Paper Critique: ‘Trade and Geography in the Spread of Islam’

Michalopoulos, S., Naghavi, A., & Prarolo, G. (2018). Trade and Geography in the Spread of Islam. The Economic Journal, 128(616), 3210–3241.

Michalopoulos et al. (2018) are a research team consisting of Stelios Michalopoulos, who studies political economy at Brown University, Alireza Naghavi, who studies trade and geography and the effects of economics on geographic conditions at the University of Bologna, and Giovanni Prarolo, who is from the same facility. The works of each author are widely published, and include studies of political economy and international trade, and economic migration. In addition, the researchers have worked together on several research publications, involving geographic trade considerations and economic conditions, including inequality studies.

The current research investigates the historical factors that led to the spread of Islam. Michalopoulos et al. (2018, p. 3210) contend that “[r]eligion is significantly correlated with a range of economic and political outcomes both within and across countries”. Using this observation, the authors analyze the spread of Islam and investigate the historical relationship between Islamic culture, religion and trade. Moreover, it is argued in the paper that Islam diversified within regions ecologically similar to the Arabian Peninsula. Finally, an empirical analysis of trade is presented to argue that a country located on a historical trade route is more likely to be Islamic.

Selected historic trade routes, including those along the Silk Road, the Red Sea, Northern Africa, and Persian Empire, are highly correlated with the integration of the Muslim religion. The combination of trade advantages, including religious cooperation and the ability to trust individuals with similar cultural foundations, is argued to be a source of valuable contacts that can expand trade and enable economic growth. Moreover, it is noted that the “rules governing commercial activities naturally favouring Muslims over non-Muslims” (Michalopoulos et al., 2018, p. 3213) would allow the proselytization of Islamic ideas as Sufi preachers could easily travel across trade routes performing missionary activities.

The paper analyses the systematic development of commerce across the global trade routes analyzing both the pre-Islamic ports, harbors and Roman roads and comparing that to the later pre-1800 trade activity that developed, demonstrating the correlation between the pre-colonial economic exchange routes. With this information, the authors present a cross-country analysis and a potential model of the uptake of Islam against a linear regression modelled over the distance from the main societies that had converted to Islam and the distance from the trade route.

Unfortunately, the statistical model is weak, with the statistical significance pushed to the  level to obtain the outcome sought by the scholars. Table 2 (Michalopoulos et al., 2018, p. 3225) shows that all reported results are insignificant at . Similarly, table 3 (Michalopoulos et al., 2018, p. 3226) is only significant for values related to tests of ethnic groups outside of the primary research question. Nevertheless, the main hypothesis of distance to trade routes, distance to the coast, and distance to Mecca are all reported at values over, meaning that the researchers cannot reject the null hypothesis of no significant impact.

Vermeesch (2009, p. 443) references the standard methodology used in statistical hypothesis testing and points out that results will be invalidated if sufficient data is applied. In response, Sornette and Pisarenko (2011, p. 64) argue that carefully constructing a hypothesis test is critical. These points are crucial to the analysis of the paper by Michalopoulos et al. (2018), as rejecting the null hypothesis is rarely achieved at any level of significance that warrants consideration. As Curran-Everett and Benos (2004, p. R247) state, “[i]f  you plan a study in the hopes of finding an effect that could lead to a promising scientific discovery, then 0.1” may be appropriate. However, this confidence level only provides low certainty against an effect.

Consequently, while the study provides some interesting preliminary results, the forces behind the formation and spread of Islam are inadequately described by the model presented. The research demonstrates that the spread of Islam has correlations with the trade routes, but the results may purely be correlational because of the migration paths of people, and no examination of causality has been conducted. The paper demonstrates a promising area of study, and more research would need to be conducted to potentially conclude that the integration of Islamic and non-Muslim groups is related to movement and trade conditions, as posited.


Curran-Everett, D., & Benos, D. J. (2004). Guidelines for reporting statistics in journals published by the American Physiological Society. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 287(2), R247–R249.

Michalopoulos, S., Naghavi, A., & Prarolo, G. (2018). Trade and Geography in the Spread of Islam. The Economic Journal, 128(616), 3210–3241.

Vermeesch, P. (2009). Lies, Damned Lies, and Statistics (in Geology). Eos, Transactions American Geophysical Union, 90(47), 443–443.

Never miss a story from Craig Wright (Bitcoin SV is the original Bitcoin)