This post is a response to the questions and comments
that my fellow demographers Elizabeth Leahy Madsen, Jack Goldstone, and Jennifer Dabbs Sciubba put to my recent assessment of Tunisia’s chances for democracy
. I’ve divided my response to address the principal questions. (Note: Throughout, “young adults” refers to 15-24 year olds, and “working age population” refers to those between 15-64 years old.)
1. Tunisia’s age structure is still quite young – aren’t the effects associated with youth bulge still at work?
Tunisia’s age structure (median age, 29 years) is in the early stages of transiting between the instability that typically prevails in countries with youthful age structures (median age <25), and the stability that is typical of mature structures (median age between 35 to 45). I classify Tunisia’s age structure as intermediate (between 25 to 35). The Jasmine Revolution has featured a mix of both types of sociopolitical behavior: some violence and property damage, perpetuated by both the state and demonstrators; evidence of a mature, professionally-led institution (the Tunisian army); and demonstrations that are peaceful and in which women and older people participate.
By any legitimate youth bulge measure, Tunisia’s age-structure is similar to that of South Korea, Taiwan, and Thailand’s during the mid-1990s. In 2010, Tunisia’s proportion of young adults in the working age population was almost precisely the same as South Korea’s in 1993. South Korea’s median age in 1993 was 28.5, compared to Tunisia’s 29 today.
Consistent with these numerical similarities, Tunisia’s political behavior is not much different than the sociopolitical events observed in South Korea, Taiwan, or Thailand in the early 1990s when these countries experienced an age structure of similar maturity. As these countries matured, political violence and destructive protests occurred less frequently but dissidence lingered in more peaceful, isolated incidents. For example, South Korea experienced several deadly anti-U.S. demonstrations in 1988-89 and only a few minor incidents in the early 1990s. By 1993, the public was demonstrating against extremism, and public protest had turned peaceful and symbolic.
Even today, public protest in Thailand (median age, 33 years) is commonplace. Thai political rights are constrained, and accordingly the country has been dropped in Freedom House’s annual assessment of political freedoms from “free” to “partly free.” However, unlike the student and worker riots that occurred sporadically throughout the 1970s and 80s, Thai demonstrations are typically non-confrontational political rallies attended by T-shirt-clad grandparents and families who bus into town for the event.
2. Why use the population’s median age rather than the youth bulge measures that political demographers (including me) have previously employed with considerable success?
I’m experimenting, and so far, I’ve found that median age replicates prior published results concerning civil conflict and stable liberal democracy. Median age’s ability to span the entire length of the demographic life-cycle of the state is its primary advantage. Youth bulge indicators do not; neither do indicators focused on working-age adults, nor those measuring seniors. In addition, median age integrates many factors that change in parallel with the age-structural transition, including income, education, women’s participation in society, secularization, and technological progress. Still, median age doesn’t track the youth bulge measures perfectly, so why use it?
Currently, foreign affairs policymakers see few linkages between a country’s (1) risk of civil and political violence; (2) its propensity to accumulate savings and human capital; (3) its chances of attaining stable liberal democracy; (4) the challenges that arise to adequately funding pensions and senior healthcare; and (5) rapid ethnic change in low fertility societies. Political demographers understand that these effects on the state are indeed related and that the rising and ebbing probabilities associated with these effects occur sequentially. Understanding this sequence is key to understanding the world’s international relations future. Median age allows political demographers to view that sequence.
As Jennifer Sciubba points out, the disadvantage of median age is its apparent lack of resonance with theories that have historically informed political demography, including theories of cohort crowding, dependent support, and life-cycle savings. I don’t believe that political and economic demographers should (or will) abandon these indicators, which help them observe the inner workings of age-structural phenomena. Nonetheless, I find it useful to make analysts aware of the advantageous and disadvantageous pressures that age structure exerts across the entire demographic life cycle of the state.
3. Can the deposed regime’s multi-faceted problems be captured by the age-structural transition?
No, age-structure cannot account for leadership – clumsy or deft, corrupt or honest. The method is limited to predictions that draw on sociopolitical behaviors that are associated with age structures, or by knowledge gained from deviations from these predictions. Nonetheless, I question the value of many of the after-the-fact observations of the Ben Ali Regime. Tunisia’s fallen regime was indeed oppressive, corrupt, and nepotistic – but so are most authoritarian regimes in Asia and Africa. Lack of job creation and preferential access to employment are valid grievances across much of the developing world, but the lack of informal sector statistics renders country comparisons difficult.
Some analysts have hypothesized that global warming has been a contributor (also applied to Egypt), others point to economic globalization and pressures on Tunisia’s middle class. Whether these assertions are wrong or right, it is difficult for me to see how such post hoc observations add to analysts’ knowledge of regime change or democratization or help them explain why other countries with similar problems have not undergone similar sociopolitical dynamics. In contrast, hypotheses based on quantitative relationships between age structural indicators and sociopolitical behaviors generate testable and repeatable predictions that can be checked and held accountable after an event.
What I find most surprising about the age-structural approach to predicting liberal democracy is how often states ascend to liberal democracy as they approach, or pass, the 0.50 probability mark. If either “triggers” of regime change or key institutions were very important, this observation would be unlikely.
A Concluding Note on Political Demography
If political demographers are serious about advancing policymakers’ ability to understand the present and future of global politics and security, political demography will have to become a scientific discipline – a field of study in which assertions are consistently tied to data and tested whenever possible. In many cases, though not all, we’re lucky – demographers provide us with timely estimates and projections at the national level. Nonetheless, for our field to succeed, political demographers must take full advantage of these data, encourage sub-national data to be collected and published, and make a clean break from the tradition of conjecture that currently pervades international relations.
Richard Cincotta is a consulting political demographer for the Wilson Center’s Environmental Change and Security Project and demographer-in-residence at the Stimson Center.
Sources: CIA World Factbook, Freedom House, Jadaliyya.
Image Credit: Adapted from “Viva the Tunisian Revolution,” courtesy of flickr user freestylee (Michael Thompson).