Regional Forecasting Table, 2016: Middle East & North Africa (MNA)
- 20 states, populations over 500,000
- Freedom Score & Freedom Status: Freedom House, Freedom in the World, 2017. (New York, FH).
- Median age: UN Population Division, World Population Prospects, the 2015 Revision. (New York, UN); for GCC States (Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, UAE), median age is for citizen-residents only, the source of which were unpublished age distributions from the US Census Bureau’s International Program Center.
- Probability of FREE (model): Cincotta, R. “Demography and Early Warning: Gauging Future Political Transitions in the Age-structural Time Domain,” J. Intelligence Analysis, 22(2): 129-148.
Note: This table shows how the 2008 forecast for the rise of at least one liberal democracy (between 2010 and 2020) among the five coastal North African states was calculated. Tunisia was projected to have the highest probability by 2020; Egypt the lowest. When the five probabilities of not being assessed as FREE (1.0-p) were multiplied together, the probability that one FREE assessment would not emerge was less than 0.03 — thus, a reasonable bet, despite the small number of cases. The forecast was realized in 2015 (Freedom House, 2016).
Several problems with this analysis:
(1) The “small number problem”. The more cases that one can lump together, the better. With few cases, its best to wager when the probability gets very low (for example, p<0.05)
(2) Libya, a state ruled by a politically monopolistic ideologue, was unlikely to switch non-violently (or with a minimum of violence) to a more liberal regime. Moreover, Libyan demographic data is notoriously uncertain. The UN has never been able to obtain a regional breakdown of Libyan data. There is a chance that it is not representative of Libyan conditions outside the major cities.
(3) Morocco’s monarchy is a source of uncertainty. The behavior of absolute monarchies in “age-structural time” is not very well tested. Data from the Authoritarian Regime Database indicate that these regimes typically do not last past a median age of 35 years. However, this hypothesis has yet to be corroborated by other means.