Even though some Fascist politicians were self‐identified Catholics, colonizing Poland took precedence over any Christian solidarity, so Polish churches had to vanish:

The nature of repression differed between the annexed territories, intended as a Lebensraum for Germans, and the General Government, which served as a reservoir of Polish laborers. In the former, all traces of Polish identity would have to disappear. This was accomplished through the arrests and executions of Catholic priests, the closure of churches, and restrictions on Catholic practices. Moreover, schools that served Catholic and Jewish populations were closed.

The Catholic Church was the main target because it was considered a bastion of Polish identity. As early as July 1939, the Army High Command declared the Catholic clergy as “primarily responsible for nationalistic rabble‐rousing” (Huener 2021, 53). The image of the “agitator‐priest” (Hetzkaplan) as an enemy of Germandom drew on a legacy of Prussian animosity toward Catholicism during the partitions of Poland (Huener 2021).

The Warthegau, which had the largest Polish population within the annexed zones, experienced the brunt of these policies (Epstein 2010, 2). Hitler wished the Gau “to become flourishing German land in ten years” (Epstein 2010, 5), and he found a willing executioner in Greiser, an “anti‐Polish Nazi zealot” from the formerly Prussian province of Posen.

Greiser succeeded in decimating the Polish clergy in the Warthegau: of the 2,100 secular and religious clergy, 133 (6%) were killed in the Gau territory, 1,523 (73%) were arrested, and 1,092 (52%) were sent to concentration camps, where two thirds died (Huener 2021, 208).

As a result, entire districts had no priests to serve parishioners during the war (Huener 2021, 167). Charitable and educational institutions previously maintained by the Catholic Church were abolished. Some 97% of all church buildings and shrines were closed, desecrated, or destroyed (Huener 2021, 2). [Fascist] authorities also restricted the times of worship, prohibited public displays of faith, and tried to prevent individuals from travelling outside of their parish to attend services elsewhere.

Greiser closed Polish schools, although this policy was applied less consistently than repression against the Church. Instead of going to school, Poles aged 14 and above were forced to work. Poles were strictly segregated from Germans and faced restricted hours for using public baths and entering shops and markets. Public spaces and cafes were designated “for Germans only.”³

Greiser also began deportations into the General Government in order to create space for German settlers. However, these slowed as [the Fascists] realized that they needed Polish laborers and farmers. From mid‐1940 onward, thousands of Poles were put to work in the Warthegau or in the Altreich.

On the other hand, while the General Government also experienced large‐scale violence and roundups for forced labor, de‐Polonization was limited. There was “an unresolved conflict in [Axis] minds” over whether this region should become a pure German colony (Lukas 2012, 32). For the time being, [the Axis] administration sought to control rather than eradicate Polish culture. As Gąsiorowski (2010, 72) states, “despite the difficult food and material situation and the growing terror in the General Government, [one] was still allowed to be a Pole.”

Allowing religion was viewed as important for preventing unrest. It is estimated that 95% of the clergy in the General Government remained in the same parish during the war (Lukas 2012, 15), a marked contrast to the Warthegau. Religious life continued uninterrupted [in the General Government]; there were no mass closings of churches or restrictions on religious services [there], although training new priests was prohibited (Kłoczowski, Müllerowa and Skarbek 1986, 354–355).⁴

[…]

The arrest and execution of clergy, along with the closure of churches, especially on the scale witnessed in the Warthegau, significantly diminishes the supply of religious services. We know that vacancies created by [Fascist] arrests and executions typically remained unfilled until the war’s end.⁵

While the “orphaned” parishes were occasionally visited by the surviving priests from elsewhere, who secretly baptized children and officiated marriages, and some religious rites were perfomed by lay Catholics, an average repressed community experienced a disruption in religious practices (Huener 2021, 243).

For example, the deportation of Father Ignacy Bronszewski from his parish of Białotarsk (Włocławek diocese) to the General Government in March 1941 resulted in the interruption of worship until February 1945. As a result, most parishioners stopped attending church altogether (Huener 2021, 234). Even in parishes where the priests avoided persecution, the availability and quality of religious services were reduced.

As noted above, the [Fascist] government restricted days of the week and hours of the day when the churches could operate. Eventually, most churches were closed, vandalized, or destroyed. In addition, overwhelmed with the demand from outside their parish, some priests asked their parishioners to attend services less frequently and to keep their confessions brief (Huener 2021, 244).

Supply shifts can have important consequences on future religious behavior (Finke and Iannaccone 1993). In our context, due to habit formation, those who were unable to attend religious services during the war may not have returned to their old customs. Intergenerational religious preference transmission and peer effects (Patacchini and Zenou 2016) can lead to changes in the religious practices of future generations. Hence, this channel predicts a drop in religious observance, which can persist for decades.

[…]

Figure 7 shows a clear discontinuity in average mass attendance across four survey years (1991, 1995, 2001, and 2015). The results in Table A.6 show an effect of −6.9 percentage points, or 0.45 standard deviations, on average attendance. The estimate is largely driven by the first two years of our sample (see Figure A.1). In particular, we estimate an effect of −11.3 percentage points for the first year for which systematic data exist, which represents more than 0.5 standard deviations.

The estimates for all survey years are negative, albeit they gradually diminish in magnitude and significance. Hence, the persecution of clergy appears to have weakened the norm of church attendance, with effects lasting into the early years of the post‐transition period. […] As shown in Table 2, municipalities with high rates of priest victimization had lower church attendance [after 1989].

This relationship holds for the average across our sample period and for 1991 and 1995 in particular, the years for which we observe more sizable treatment effects in the geographic RD framework. The result is consistent with the religious supply channel detailed in Section 3.

We should note that in 44% of municipalities, all priests were removed, i.e. there is limited variation in the proportion of victimized priests, our main explanatory variable. This proxy for reduced supply of religious services also neglects the interruption of supply following the destruction or closure of church buildings, which was almost universal in the Warthegau.

Furthermore, even though mass attendance is a significant predictor of PiS [Law and Justice] and LPR [League of Polish Families] vote shares, we find higher support for these parties in municipalities with higher rates of priest victimization (Table 2). The estimate is significant at a 10% level for the 2005 election, but not for other years. We interpret this pattern as suggestive evidence for the martyrdom channel.