SPECIAL REPORT

Population Age Structure and Its Relation to Civil Conflict: A Graphic Metric

ecent studies suggest that a large “youth experience conflict in the next decade. bulge”—a youthful population age • Of countries without recent civil conflict, 24 RICHARD P. R structure—can increase the risk of the percent of all states with more than 60 per- CINCOTTA and onset of civil conflict and political violence cent of their population under 30 years of ELIZABETH LEAHY (Urdal, 2006, Cincotta et al., 2003). These stud- age (i.e., a young age structure) experienced ies exclude states with a recent history of civil at least one incident of civil conflict during Richard P. Cincotta is a conflict, reasoning that they are already highly the following decade. Among countries with demographer and consultant vulnerable to persistent and re-emerging violence less than 60 percent under 30 years of age, whose research has focused (Collier et al., 2002). Can these two quantifiable just 7 percent experienced civil conflict. on the course of the demo- variables—population age structure and recent Conclusion: One-fourth of all non-con- graphic transition and on pat- history of civil unrest—be used to project risks of flict countries with young age structures terns of human migration. His civil conflict a decade into the future? will likely experience a new civil conflict published work strives to We conducted separate analyses of three during the next decade. characterize these trends and decades—the 1970s, 1980s, and 1990s—using • About 86 percent of all countries that experi- their relationships to natural the UCDP/PRIO Armed Conflict Dataset, enced a new outbreak of civil conflict had age resource dynamics, to human which identifies conflicts with at least 25 battle- structures with 60 percent or more of the pop- health trends, and to the related deaths per year (Gleditsch et al., 2002; ulation younger than 30 years of age. political stability and legitima- UCDP/PRIO, 2006) and age-structural data Although country age structures in many cy of states. from the UN Population Division (2005). regions matured between 1970 and 1999, this Countries with average immigration or emigra- “outbreak benchmark” remained virtually con- Elizabeth Leahy is a research tion rates exceeding 0.5 percent between 1970 stant. Conclusion: This “60-percent-under- assistant at Population Action and 2000 were removed, as were the five most 30” benchmark could serve as a means to International (PAI), where she populous countries (China, , Indonesia, identify and track a state’s demographic researches international USSR/Russia, and the United States) and coun- risks of civil conflict (see figures below). demographic trends and poli- cy. She is the author of The tries with populations less than 100,000. Policy Implications Shape of Things to Come: Why Age Structure Matters to Findings a Safer, More Equitable World If these relationships accurately reflect some of (PAI, 2007). She is a graduate • Of all countries that experienced civil con- the risk factors of civil conflict, future risks could student in international affairs flict during the last five years of the prior be reduced by (a) supporting sustained efforts to at George Washington decade, 76 percent experienced civil conflict mediate incipient civil conflicts; and (b) by sup- University. for at least one year during the following porting programs and policies that promote decade. Conclusion: Three-fourths of all advancement along the path of the demographic countries with recent conflict will likely transition in countries with young age structures. 55

SPECIAL REPORT • POPULATION AGE STRUCTURE AND ITS RELATION TO CIVIL CONFLICT Figure 1: Path of the Demographic Transition, 1970 and 2050

Each point represents a coun- 80% try’s position, plotted in terms 1970 (estimate) of the most mature (aged 60+ 2050 (medium projection) Demographic Transition years) and least mature (aged 0 60 to 29 years) segments of the population. The demographic transition function (y = -20.47Ln(x) + 102.59) plots 40 the average path of expected maturation of population age structures during the course of of total population 20 the demographic transition— the change from high birth/high Ages 0–29 as a proportion death rates to low birth/low death rates. 0 0 10 20 30 40 50%

Ages 60+ as proportion of total population

Figure 2: Distribution of Outbreaks of Civil Conflict, 1970-1999

100 The distribution of outbreaks of 1970–79 civil conflict (in countries with- 1980–89 out recent conflict) along the 1990–99 Demorgraphic Transition path of the demographic transi- 80 tion during three decades of analysis: 1970-79, 1980-89, 1990-99. Note the clustering of outbreaks in the region of 60 youthful age structures.

40 of total population Ages 0–29 as a proportion 20

0 0102030

Ages 60+ as proportion of total population 56

ECSP REPORT • ISSUE 12 Figure 3: Number of Outbreaks of Civil Conflict by Age Structure, 1970-1999

< 60% youth >_ 60% youth 25 The ratio between the number of countries that experienced 20 outbreaks of civil conflict on the younger side of the outbreak benchmark—where 60 percent 15 of the population is under age 30—ranged from 11:1 (1980-89) to 5.25:1 (1970-79). 10 of civil conflict

Number of outbreaks 5

0 1970–79 1980–89 1990–99

Figure 4: Countries With Age-Structural Risks, 2005

100% The zone of greatest concentra- Zone of Risk tion of outbreaks of civil conflict Demographic Transition Countries is shaded, and lies among age 80 structures that are younger than the outbreak benchmark— where 60 percent of the adult population is below the age of 60 30 years. Country positions on this graph are for 2005 (accord- ing to data from the UN Population Division, 2005). The 40 majority of states in this zone

of total population are in sub-Saharan Africa and western Asia. Ages 0–29 as a proportion 20

0 0 5 10 15 20 25 30%

Ages 60+ as a proportion of population 57

SPECIAL REPORT • POPULATION AGE STRUCTURE AND ITS RELATION TO CIVIL CONFLICT Countries With Very Young and Youthful Age Structures, 2005

Swaziland Angola Ethiopia Micronesia Uzbekistan Zimbabwe Congo Mozambique Mauritania Uganda United Republic of Solomon Islands Gabon Bangladesh Rwanda Tanzania Guinea-Bissau Bhutan Tonga Zambia Haiti Timor-Leste Oman Gambia Lesotho Cape Verde Afghanistan Turkmenistan Kenya Côte d'Ivoire Cambodia Egypt São Tomé and Principe Senegal Palestinian Territories Belize Bolivia Burundi Nigeria Guatemala Guinea El Salvador Mali Arab Republic Namibia Sudan Saudi Arabia Yemen Syrian Tajikistan Libya Dominican Republic Burkina Faso Chad Papua New Guinea Morocco Botswana Togo Ghana Jordan South Africa Eritrea Central African Republic Honduras Vanuatu Kyrgyzstan Malawi Nicaragua Madagascar Nepal Peru Liberia Cameroon Sierra Leone Samoa Niger Comoros Djibouti Algeria Fiji Democratic Republic of Maldives Equatorial Guinea Mongolia the Congo Benin Laos Paraguay

References (2002). “Armed conflict 1946-2001: A new dataset.” Journal of Peace Research 39(5), 615-637. UN Population Division. (2005). World population Cincotta, Richard P., Robert Engelman, & Daniele prospects: 2004 revision. New York: United Nations. Anastasion. (2003). The security demographic: Uppsala Conflict Data Program (UCDP) & Population and civil conflict after the cold war. International Peace Research Institute, Oslo Washington, DC: Population Action International. (PRIO). (2006). UCDP/PRIO Armed Conflict Collier, Paul, Anke Hoeffler, & Mans Söderbom. Dataset codebook: Version 4-2006. Oslo, Norway: (2001). On the duration of civil war. Washington, UCDP & PRIO. DC: World Bank Development Research Group. Urdal, Henrik. (2006). “A clash of generations? Youth Gleditsch, Nils Petter, Peter Wallensteen, Mikael bulges and political violence.” International Studies Eriksson, Margareta Sollenberg, & Håvard Strand Quarterly 50, 607-629.

58

ECSP REPORT • ISSUE 12