A tale of minorities: evidence on religious ethics and entrepreneurship

Abstract

Does Protestantism favour entrepreneurship more than Catholicism does? We provide a novel way to answer this question by comparing Protestant and Catholic minorities using Swiss census data. Exploiting the strong adhesion of religious minorities to their denomination’ ethical principles and the historical determination of the geographical distribution of denominations across Swiss cantons, we find that Protestantism is associated with a significantly higher propensity for entrepreneurship. The estimated difference ranges between 1.5 and 3.2 % points, it is larger the smaller the size of the religious minority, it is mainly driven by prime age male entrepreneurs and it stands up to a number of robustness checks. No effects are found when comparing religious majorities, suggesting that the implications of religious ethical norms on economic outcomes emerge only when such norms are fully internalized.

Appendices

Appendix 1: The reform in Switzerland

At the beginning of the sixteenth century, Switzerland was composed of thirteen cantons—Uri, Schwytz, Unterwalden, and Lucerne (the four so-called Forest Cantons), Zug, Solothurn, Zürich, Glarus, Bern, Fribourg, Basel, Schaffhausen, and Appenzell—and a number of allied cities and dependent territories. Dependent territories were jointly administered by several cantons.

The Reformation in Switzerland is linked to the name of Ulrich Zwingli, a pastor in Zurich with an humanistic approach that was largely influenced by Erasmus’ ideas, and to a number of other leaders (e.g., Bucer, Bullinger, Calvin, Farel, Oecolampadius). Zwingli started preaching the need for reform of the Catholic Church at nearly the same time (1518) as Luther nailed his ninety-five theses to the doors of the Wittenberg Cathedral (1517). Zwingli’s ethical and moral criticisms of the Church where linked to the condemnation of the Swiss practice of serving as mercenaries for the European powers (Schaff 1882).

Although Zwingli’s and Luther’s theologies developed independently, they have many similarities, including the primacy of the scriptures over tradition, justification by faith alone and not good works, and universal priesthood without requiring the mediation of the clergy.

However, in several respects, Zwingli’s theology is the more radical. While Luther accepted Catholic tradition if it does not contrast with the scriptures, Zwingli wanted to quit any practice that was not explicitly grounded in the New Testament.³²

While in Luther’s view the new religion should be brought to the masses by the state, for Zwingli conversion should be achieved by means of persuasion and discussion, in line with his humanistic approach, as only reason should induce people to convert.³³

The Reformation spread from Zürich to many cantons, including St. Gallen (a former Catholic Prince-Abbacy), Basel, Bern, and Genève. The Forest Cantons and Zug opposed it fiercely, creating the league of the Five cantons, and called for the emperor’s support. War broke out in 1529 (the first War of Kappel), when Zürich and Bern attacked. A peace treaty favourable to Zürich and its allies was signed even before the war was fought, and the Reformation was declared free to spread in the dependent territories. In 1531 a second War of Kappel broke out, in which the five cantons attacked and eventually defeated the Reformed army and Zwingli, serving as a military chaplain, died in battle. Anticipating the Peace of Augsburg of 1555, the peace treaty established a principle similar to the “cuius regio eius religio”, where each canton was free to set its own religion. The primacy of Catholicism was favoured in the dependent territories and the Prince-Abbacy of St. Gallen was restored (Meyer 1976).

After Zwingli’s death, Calvin systematised and formalised the Swiss Reformed theology. Genève, Zürich, Bern, and Basel defined and agreed upon the common traits of their faith by means of the mutual consensus (1549), attenuating the more extreme spikes of Zwinglian doctrine. Next, in 1566 Bullinger prepared the (second) Helvetic Confession, which was published by the Elector Palatinate and adopted not only in the entire Reformed Switzerland but also in Scotland, France, Hungary, and Poland.³⁴

Religious tension continued in Switzerland until the eighteenth century but without major conflicts. The Catholic cantons promoted the re-catholicisation of the dependent territories that had previously turned to the Reformation. Glarus had two religions and two legislative assemblies, one representing each religious community, and Appenzell was split into two cantons according to religious lines. The Duchy of Savoy continued the war to reconquer Genève, but the city of Genève remained a Reformed enclave.

Catholicism was restored also in the Valais and in Solothurn. The Prince-Bishop of Basel regained some territories southwest of Basel, which turned to the Reformation in 1529. The Capuchins and the Jesuits, expressive of the Catholic Counter-Reformation, established schools and convents and largely favoured the process of re-catholicisation (Schaff 1882).

Hence following the application of the cuius regio eius religio principle, which aimed to establish a religious uniformity at least within each of the historically thirteen cantons, the Counter-Reformation was effective in reintroducing Catholicism also in parts of the reformed cantons. Furthermore, the initial effort of conversion to the Reformation and the subsequent re-catholicization determined the creation of a more mixed religious distribution in the dependent territories. Both phenomena favoured the formation of religious enclaves and minorities in most parts of the country. According to (Schaff 1882), Glarus, St. Gallen, Appenzell, Thurgau, and Aargau were nearly equally divided between the two denominations. Of the twenty-three subject towns and districts, only two became wholly Protestant, sixteen remained Catholic, and five were divided. In the Graubünden nearly two-thirds of the population adopted the Reformation.

Today the Reformed churches in Switzerland remain separate, cantonal units. The German churches are more in the Zwinglian tradition, the French more in the Calvinist tradition, but the differences are minor. All cantonal churches are federated in the Federation of Swiss Protestant Churches, which represents the common interests of its member churches on the national and international levels. Theologically the fundamental principles of the federation coincide with the cornerstones of Protestantism that we have already discussed: the primacy of the scriptures, salvation by faith alone, and universal priesthood.

The Federation of Swiss Protestant Churches reports five key tenets on its website:³⁵

• 1. solus Christus (By Christ alone): The love of Almighty God is evident in Jesus Christ, in his birth in poverty, his life, his preaching and healing, his crucifixion and resurrection. Jesus Christ opens the door to God, His and our Father in Heaven. He calls us to follow in His footsteps.

• 2. sola gratia (By grace alone) and sola fide (By faith alone): God’s salvation does not have to be earned; it is freely given to mankind (sola gratia). The only demand on us is that we accept it and receive the gift in faith (sola fide).

• 3. sola scriptura (By scripture alone): In questions of a Christian life and faith, the Reformed Churches are not guided by doctrines. They also have no authority to decree any binding directives. Their only guideline and source of knowledge are the scriptures of the Old and New Testaments.

• 4. soli Deo gloria (Glory to God alone): In praising God, we express the fact that we have Him to thank for everything. His mighty deeds and works are behind everything on earth. And so by praising God, we can move beyond ourselves and experience ourselves as part of a reality that transcends us.

• 5. The priesthood of all believers: everyone is received into the priesthood through baptism. The Reformed Churches do not differentiate between clergy and laymen but between general and official practice of work as a priest. Clergy are officially legitimised for this, meaning they are officially authorised to perform official duties for everything and everyone. Their office holds the specific task with which all who have been baptised are familiar: making the Gospel seen and heard throughout the world by word and deed.

Appendix 2: Placebo test

An important concern regarding the validity of our results is that the model may not pick up the effect of religious ethics but the effect of possibly differential characteristics of the territories where the religious groups reside, which could be more or less favourable to entrepreneurship. To address this concern we performed a placebo test which randomly reshuffles individual religious affiliation within each district, preserving the shares of each religious denomination, without thus altering the geographic distribution of majorities and minorities. As a result of reshuffling, any association between the reshuffled variable (religious denomination) and the rest of the dataset is random. On reshuffled data, interactions $M_{idt}\times P_{idt}$ and $m_{idt}\times P_{idt}$ were generated anew and model (1) was estimated to check whether a significant effect of Protestantism resulted. The presence of an effect would be an indication that the supposed influence of religious ethics on entrepreneurship is in fact due to confounders, and not to the respondents’ religion denomination. We repeated this operation 500 times. The effect of Protestantism was significantly different from zero only in 2 % of the repetitions, thus dismissing the concern that such an effect is due to non-religious factors.

Appendix 3: Additional tables

See Tables 11, 12, and 13.

Table 11 Religion and entrepreneurship—baseline model

Table 12 Alternative thresholds

Table 13 Proportions of minorities and majorities by geographical aggregation

Appendix 4: Additional figures

See Figs. 7, 8, 9 and 10.

Fig. 7

Geographical distribution of Protestant minorities defined at the district level at alternative minority thresholds

Fig. 8

Geographical distribution of Protestant minorities defined at the district level at alternative minority thresholds

Fig. 9

Minority Protestant municipalities confining with majority Protestant areas, under alternative criteria

Fig. 10

Minority Catholic municipalities confining with majority Catholic areas, under alternative criteria