VoxEU Column Development Economic history Education Europe's nations and regions

Devotion and development: Religiosity, education, and economic progress in 19th-century France

Religion has had a complex relationship with technological progress throughout history, but there is scant empirical evidence on how conservative religious values may have affected the spread of new ideas and, by extension, economic development. This column examines the influence of the Catholic Church on technical education in France during the Second Industrial Revolution. It finds that areas with higher ‘religiosity’ had lower levels of industrial and economic development, suggesting that conservative religion can hamper economic development when it prevents primary schools from adopting technical education.

Religion has played a primary role in human societies for millennia and continues to do so for billions of people around the globe.  A rich literature – starting with the pioneering work of Max Weber (1905) – has pointed to different channels through which religion can affect economic development. Religion’s relationship with scientific-technological progress has been particularly complex throughout history, and clashes between religious doctrines and scientific development remain important in many countries today (Bénabou et al. 2015). 

Recent studies have analysed the hindering role that religion can play for the diffusion of new ideas and innovation, focusing on different settings and religious affiliations (Berman 2000, Mokyr 2011, Bénabou et al. 2015, Carvalho 2013, Carvalho et al. 2017, Iyigun et al. 2018). Most of these articles adopt historical and/or theoretical viewpoints, but there is scant empirical evidence on how conservative religious values can hamper the spread of new ideas and affect economic development. It is challenging to measure religion and to find a context in which to study its interaction with the adoption of ‘useful knowledge’. 

In a recent paper (Squicciarini 2019), I focus on 19th-century Catholicism during a crucial phase of modern economic growth: the Second Industrial Revolution (1870-1914) in France. In this period, Western economies started to adopt transformative, skill-intensive technologies. Providing technical education to the masses in primary school became an essential component of the industrialisation process (Galor and Moav 2006). 

At the same time, the Catholic Church was promoting a conservative, anti-scientific programme and hindered the introduction of the technical curriculum, while pushing for religious education. This tension was particularly strong in France, which experienced spectacular scientific and economic development, and where the relationship between the Church and science had been exacerbated by the 1789 French Revolution. While 98% of the French population was Catholic, there was large pre-existing variation in the intensity of Catholicism (which I refer to as ‘religiosity’). I exploit this variation to study the differential diffusion of technical education and industrial development in about 80 French departments (districts) and 230 French cantons (metropolitan areas).

To conduct the empirical analysis, I assemble a rich dataset from historical archives and secondary sources. My main measure of religiosity is the share of refractory clergy in 1791. This represents the share of French clergy that did not swear the oath of allegiance to the Civil Constitution promoted by the revolutionary government but instead confirmed their loyalty to the Catholic Church. Since a clergyman’s decision to accept or reject the oath was largely determined by the religious attitude of the local community, the share of refractory clergy reflected religiosity at the local level (Tackett 1986). To further validate this measure, I use six other indicators for Catholic intensity and provide evidence of a stable spatial distribution of religiosity over time (Figure 1).

Figure 1 Spatial distribution of religiosity in France

Notes: The figure shows the spatial distribution of religiosity in France. The left panel uses data on the share of refractory clergy in 1791 (Tackett 1986) and the right panel uses data on church attendance in 1945-1966 (Isambert and Terrenoire 1980). The maps are obtained using quartiles of the respective variables.

Empirical results 

Locations with higher religiosity had lower industrial and economic development during the Second Industrial Revolution, but not before. I first provide cross-sectional evidence at the department and at the canton levels, using a host of outcome variables (Figure 2). These findings are supported by a difference-in-differences analysis, showing that the more religious departments had significantly lower industrial employment in the post-1870 period – and suggesting that pre-existing variation in religiosity started to matter when skill-intensive technologies were introduced. 

Figure 2 Religiosity and industrial employment before and during the Second Industrial Revolution 

Notes: The figure plots the share of industrial employment against the share of refractory clergy, after controlling for a set of baseline controls (see Squicciarini 2019 for details). The left panel examines the share of industrial employment in 1866 (i.e. before the Second Industrial Revolution). The right panel examines the share of industrial employment in 1901 (i.e. when the Second Industrial Revolution had been on its way for some decades).

What explains the negative relationship between religiosity and industrialisation after 1870? In these decades, the contribution of human capital to industrialisation changed dramatically: contrary to the First Industrial Revolution – when the upper tail of the skill distribution was crucial for industrial development and worker skills mattered less (Mokyr 2005, Squicciarini and Voigtländer 2015) – the more sophisticated industrial machinery of the Second Industrial Revolution required a technically skilled workforce to be operated and maintained (Galor and Moav 2006). Consequently, the French state took an active role in primary education, promoting a more technical curriculum to form a skilled labour force. While educational policies were adopted at the national level, religiosity played a key role in their local implementation. 

I find that the more religious locations experienced a slower adoption of the technical curriculum and a push for Catholic education. Historical records suggest that this was driven by parents’ preference for religious education for their children, especially when, starting in the 1870s, the differences between the two education systems increased (see Figure 3) and the Catholic identity was being threatened by the introduction of the secular, technical curriculum (Grew and Harrigan 1991). 

Figure 3 Religiosity and Catholic education

Notes: The figure plots the per-period coefficient of the share of Catholic schools on religiosity. The baseline time-period is 1851. Data on the share of Catholic schools are not available for the year 1856. The bars represent 90% confidence intervals. 

I investigate the role of secular versus Catholic education for the industrialisation process. Using detailed panel data from 1871 to 1911, I show that the share of Catholic schools was negatively and significantly associated with employment in industry about 10 to 15 years later, while lagged industrialisation does not predict the type of schooling later on. 

This specific time pattern speaks against concerns of reverse causation and is consistent with an effect of education type on industrial employment. Thus, the type of primary education (Catholic versus secular) – whose choice likely depended on parents’ religious preferences – seems to be crucial for the diffusion of ‘useful knowledge’ for the formation of a skilled labour force, and hence for industrial and economic development.

Interpretation and discussion

When interpreting the above findings, I do not claim that the type of primary education is the only mechanism that explains the negative relationship between religiosity and industrial-economic progress during the Second Industrial Revolution. The anti-scientific programme of the Church could have been manifested in other aspects of people’s lives: for example, the opposition to vaccinations and birth control, as well as the proscription of the use of electricity in churches. I provide some evidence for this, showing that the more religious departments had a lower vaccination rate and higher fertility. However, when studying the role of these alternative effects, I find that they are uncorrelated with industrialisation. 

There is also a large set of potentially confounding factors. Among them, government investments could have been an additional important driver of industrialisation, especially if the government was differentially devoting resources to secular- versus Catholic-oriented departments. I show that state spending is unlikely to fully explain the adoption of skill-intensive technologies during the late-19th century or to confound my results. 

Thus, the empirical evidence, together with a rich historical record, makes it likely that education was the primary mechanism that drove the negative relationship between religiosity and economic outcomes during the Second Industrial Revolution. 

In addition, I shed light on the main differences between secular and religious educational curricula – with the former introducing ‘technical’ subjects and becoming increasingly modern and professional, and the latter remaining “the bastions of a Catholic subculture” (Grew and Harrigan 1991: 221). I discuss and explore other critical educational dimensions that could distinguish Catholic from secular education and affect industrial employment. For instance, the two school systems did not differ in terms of student attendance and schools’ financial resources, further highlighting the key role of educational content for human capital accumulation. 

While the subjects included in the secular school curriculum (stretching from basics of arithmetic, geometry, and the metric system to their practical applications to objects and tools) were clearly relevant for the accumulation of ‘economically useful’ knowledge, I do not argue that the hours spent studying these topics turned schoolchildren into skilled workers. Children usually entered a profession after finishing primary school and they still needed to learn their exact tasks on the factory floor. However, the ‘economically useful’ knowledge acquired could have prepared them to better learn the extra knowledge required in their profession. 

The differences between Catholic and secular schools are not restricted to the subjects studied, but comprise broader attitudes that were (or were not) beneficial to the accumulation of ‘economically useful’ human capital. Thus, I emphasise a broad conception of the Catholic school curriculum, but one that is clearly distinct from a secular curriculum. 


These findings have important implications for economic development today since many developing countries – where religion plays a primary role in the personal and public spheres – are also experiencing large-scale technological progress, similar to the process of development in Western Europe during the Second Industrial Revolution. Three main implications emerge. 

First, the relationship between religion and economic development becomes negative when religion clashes with and hinders the adoption of ‘economically useful’ knowledge. Second, the intensity of religion is key in that it determines the importance given to religious norms and the degree of resistance to new ideas and innovative activities if these clash with religious values. Third, the mechanism through which religion can affect economic development is by affecting the content of education, and thus the accumulation of human capital among the population.


Bénabou, R, D Ticchi and A Vindigni (2015), “Forbidden fruits: The political economy of science, religion and growth”, NBER Working Paper 21105.

Berman, E (2000), “Sect, subsidy, and sacrifice: an economist’s view of ultra-orthodox Jews”, Quarterly Journal of Economics 115(3): 905–953.

Carvalho, J-P (2013), “Veiling”, Quarterly Journal of Economics 128(1): 337–370.

Carvalho, J-P, M Koyama and M Sacks (2017), “Education, identity, and community: Lessons from Jewish emancipation”, Public Choice 171(1–2): 119–143.

Crabtree, S (2010), “Religiosity highest in world’s poorest nations”, Gallup. 

Galor, O, and O Moav (2006), “Das Human-Kapital: A theory of the demise of the class structure”, The Review of Economic Studies 73(1): 85–117.

Grew, R, and P Harrigan (1991), School, state, and society: The growth of elementary schooling in nineteenth-century France – A quantitative analysis, Michigan: University of Michigan Press. 

Iyigun, M, J Rubin and A Seror (2018), “A theory of conservative revivals”, IZA DP 11954.

Mokyr, J (2002), The gifts of Athena: Historical origins of the knowledge economy, Princeton: Princeton University Press.

Mokyr, J (2005), “The intellectual origin of modern economic growth”, Journal of Economic History 65(2): 285–351.

Mokyr, J (2011), “The economics of being Jewish”, Critical Review 23(1-2): 195–206.

Squicciarini, M P (2019), “Devotion and development: Religiosity, education, and economic progress in 19th-century France”, CEPR Discussion Paper 13877.

Tackett, T (1986), Religion, revolution, and regional culture in eighteenth-century France: The ecclesiastical oath of 1791, Princeton: Princeton University Press.

Weber, M (1905), The Protestant ethic and the spirit of capitalism, London: Routledge.


[1] The 2009 Gallup surveys show that the global median proportion of people who said religion is important in their daily lives was 84%. In 10 countries and areas, at least 98% of people said religion is important in their daily lives and this percentage was as high as 65% in the United States (Crabtree 2010).

[2] Following Mokyr (2002), I refer to ‘useful knowledge’ as knowledge that is ‘economically useful’, i.e. necessary for economic development.

17,954 Reads