Criterion: If a country where the population is >50.0% Muslim as of May 03, 2026 is less than or equal to 50.0% Muslim before 2040, the market resolves YES. Otherwise, it resolves NO.
Source: https://en.wikipedia.org/wiki/Islam_by_country#Countries
Eligible Countries: Here are the countries that currently have a Muslim majority and to which the criterion applies: Maldives, Mauritania, Iran, Somalia, Afghanistan, Djibouti, Western Sahara, Algeria, Morocco, Comoros, Niger, Tajikistan, Tunisia, Palestine, Azerbaijan, Jordan, Senegal, Yemen, Libya, Mayotte, Pakistan, Gambia, Mali, Saudi Arabia, Sudan, Iraq, Turkmenistan, Kosovo, Turkey, Bangladesh, Egypt, Guinea, Uzbekistan, Indonesia, Syria, Oman, Brunei, Kyrgyzstan, Sierra Leone, Qatar, Kuwait, Bahrain, United Arab Emirates, Kazakhstan, Lebanon, Cocos (Keeling) Islands, Burkina Faso, Malaysia, Chad, Eritrea, Albania, Bosnia and Herzegovina.
1,000I think the question is kind of uninteresting in its current form - a better version would require that the proportion of Muslims also decrease meaningfully. As stated, it is to a large extent a bet on the amount of noise in estimation/reporting of religious affiliation in Eritrea, Albania and Bosnia and Herzegovina - if it is large, this is very likely to resolve YES regardless of the actual trends in religious affiliation.
@AIBear This could be an issue, although at this time, the market does not think it is all that likely. Albania's last statistic was its 2023 census, so its sample is nearly the entire population (ChatGPT claims that around 85%, around 2 million, people responded to the religion question). Bosnia and Herzegovina and Eritrea might have a different issue: they sourced the CIA World Factbook, which is now defunct, and the data is old. Nigeria is also an unreliable source, as are most Muslim countries. Actually, I think this question is interest because the data on Islam is so unreliable.
@TheUserU2 Bosnia and Herzegovina numbers in the CIA fact book match the 2013 census (only taking the "Islam" category, though if you count "Muslimanski" also as Islam, you get 51.3%). While a census gives you no sampling error, you would still expect variability over time due to willingness of participants to disclose/report their religious affiliation independent of actual religious attitudes changes.