Faith Under Fire: the Effects of Global Repression on the COG
By Mark-Phillip Pebworth & Daniel Russo
Intro
In our last analysis, we looked at where the Church of God (COG) is located worldwide and some possible explanations relating to geopolitics. We found that the number of COG congregations within a country tracks very well with Cold War-era alliances. We also identified a clear relationship between COG size and country and surprising absences in places like South Korea, Japan, West Africa, and Eastern Europe (Figure 1).
In this release, we will explore possible effects of 1) governmental repression and 2) societal hostility on where COG congregations are located globally.
Governmental repression includes efforts to control religious groups and individuals such as registration requirements, discriminatory policies, and outright bans on certain faiths. Societal hostilities, on the other hand, involve activities like vandalism of religious property, desecration of sacred texts, and violent assaults resulting in deaths and injuries.
Figure 1: Number of COG congregations per country. Countries are color-coded by the number of congregations found within that country.
Effect of Governmental Repression
Governmental persecution data shows a significant relationship between higher levels of religious persecution and fewer COG congregations (Figure 3a). For example, the lack of known COG congregations in much of North Africa, the Middle East, and Asia correlates with the areas where repression is strongest.
Figure 2: Government-mediated Religious Repression worldwide by country. The darker red the color, the greater the religious repression. This metrics measures religious repression against any religious group within that country.
3a. 
3b. 
Figure 3: Government repression, but not societal hostility, may influence the number of COG congregations present in a country. a, Violin plot of government-mediated religious repression in countries with and without a COG congregation. Each dot represents a country. The higher the dot, the more government repression occurs in that country. The colored shape in the background reveals how many countries have a given governmental repression index. The wider the shape, the more countries have that level of government repression on the y-axis. Most countries with churches have low governmental repression, while countries without churches have a much higher chance of having high governmental repression. A two-sided Wilcoxon test returned a p-value of 0.0064. b, A violin plot of societal hostility against religion for those countries with and without a church presence. Both violin plots and two-sided Wilcoxon tests revealed no significant difference between countries with and without a COG presence.
Figure 4: There are numerous countries with low governmental repression, low societal hostility, high populations, and no Church of God presence. a, A closeup plot of countries with low government repression and no churches. The size of the dot represents the population of that country in 2022. The y-axis represents the relative societal hostility of that country, and the x-axis represents the relative government-mediated religious repression. b, Same plot as a, but represents all countries without a church presence.
The violin plots above (Figure 3) illustrates just how significant the impact of governmental repression is on number of Church of God congregations. Each dot within the violin plot represent a country. If the dot falls higher up on the plot (on the y-axis), it has higher government repression. The width of the background shape, or ‘violin’ represents how many countries have a given amount of government repression. For countries without a COG congregation (red shape), the violin has two bulges, one representing countries with high repression and another for lower. For countries with a COG congregation, there’s only one bulge, representing countries with low government repression. When tested significantly, the numbers show that we have a significantly greater presence in nations with low governmental repression of religion. However, it does not explain the lack of COG presence in several counties with high governmental repression nor our absence in many nations with low governmental repression like former Soviet states and Japan.
Effects of Societal Hostility
Conversely, while there does appear to be a relationship between governmental repression and COG presence, there is no significant relationship between the level of societal hostility and COG presence. While societal hostility certainly affects quality of life for brethren in those nations where hostility is high, there is no measurable effect on Church of God presence.
Limitations
There are several limitations for this study. For one, persecution statistics are for religion as a whole, not specific to Sabbatarian or even “Christian.” We relied on overall religious repression statistics, because even ‘Christian’ nations have historically labeled the Church of God as “heretics.”
Additionally, religious persecution statistics are approximately 10 years old, on average, and newer statistics could not be found. Also, religious persecution index does not take into account other factors, including active warfare, societal violence and crime, and governmental corruption, all of which could significantly impact the ability to safely preach the message in person. These factors would not influence our ability to preach the gospel through a digital presence, but could significantly impact an individual’s access to the means by which to come in contact with that digital presence.
Areas for Further Analysis
A number of follow-up analyses are being considered for future CGN data releases, including several focused on the relative distribution of church congregations within a country and the accessibility of that country’s population to church services- i.e. what percentage of a country’s population has a COG church within an hour or so? These analyses would address to what extent a country’s population is being served by the Church locations present, or if there’s potential to expand. It would also be valuable to analyze to what extent the pattern of dispersion outlined in the above analysis of the COG mirrors that of mainstream Christian denominations.
Methodology & Code Availability
All plots represent the total number of congregations across all organizations (or the given set) for each country. When p-values are reported, a two-tailed Wilcoxon Ranked Sum test was conducted.
References
Pew Research. “Latest Trends in Religious Restrictions and Hostilities.” Last modified February 26, 2015. https://www.pewresearch.org/religion/2015/02/26/religious-hostilities/
World Bank Group. “Country population statistics.” Last modified 2023. https://data.worldbank.org/indicator/SP.POP.TOTL