, an absolute failure’.

Despite this proclamation, there is evidence that multiculturalism has modest positive effects for first generation immigrants, though no discernible effects for the second generation. What’s more, individual policies, whether multiculturalist or assimilationist, have not been sufficiently examined.

Our study contributes empirical evidence to this debate on integration. Our analysis of the effect of the 2004 headscarf ban provides causal evidence that policies with an assimilationist character can hinder integration. While they may succeed in their narrower goals (women took off the headscarf in the school, after all), we should consider also their broader consequences, which, in cases like this one, seem to be negative.

This column summarises ‘Political Secularism and Muslim Integration in the West: Assessing the Effects of the French Headscarf Ban’ by Aala Abdelgadir and Vasiliki Fouka.