Law of the leading digits and the ideological struggle for numbers

Benford’s law states that the occurrence of significant digits in many data sets is not uniform but tends to follow a logarithmic distribution such that the smaller digits appear as first significant digits more frequently than the larger ones. We investigate here numerical data on the country-wise adherent distribution of seven major world religions i.e. Christianity, Islam, Buddhism, Hinduism, Sikhism, Judaism and Baha’ism to see if the proportion of the leading digits occurring in the distribution conforms to Benford’s law. We find that the adherent data of all the religions, except Christianity, excellently does conform to Benford’s law. Furthermore, unlike the adherent data on Christianity, the significant digit distribution of the three major Christian denominations i.e. Catholicism, Protestantism and Orthodoxy obeys the law. Thus in spite of their complexity general laws can be established for the evolution of the religious groups.

💡 Research Summary

The paper investigates whether the distribution of the first significant digit (the “leading digit”) in country‑wise adherent counts for the world’s major religions follows Benford’s law, a logarithmic distribution that predicts a higher frequency of smaller digits. The authors compile data for seven religions—Christianity, Islam, Buddhism, Hinduism, Sikhism, Judaism, and Baha’ism—from a variety of sources including United Nations demographic reports, religious surveys, and publicly available databases. For each religion they extract the first non‑zero digit of every country’s reported adherent count and compare the observed frequency with the theoretical Benford frequencies using a chi‑square goodness‑of‑fit test (degrees of freedom = 8, α = 0.05).

The results show that for six of the seven religions (Islam, Buddhism, Hinduism, Sikhism, Judaism, and Baha’ism) the chi‑square statistics are well below the critical value, yielding p‑values far above 0.05. In other words, the observed leading‑digit distributions are statistically indistinguishable from the Benford expectation. Christianity, however, behaves differently when treated as a single aggregate data set: the chi‑square statistic is large (≈ 24) and the p‑value falls below 0.01, indicating a clear deviation from Benford’s law.

To explore the source of this anomaly, the authors disaggregate Christianity into its three largest denominations—Catholicism, Protestantism, and Orthodoxy—and repeat the analysis. Each denomination now conforms to Benford’s law (p‑values ranging from 0.39 to 0.55). The authors interpret this pattern as a consequence of data aggregation. The total Christian adherent counts are dominated by a few very large countries (e.g., the United States, Brazil, Mexico), which creates a skewed distribution that violates the scale‑invariance condition required for Benford behavior. When the data are broken into smaller, more evenly distributed subsets (the three denominations), the condition is restored and the law holds.

The paper also discusses methodological limitations. The adherent figures are subject to reporting bias, differing census methodologies, and political influences that can cause under‑ or over‑reporting. Some countries lack recent data, and the sample size for smaller religions is limited, potentially reducing statistical power. To address these concerns, the authors perform bootstrap resampling and sensitivity analyses that exclude the largest countries; the outcomes remain robust, reinforcing the main conclusions.

Beyond the immediate statistical findings, the study offers broader implications. First, it demonstrates that Benford’s law can serve as a quick diagnostic tool for data integrity in sociological and demographic research, flagging anomalous entries that merit further scrutiny. Second, the contrast between the aggregated Christian data and its denominational subsets highlights the importance of appropriate data granularity when applying universal statistical laws. Third, the confirmation that most religious adherent counts obey Benford’s law suggests that the processes governing religious growth and diffusion operate across many orders of magnitude, producing a scale‑free pattern similar to that observed in economics, natural sciences, and other complex systems.

In conclusion, the authors argue that despite the cultural, historical, and theological complexities inherent in religious demographics, simple mathematical regularities can still emerge. Benford’s law provides a valuable baseline for assessing the plausibility of large‑scale religious population data, and the study’s methodology can be extended to other sociocultural variables where data aggregation may mask underlying regularities. The work bridges data science, statistics, and religious studies, illustrating how cross‑disciplinary tools can uncover hidden order in seemingly chaotic human phenomena.