The Salem witch trials of the 1690s have an iconic place in American lore. But before the Salem witch hunt, there was the “Great Hunt”: a larger, more prolonged European phenomenon between 1560 and 1630 that led to 80,000 accusations and 40,000 deaths.

Why’d it happen? Well, as with the Salem witch trials, there are a lot of theories. In the past, scholars have suggested that bad weather, decreased income, and weak government could have contributed to the witch trial period in Europe. But according to a new theory, these trials were a way for Catholic and Protestant churches to compete with each other for followers.

In a forthcoming Economic Journalarticle, economists Peter Leeson and Jacob Russ lay out their argument that the two churches advertised their finesse at persecuting witches as proof that they were the best church to join if you wanted protection from Satan. Witches, after all, were doing the bidding of Satan; so getting rid of them was a way to protect people from him.

“Similar to how contemporary Republican and Democrat candidates focus campaign activity in political battlegrounds … historical Catholic and Protestant officials focused witch-trial activity in confessional battlegrounds during the Reformation and Counter-Reformation to attract the loyalty of undecided Christians,” Leeson and Russ write. These “battlegrounds” were places where Protestantism had made inroads, giving Christians a choice about which church they wanted to belong to.

To bolster their point, the authors point out that from about 900 to 1400, the church didn’t want to acknowledge the existence of witches; and consequently, it didn’t try people for witchcraft. In 1258, Pope Alexander IV even prohibited the prosecution of witchcraft. Yet a few centuries later, the church reversed its decision. According to the economists, it was because of the Protestant Reformation.

Beginning in 1517, the Reformation split the church into two factions: Catholic and Protestant. Suddenly, these two churches had to compete with each other for followers, and they did so by using the attention-grabbing witch trials as perverse advertisements for their brand.

Leeson and Russ argue that this helps explain why areas where Protestantism spread saw more witch trials than solidly Catholic regions. Germany, where Protestantism began, accounted for 40 percent of these persecutions. Switzerland, France, England, and the Netherlands—all countries where Protestantism spread—accounted for 35 percent. But only six percent of persecutions took place collectively in Spain, Italy, Portugal and Ireland, all regions that were more solidly Catholic.

The economists argue that witch hunts declined in the late 17th century thanks to the Peace of Westphalia. That 1648 treaty ended two religious wars, including the Thirty Years War, and established a new balance of power in Europe. It also gave Protestantism and Catholicism a religious monopoly on certain regions, eliminating the need to compete for followers by persecuting witches.

Still, some witch trials did continue between 1650 and 1700. Leeson and Russ suggest this may have been because people had become accustomed to witch trials, and sincerely believed them to be a way of protecting their communities from Satan.

Using witch trials to attract followers is only possible when the belief in witches is widespread. In the same vein, people “will only continue to demand witch trials if that belief continues,” Leeson and Russ write. The scientific revolution “may have eventually eroded popular belief in witchcraft, eroding popular demand for witchcraft prosecutions along with it.”