Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Treatment of missing data determined conclusions regarding moralizing gods


The Original Article was published on 20 March 2019

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: Moralizing gods across 12 key regions.
Fig. 2: Comparison of the original logistic regression model and a reanalysis removing unknown outcomes.

Data availability

Supplementary data are available at All software is freely available under Creative Commons License CC BY-NC-SA 4.0. Source materials are available at

Code availability

Re-analysis code is available at All software is freely available under Creative Commons License CC BY-NC-SA 4.0.


  1. 1.

    Whitehouse, H., et al. Complex societies precede moralizing gods throughout world history. Nature 568, 226–229 (2019).

    ADS  CAS  Article  Google Scholar 

  2. 2.

    Seshat: The Global History Databank (Evolution Institute & Seshat Project, 2015);

  3. 3.

    Watts, J. et al. Broad supernatural punishment but not moralizing high gods precede the evolution of political complexity in Austronesia. Proc. R. Soc. Lond. B 282, 20142556 (2015).

    Google Scholar 

  4. 4.

    Little, R. J. A. & Rubin, D. B. Statistical Analysis with Missing Data (John Wiley & Sons, 2014).

  5. 5.

    McElreath, R. Statistical Rethinking: A Bayesian Course with Examples in R and Stan (Chapman and Hall/CRC, 2020).

  6. 6.

    Gelman, A. & Hill, J. Data Analysis Using Regression and Multilevel/hierarchical Models. (Cambridge University Press, 2007).

  7. 7.

    Wheatley, P. The Pivot of the Four Quarters: A Preliminary Enquiry into the Origins and Character of the Ancient Chinese City (Edinburgh University Press, 1971)

Download references


We thank H. Whitehouse and co-authors for responding to our request for clarification and to our early notification of our intent to submit this report, and for making their code, source material, and data public. We thank three anonymous reviewers as well as the Department of Human Behavior, Ecology and Culture at the Max Planck Institute for Evolutionary Anthropology, H. Colleran, J. C. Jackson, R. McElreath, E. Ready and J. Watts for feedback, and A. Ashtari, A. Barnett and T. Hwang for research and administrative support.

Author information




R.S., B.G.P., M.M., M.L., J.H., R.D.G., B.B. and Q.D.A. designed the reanalysis. R.S., M.L., B.G.P. and B.B. performed the re-analyses. R.S., M.L. and B.G.P. reviewed code. E.S. and M.W.M. reviewed and vetted historical coding. All authors wrote the manuscript.

Corresponding author

Correspondence to Bret Beheim.

Ethics declarations

Competing interests

R.D.G., J.H., M.M., M.W.M., E.S. and R.S. are involved in the Database of Religious History (DRH) project, another freely available online historical database.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 The first appearance of writing and moralizing gods across NGAs.

The solid line indicates when writing and moralizing gods (MGs) are first recorded in the same century, and the dashed lines show when writing appeared 100 years before moralizing gods and when moralizing gods appeared 100 years before writing. NGAs are coloured by whether social complexity data are available both before and after the appearance of moralizing gods or not. Only natural geographic areas with social complexity data available both before and after the appearance of moralizing gods were included in the analysis (and only these natural geographic areas are shown in Fig. 1). It must be noted that while writing first appears at 2500 bc in the Kachi Plain, it is absent for the subsequent two polities in the dataset, and does not reappear until 300 bc —the same time as the first appearance of moralizing gods.

Extended Data Fig. 2 Distribution of social complexity score by ‘moralizing gods’ outcome status.

Before statistical analyses were performed in Whitehouse, et al.1, all ‘unknown’ or ‘suspected unknown’ (NA) cases were treated as moralizing gods ‘absent’ (0) without explicit description in the manuscript. In box plots centre line shows median, box limits indicate upper and lower quartiles and whiskers span 1.5× interquartile range). N = 801 observations.

Extended Data Fig. 3 Social complexity before and after the appearance of moralizing gods.

Dots represent mean social complexity as calculated by Whitehouse, et al.1 (a combination of population and territory size, infrastructure, hierarchy, and other factors, standardized between 0 and 1) collapsed across natural geographical area. Data are mean ± s.e.m. The shading of lines connecting the dots in a reflects the weight that the difference (d1,..., d7; on d1 and d2 are shown) between the social complexity at time point n and time point 0 in the t-test analysis performed by Whitehouse, et al.1 (that is, differences in social complexity are highest around time 0, hence driving the forward bias). Note that the increase in social complexity from time point −100 to 0 is coded as pre-moralizing gods, while the complexity often arrives via conquest or mission together with moralizing gods. a, The ‘first appearance’ of moralizing gods (MGs) in the archaeo-historical records follows a sharp increase (39%) in social complexity in the 12 geographical areas. b, The sharp increase in social complexity just before the appearance of moralizing gods is partially caused by ascribing properties of conquerors to the conquered regions in the Deccan, Kachi Plain and Sogdiana regions. c, Similarly, regions receiving moralizing gods via mission (Kansai, Niger Inland Delta and Orkhon Valley) experience a sharp increase in social complexity. d, The remaining six natural geographical areas where moralizing gods were not first recorded through conquest by a larger empire or through mission show a steady rise in social complexity.

Supplementary information

Supplementary Methods

This file provides additional ethnographic justification and statistical details for the re-analyses of Whitehouse, et al. (2019), including Supplemental Figures 1-5, Supplementary Tables 1-4 and Supplementary References.

Reporting Summary

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Beheim, B., Atkinson, Q.D., Bulbulia, J. et al. Treatment of missing data determined conclusions regarding moralizing gods. Nature 595, E29–E34 (2021).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links