SLIDE 1
Social Conflict and a Migrant Youth Bulge in Urban Sub-Saharan Africa - - PDF document
Social Conflict and a Migrant Youth Bulge in Urban Sub-Saharan Africa - - PDF document
Social Conflict and a Migrant Youth Bulge in Urban Sub-Saharan Africa Ashira Menashe-Oren ABSTRACT Sub-Saharan Africa (SSA) is the region of the world with the greatest expected future population growth and where urbanisation is expected to
SLIDE 2
SLIDE 3
3 2008; Urdal and Hoelscher 2009; Yair and Miodownik 2014). Two particular methods have been most commonly used to operationalize the YB. One is as the proportion of young adults in a population,
- ften measured as the number of 15 to 24 year olds of the total adult population. The other is a YB as
a relative cohort, where large relative cohorts make it easier for conflicts to erupt (Staveteig 2005). Described as an incendiary factor of the Arab Spring (Hvistendahl 2011), the YB is an enabling factor to violent conflict. The focus on young adults is based on a diverse literature claiming that young adults are relatively easily mobilised, with fewer responsibilities to families and careers – frequently not yet married and not fully integrated into the job market (Mesquida and Wiener 1999). The opportunity costs for political violence are low, especially amongst large cohorts (Macunovich 2000). According to Easterlin’s relative cohort size hypothesis, when a relatively large cohort comes of age economic frustrations emerge- from strains on the education system, unemployment and reduced wages. This may in turn enable political instability and armed conflict (Staveteig 2005). With high unemployment and lack of opportunities, especially in the formal work sector, the alternative costs to engaging in violent action are low (Collier and Hoeffler 2004). Turning to civil conflict can also be considered a legitimate way to redress perceived economic, political and social inequalities when there is little to
- lose. Young adults may be alienated and marginalised (Sommers 2010), politically excluded or with
unmet expectations. Deprived youth may aspire for something better and thus be motivated to take action (Pinard 2011). Notably, when referring to a youth bulge, it is most often a male YB. Proportionately large male youth cohorts have been found to have a significant effect on regime type and change, with democracies more likely to collapse (Weber 2013). Similarly, male youth bulges affect the frequency and severity of conflicts (Mesquida and Wiener 1999). Men are more susceptible to violence from a behavioural ecology perspective, as they strive for mate acquisition (Mesquida and Wiener 1999), and have greater taste for risk (Wilson and Daly 1985). This is also evident by the accident hump in male
SLIDE 4
4 mortality profiles by age where young adult men have higher mortality rates (Hannerz 2001). Furthermore, skewed sex ratios (a high proportion of men to women in a population) may pose a security threat with vast implications on marriage markets, drug use, crime and prostitution (den Boer and Hudson 2004; Dyson 2012). Migration and Urban Youth Bulges Urbanisation is a result of the demographic processes of mortality, fertility and migration. Some studies show that the primary component of urbanisation is urban natural increase (Preston 1979), while many find that urban growth is predominantly due to rural-to-urban migration (Keyfitz 1980; Rogers 1982). Urban age structures tend to be concentrated in the production and reproduction ages during earlier stages of the demographic transition. The age selectivity of rural-to-urban migration which occurs disproportionately amongst the young (Montgomery, 2003), and particularly amongst males, reinforces an urban youth bulge. The proportion of young migrants in the urban population may be fundamental to understanding the effect of an urban YB on social unrest for a number of reasons. Firstly, migration may shift the ethnic composition of urban populations and promote conflict (Fearon and Laitin 2011; Goldstone 2002). The heterogeneity of urban areas can be a source of instability, when several ethnic, religious or regional groups are in close social contact (Cincotta et al. 2003). Secondly, associated with urban growth, the job market and economy may struggle to keep up with incoming migrants (Goldstone 2002). A surge
- f young workers contributes to under employment and low wages. Also, coming from rural areas,
migrants are less educated than their urban counterparts (Sahn and Stifel 2002), making it harder for them to find opportunities in the formal sector of the economy. Migrants are thus more likely to experience economic marginalisation and relative deprivation (Gizewski and Homer-Dixon 1995). Low paid migrants are peripheralised through market relations and excluded from different segments of society (Cook 2015). Thirdly, migrants may feel alienation and marginalisation in cities (Cook 2015), finding it hard to adjust socially and psychologically (Gizewski and Homer-Dixon 1995). Finally, by
SLIDE 5
5 moving to the city migrants have greater opportunity for collective political action and mobilisation (Gizewski and Homer-Dixon 1995), potentially being able to raise grievances that previously affected them or continue to affect their rural family and community. I hypothesise that a relatively higher proportion of young rural-to-urban migrants (a migrant YB)
- verall increases the likelihood and frequency of social conflict in urban settings. In particular:
1) A large migrant YB neither increases nor decreases the likelihood and frequency of conflict related to the economy, jobs and assets. Naturally these issues are mostly dealt with in strikes. While it is expected that migrants have trouble in the job market, facing unemployment, it is not expected that their economic marginalisation would be reflected in participation in strikes. Labour strikes are generally held by people employed in the formal sector. Nonetheless, a large proportion of migrants may create more competition for non-migrants in the job market, leading to non-migrant dissatisfaction. 2) A large migrant YB increases the likelihood and frequency of conflict related to ethnic discrimination and religious issues. Considering migrants are typically from different ethnic groups and that cities are often ethnically heterogenic, violent urban conflict may evolve when
- ne ethnic group has a grievance against another (Esteban et al. 2012; Fearon and Laitin 2011;
Higashijima and Houle 2017). 3) To a lesser extent, a large migrant YB may increase the likelihood and frequency of conflict related to human rights and democracy as well as elections. When migrants experience relative deprivation, they may pertain to human rights (such as freedom of speech or rights to trade), driving them to participate in such social conflict. 4) Finally, a large migrant YB is expected to increase somewhat the likelihood and frequency of social conflict related to domestic war, violence and terrorism. Migrants may protest wars fought in rural regions over natural resources, affecting their family and place of origin. DATA AND METHODOLOGY
SLIDE 6
6 The Social Conflict Analysis Database (SCAD) contains data on social conflict in 42 Sub-Saharan African countries from 1990 to 2014, covering 7200 instances of protests, riots, strikes and government related violence (Salehyan et al. 2012). 1 The data is based on systematic aggregation of press reports. This has been found to be an appropriate means of analysing low intensity conflicts common in SSA (Bocquier and Maupeu 2006). Politically significant conflict events are identified through keyword searches of news wires. These events are mostly large-scale, drawing many participants, including significant acts of violence and threatening to political stability. On average there are four conflict events every year per country. The data is georeferenced and indicates whether the conflict was in rural or urban locations. The conflict data is merged with population data to see how urban youth bulges may affect conflict
- ccurrence, creating a cross-section time series dataset. United Nations Department of Economic and
Social Affairs’ (UN-DESA) Population Division data on urban and rural populations by age and sex - URPAS (United Nations 2014a) - is used from 1980 to 2015 to calculate YBs for men for every five years and for each country. The URPAS data provides best estimates of rural and urban populations by age and sex, building on census data, population registers and demographic interpolation (United Nations 2014b). The data also allows for measuring migration estimates using the census survival ratio method (CSRM), to obtain a measure of the proportion of migrants among the urban YB. The CSRM provides an approach to estimate migration flows between rural and urban areas (Hamilton and Henderson 1944; Preston 1979), especially useful considering little data is available of such migration in SSA (de Brauw et al. 2014). The standard CSRM approach estimates survivorship for each age group between two censuses exactly ten years apart for the population as a whole. These total cohort survival ratios are the backbone upon which the migration estimates are based. The survivorship levels are adjusted for the urban population.2 The urban cohort survival ratios are then
1 SCAD does not include cases of conflict or violence that are coded as civil war in the Uppsala Armed Conflict
Database.
2 Urban survival is assumed to be 25% higher than rural.
SLIDE 7
7 used to predict the expected number of people in each urban age group at the time of the second
- census. The difference in the expected number of urban people and the actual number measured in
the second census provides an estimate of net rural-to-urban migration. I use an adjusted CSRM with the URPAS data, estimating migration every five years, based on rural cohort survival ratios. This approach is preferable and may be more robust in the SSA setting where urbanisation levels are low (Menashe-Oren and Stecklov forthcoming ). Important limitations to the CSRM, valid for this adjusted approach too, have been discussed elsewhere particularly regarding potential bias from international migration and reclassification (Menashe-Oren and Stecklov forthcoming; Moultrie et al. 2013; Preston 1979). It is worth noting that the migration estimates are a combination of both net rural-to-urban migration as well as reclassification, where formerly rural areas are redefined urban once they pass a threshold of “urbaness”. The CSRM produces rural-to-urban migration estimates for all 42 SSA countries with conflict data for every five year interval between 1990 and 2015. The migration estimates are the annual net number
- f rural-urban migrants – where a positive number indicates there were greater flows from the rural
to urban sectors than from the urban to rural. These migrants are relatively new in urban areas- having moved on average two and a half years ago. Thus the proportion of migrants in urban populations estimated relates to recent migrants. The proportion would be even higher if considering migrants who moved earlier. A migrant YB is measured as the ratio of young adult migrants aged 15 to 24 to older working aged adults aged 25 to 59 in the urban sector.3 Regression models are utilised to predict the relative contribution of rural-to-urban migration as part
- f the urban YB to cases of conflict. A generic version of the model is:
Pr(𝑧𝑗,𝑒) = 𝛽 + 𝑦𝑗,𝑢−10𝛾1 + 𝑨𝑗,𝑢𝛾2 + 𝑑𝑗,𝑢 + 𝜁𝑗,𝑢
3 The youth bulge has been measured in various ways - aged 15-29 or aged 15-24, as percentage of total adult
population or out of total population. I test for sensitivity and run the main model with a bulge defined by ages 15-29.
SLIDE 8
8 where Pr(𝑧) is the outcome of interest - the probability of urban social conflict. 𝑦𝛾1 is a vector of the key explanatory variables; depending on the model it represents the urban YB, the proportion of rural- to-urban migrants aged 15 to 24 to adults in urban area aged 25 to 59 and the non-migrant urban YB. 𝑨𝛾2 is a vector of the control variables, c indicates the time-invariant fixed effects on countries and 𝜁 indexes residuals. The t signifies temporal effects and i country-level analysis. The models are run in Stata 11 using fixed effects (FE) binary logit regression models. Fixed effects Poisson models are also run when the outcome is the number of conflicts. Country FE models account for between-country sources of heterogeneity and were chosen to control for time invariant variables such as ethnic and religious fractionalisation and climate. A Hausman test indicated that a FE model is preferred to a random effects model– the test was significant at 0.01. The models are run for all social conflict cases in urban settings and are also separated according to the main issue at the source of the conflict – 1) human rights and democracy, 2) economy, jobs and assets, 3) domestic war, violence and terrorism, 4) elections, 5) ethnic or religious discrimination and 6) other or unknown issues including environmental degradation, foreign affairs and education. The key explanatory variables (youth bulges) are lagged by five years to account for temporal dependence between observations within countries. The models take into account time (1990 to 2015), country level economic development, political regime, unemployment, urban population size (logged), percent
- f population in urban settings and the urban youth sex ratio.
Economic development can predict conflict as a poor country is constrained in meeting the demands
- f its citizens, while a wealthy country can easily distribute resources and dampen any dissatisfaction
(Fearon and Laitin 2003). It is measured as gross domestic product (GDP) per capita in $US (International Monetary Fund 2014) – standardised to mean of zero and standard deviation of one because of high variability. The effect of regime type is argued as a measure of weak governance; established democracies or harsh autocracies are both less likely to experience conflict than unstable regimes (Hegre et al. 2001). A democracy-autocracy score is taken from the Polity IV Project of the
SLIDE 9
9 Centre for Systemic Peace and provides a convenient measure of general regime effects (Centre for Systemic Peace 2015). The score ranges from -10 (strongly autocratic) to +10 (strongly democratic). The mean polity score for African countries included in the analysis is 1.41. Although questionable, unemployment is often cited as a strong case for people to engage in violence (Cramer 2011; Goldstone 2002; Hvistendahl 2011) - from gang participation to civil wars, including low opportunity costs of violence (Collier and Hoeffler 2004). Unemployment is measured as the percent of male labour force unemployed (World Bank 2016). Population size can account for conflict when the larger the population the greater likelihood of fractionalisation and the more recruits available. The proportion of the population urban, like urban growth, can affect violent conflict (Cincotta et al. 2003; Goldstone 2002; Wirth 1938), especially when interacting with other factors (Gizewski and Homer- Dixon 1995). A sex ratio with higher proportions of men, can affect marriage markets and increase undesirable behaviours (Dyson 2012). A youth sex ratio is measured as the proportion of 15-24 year
- ld males to females in the urban sector. The measures of population size, proportion urban and sex
ratios are based on the URPAS data. Ethnic and religious fractionalisation are not included in the models as it is unvarying over the 25 year period analysed, but is controlled for in the FE models. RESULTS Initially, it is worth considering the dependent variable, probability and number of social conflicts, and the main independent variables, the migrant YB and the non-migrant urban YB, for the main 1008 sample (42 countries over 24 years). The probability of at least one social conflict (on any issue) in an SSA country is 0.72 over the whole period, ranging from a low of 0.45 in 1990 to a high of 0.9 in 2000. The mean number of conflicts per year is 4.4, from 1.7 on average in 1990 to 11 in 2012. Although analysis in this article is according to the main issue at the source of the conflict, these issues are often dealt with by specific means including demonstrations, riots and strikes (spontaneous or
- rganised) and pro- anti- extra- or intra-governmental violence (grouped here as “other”). Table 1
maps the conflicts by issue and type of conflict. Urban conflicts regarding human rights and democracy
SLIDE 10
10 are dealt with mostly through demonstrations (71%). Conflicts on economic issues are expressed through strikes (36%) and demonstrations (34%). 47% of election related conflicts are demonstrations and 30% are riots. Demonstrations are in general a preferred method of social conflict, composing 45% of all social conflict in SSA. As a means of advocacy or protest based on social networks and
- rganisation, demonstrations may be preferred to riots and strikes as they are largely more peaceful
(McPhail and Wohlstein 1983). Table 1: Main issue of conflict and type of conflict
Human rights/ Democracy Economy/ Jobs/ Assets Domestic war/ Violence/ Terror Elections Ethnic/ Religious Other/ Unknown Total Demonstration 611 249 190 209 79 650 1,988 % 70.96 33.65 45.13 47.07 20.73 41.69 45.12 Riot 132 121 70 132 126 281 862 % 15.33 16.35 16.63 29.73 33.07 18.02 19.56 Strike 51 267 16 15 2 69 420 % 5.92 36.08 3.8 3.38 0.52 4.43 9.53 Other 67 103 145 88 174 559 1,136 % 7.78 13.92 34.44 19.82 45.67 35.86 25.78 Total 861 740 421 444 381 1,559 4,406 % 100 100 100 100 100 100 100
Turning to the key explanatory variables on population composition, the urban population has proportionately more young adults (aged 15-29) than the rural population – 31% compared to 26%. The urban population is distributed by age as seen on the left of Figure 1; on average in 2015, almost 27% of the urban population is under age ten and merely 3.6% over age sixty. 58% of the urban population is of working ages (15-59). In contrast, on the right side of the pyramid in Figure 1 is the age distribution of rural-to-urban migrants in SSA in 2015. The migrant population is less evenly distributed between age groups and has high proportions of ten to twenty year olds – 44.4%. The proportion of migrants in the urban population on average for SSA countries peaks for age groups 10-
SLIDE 11
11 14 for women (25% of the urban population), and 15-19 for men (28% of the urban population), as seen in Figure 2. Migrants are clearly an important feature of the urban YB. Figure 1: Mean Male Urban and Migrant Population Composition for Countries in Sub-Saharan Africa for 2015 Figure 2: Mean Proportion Migrants of Urban Population by Age for Sub-Saharan Africa 2015
SLIDE 12
12 The mean male urban YB is 0.65, that is, 65% of the male adult working population is aged 15-24 over all time periods and countries. The mean male migrant YB is 11%, ranging from -1% to 67%, and the non-migrant YB is 55%, ranging from -1% to 89%. Between 1990 and 2013 the migrant and non- migrant urban YBs have seen declines, reflecting changes in the urban population as it transits to lower fertility. When examining the effect of an urban youth bulge (not seperated into migrants and non-migrants)
- n social conflict in SSA, results indicate that a male YB increases the odds of conflict (Table 2). An
increase in the urban YB significantly raises the odds of urban conflict by a factor of 44, when only accounting for time (“Null Urban” model). When control variables of population size, proportion urban, sex ratio, GDP, polity and unemployment are included in the model, “Urban OLS”, the odds become insignificant. However, in a country fixed effects model, with an increase in the urban male YB the odds of social conflict increase by a factor of 477 within countries. In this model (“Urban FE”) all
- ther factors included in the model are insignificant. In the OLS models when time-invariant effects
are not removed, population characteristics and unemployment significantly affect the probability of
- conflict. An increase in population size raises the odds of conflict; a higher sex ratio, with
proportionately more men also increases the odds of social conflict. With an increase in the proportion of a country’s urban population the chances of conflict decrease by around 80%. This may be explained by urbanisation reflecting economic growth (Black and Henderson 2011; Eaton and Eckstein 1997). Decomposed to migrant and non-migrant youth bulges, results indicate that within countries a non-migrant male urban YB significantly increases the odds of conflict by a factor of 173. The migrant YB has no significant effect on overall social conflict in SSA.4
4 Three other models tested for sensitivity of these results- using a non-lagged youth bulge, a youth bulge
defined as 15-29 year olds and a female youth bulge. None of these models indicated a significant effect of migrant youth bulges on conflict. It is worth noting however, that a female youth bulge reduces the odds of social conflict by 20%.
SLIDE 13
13 Table 2: The Probability of Urban Conflict in 42 Sub-Saharan African Countries (Logistic Models)
Null Urban Urban OLS Urban FE Null Migrant Migrant OLS Migrant FE Urban Male YB 44.338** 2.856 477.150* (32.087) (3.162) (1201.345) Male Migrant YB 2.558 0.157 5.537 (2.486) (0.213) (16.245) Non-Migrant Urban Male YB 20.939** 1.348 172.741* (14.191) (1.332) (382.535) Year 1.032** 1.021 1.060 1.032** 1.024^ 1.062 (0.011) (0.014) (0.038) (0.011) (0.014) (0.041) Logged Urban Population Size 2.143** 0.701 2.127** 0.798 (0.208) (0.731) (0.209) (0.899) Proportion of Population Urban 0.192* 0.009 0.110** 0.002 (0.141) (0.051) (0.085) (0.010) Urban Youth Sex Ratio 1.054** 1.022 1.057** 1.033 (0.015) (0.050) (0.015) (0.053) GDP 0.888 1.231 0.869 1.131 (0.115) (0.336) (0.110) (0.300) Autocracy-Democracy Score 0.977 1.008 0.971^ 1.017 (0.015) (0.033) (0.015) (0.033) Unemployment 1.080** 0.993 1.083** 0.989 (0.020) (0.070) (0.020) (0.070) Constant 0.000** 0.000^ 0.000** 0.000* (0.000) (0.000) (0.000) (0.000)
- No. of cases
1008 873 744 1008 873 744
Odds ratios (Standard errors) Two-tailed test: ** p<0.01; *p<0.05; ^p< 0.1
Examining the probability of urban conflict by main issue facilitates a more refined view on an urban YB effect on conflict. In Table 3, the central model is run according to the main issue behind the
- conflicts. A male migrant YB significantly does not increase or decrease the odds of urban terrorism or
domestic war. Though insignificant, the coefficients suggest that a migrant YB lowers the probability of human rights-democracy, elections and ethnic-religious related conflicts in the urban sector while increases the odds of economic-job-asset and other conflicts. A non-migrant urban male YB increases the odds of all urban conflicts regardless of the main issue behind the conflict (except for terror related conflicts). In particular, the odds of an economic conflict are significantly raised by a factor of 221 with an increase in the non-migrant YB. A multinominal logistic model indicated that when
SLIDE 14
14 controlling for country effects, neither a migrant nor non-migrant YB increased the relative risk of conflict regarding one issue over another (results not shown). Table 3: The Probability of Urban Conflict in 42 Sub-Saharan African Countries, by Main Issue of Conflict (Logistic Models)
Human Rights Economic Terror Elections Ethnic Other Male Migrant YB 0.012 15.396 0.000* 0.011 0.005 10.671 (0.034) (42.205) (0.000) (0.036) (0.026) (27.550) Non-Migrant Urban Male YB 49.425^ 221.385* 0.005 2.135 143.473 43.178^ (116.529) (509.590) (0.018) (5.887) (735.352) (84.558) Year 0.977 1.054 1.119* 0.934 1.059 1.023 (0.037) (0.040) (0.050) (0.041) (0.070) (0.034) Logged Urban Population Size 14.266* 0.497 0.191 2.803 6.558 1.352 (15.767) (0.578) (0.257) (3.562) (11.990) (1.328) Proportion of Population Urban 0.000** 0.000^ 18205.080 31.876 0.000* 1.522 (0.000) (0.000) (176722.712) (218.196) (0.000) (8.468) Urban Youth Sex Ratio 1.109^ 1.078 0.999 1.085 0.951 1.025 (0.061) (0.055) (0.097) (0.076) (0.109) (0.046) GDP 1.912* 1.293 0.517^ 1.092 0.657 1.036 (0.553) (0.339) (0.206) (0.342) (0.336) (0.262) Autocracy-Democracy Score 1.001 1.055^ 1.014 1.105** 0.941 0.972 (0.029) (0.030) (0.045) (0.038) (0.050) (0.028) Unemployment 1.000 1.019 0.728* 0.933 1.069 0.949 (0.074) (0.074) (0.117) (0.075) (0.146) (0.068)
- No. of cases
836 840 597 785 440 873
Odds ratios (Standard errors) Two-tailed test: ** p<0.01; *p<0.05; ^p< 0.1
Tables 2 and 3 have shown that male migrant YBs do not affect the probability of conflict while the non-migrant YB tends to increase the odds of urban social conflict. In Table 4 I examine whether the YBs increase the number of conflicts in SSA. In the OLS model an increase in the size of the migrant YB by one unit would lower the rate ratio of urban conflict by a factor of 0.026. This effect disappears in the fixed effects model which accounts for between-country heterogeneity. The number of conflicts within countries on election and ethnic-religious issues is reduced by over 99% with a unit increase in the size of the migrant YB. An increase in the non-migrant male YB significantly increases the
SLIDE 15
15 frequency of all conflicts within countries – in particular on issues of human rights-democracy, economic-job-assets, elections and ethnic-religious. According to the “All FE” model, as time progresses the incident rate is expected to change by a factor
- f 0.98, suggesting a decline in the number of conflicts by 2% every year. In addition, the population
variables are all significant in accounting for the number of social conflicts: an increase in urban population size increases the rate ratio of conflicts by a factor of 4.6; an increase in the proportion of the population urban lowers the rate ratio of conflict by a factor of 0.008; and a unit increase in the urban youth sex ratio suggests an increase of nearly 10% in conflicts. A positive relationship is also found between GDP and the frequency of conflict. Table 4: Number of Conflicts in 42 SSA Countries, by Main Issue of Conflict (Poisson Models)
All OLS All FE Human Rights FE Economic FE Terror FE Elections FE Ethnic FE Other FE Male Migrant YB 0.026** 0.393 0.092 1.051 0.046 0.014^ 0.005^ 0.915 (0.008) (0.254) (0.135) (1.581) (0.116) (0.034) (0.015) (0.979) Non-Migrant Urban YB 0.042** 7.635** 22.079* 9.564^ 0.029^ 97.536* 2.40e+07** 2.346 (0.009) (4.406) (28.826) (12.093) (0.057) (189.419) (86114512.7) (2.325) Year 1.022** 0.983* 0.953** 1.028 1.067** 0.924** 0.867** 0.988 (0.003) (0.008) (0.017) (0.019) (0.025) (0.024) (0.030) (0.013) Logged Urban Population Size 1.937** 4.577** 33.106** 0.665 0.556 12.107** 254.549** 4.005** (0.028) (1.116) (18.464) (0.417) (0.417) (9.710) (296.853) (1.660) Proportion of Population Urban 0.291** 0.008** 0.000** 0.010 2.48e+05* 0.276 0.000** 0.514 (0.058) (0.012) (0.000) (0.034) (1256815.6) (1.203) (0.000) (1.298) Urban Youth Sex Ratio 1.029** 1.099** 1.110** 1.107** 1.065 1.193** 1.059 1.080** (0.003) (0.016) (0.033) (0.033) (0.066) (0.059) (0.099) (0.028) GDP 0.881** 1.162** 1.508** 1.314* 0.713^ 0.865 1.850* 1.036 (0.027) (0.065) (0.206) (0.170) (0.141) (0.154) (0.522) (0.098) Autocracy-Democracy Score 0.993^ 1.001 0.943** 1.033* 0.953* 1.092** 1.016 1.002 (0.004) (0.006) (0.012) (0.014) (0.020) (0.023) (0.026) (0.011) Unemployment 1.009* 0.978 1.008 0.978 0.934 0.946 1.053 0.966 (0.003) (0.015) (0.040) (0.032) (0.063) (0.043) (0.065) (0.025) Constant 0.000** (0.000)
- No. of cases
873 873 850 854 597 785 440 873
Incidence rate ratios (Standard errors) Two-tailed test: ** p<0.01; *p<0.05; ^p< 0.1
SLIDE 16
16 DISCUSSION Classic sociologists regularly describe urban life as fundamentally different to life in rural areas. Amongst them, Wirth (1938) coherently defined urban life. He suggested that three key features of the urban sector may lead to violence – population size, density and heterogeneity. These features are associated with specialisation, utilitarian interpersonal relationships, increased competition and social stratification which can all promote unrest and violence. Rural-to-urban migrants play a significant role in increasing population size, density and heterogeneity. They are an important component of urbanisation to contend with when considering urban conflict. Results have shown that migrants are a particularly large proportion of young adults in the urban sector yet multivariate analysis suggests that
- verall in SSA a migrant YB is not a contributing factor to social conflict. While a migrant YB doesn’t
increase the probability of urban violence, it may increase the number of conflicts, particularly ethnic- religious and election related conflicts within countries. My first hypothesis is essentially confirmed – migrant youth bulges are insignificant in explaining economic related social conflicts. However, the coefficients are positive suggesting if any effect exists, the migrant YB would increase the frequency and probability of economic conflicts within countries. In light of the urbanisation of poverty, migrants may experience material deprivation and inequalities forming a source of insecurity (Fox and Beall 2012). When faced with chronic poverty and difficulties in the job market, the cost of engaging in violence may outweigh individual costs (Collier and Hoeffler 2004). Non-migrant YBs on the other hand are significant in increasing economic conflicts. Non- migrants tend to engage in the formal job sector and thus participate in strikes. My second hypothesis is partially confirmed – migrant youth bulges do not significantly increase the
- dds of ethnic-religious conflict but they do increase the frequency of such conflict. Migrants
contribute to heterogeneity of the urban population. The ethnic heterogeneity of a population in itself doesn’t necessarily lead to violence. It is the identification with the group, strengthened by between
SLIDE 17
17 group inequalities - a combination of political, economic and cultural inequalities - that increases the number of conflicts (Higashijima and Houle 2017; Stewart 2010). Results indicate that my third hypothesis is false. A migrant YB does not increase the likelihood and frequency of conflict related to human rights, democracy or elections. Though mostly insignificant, a migrant YB admittedly lowers the probability and frequency of such conflict. Unlike economic-related conflict, the cost of participation in conflicts on human rights may not surpass individual grievances. Finally, my fourth hypothesis is also falsified, a migrant YBs decreases the frequency of conflicts on issues of domestic war and terror. Migrant YBs significantly do not raise or lower the likelihood of such
- conflict. Domestic war and terror related conflicts likely affect the entire population equally.
CONCLUSION The aim of this article is to disaggregate the urban youth bulge by distinguishing between natural growth and migration. The study tests whether an increase in the relative proportion of young rural- to-urban migrants increases the probability and frequency of social conflict by the main issue behind the conflicts. In general, models failed to support the hypotheses. Although migrants do increase the size, density and heterogeneity of the urban population and face relative deprivation and marginalisation (Cincotta et al. 2003; Cook 2015; Fearon and Laitin 2011; Gizewski and Homer-Dixon 1995), two related factors may also be at play, compounding a relationship between a migrant YB and conflict. One, economic growth is sustained by urbanisation (Black and Henderson 2011; Eaton and Eckstein 1997), so rural-to-urban migration may positively reflect economic growth in cities. More young migrants to the urban sector suggests better opportunities in education and the job market. Better economic prospects may pull migrants to the city and allow them to integrate into the fabric of urban society relatively smoothly. Two, corresponding to migration reflecting economic growth, urban areas provide better opportunities for young adults. Despite increasing prevalence of urban poverty and
SLIDE 18
18 slums in Africa, urban populations are still better off than their rural counterparts (Awumbila et al. 2014; Oucho 2014). The urban setting provides more economic opportunities, better education (Sahn and Stifel 2003) and better child health outcomes (Fink et al. 2014). For youth, cities also provide anonymity, a resource for delaying adulthood expectations and reinvention, a space of possibility (Sommers 2010). Young rural-to-urban migrants can relatively improve their position. Hence, any grievances they may have are negligible to what has been gained by moving to the city. This has also been suggested by Buhaug and Urdal (2013) who found that population growth in cities may even lower urban disorder. A number of challenges were encountered in this study. Firstly, while CSRM is a decent method for measuring internal migration in SSA, these measures may be biased by international migration and
- reclassification. All the same, considering the lack of other comprehensive sources of internal
migration data, this paper has managed to distinguish whether the population pressure of young adults in urban settings is associated with migration or natural city growth. Additionally, although migration may be related to relative deprivation, economic marginalisation and greater ethnic heterogeneity, these factors lack disaggregated rural/urban quantitative data to include in analysis. Further research is needed with improved data on youth exclusion to examine the underpinnings of an urban YB, and in particular of a migrant YB. It is also important to expand the scope of future research by examining the effect of a migrant YB on urban homicide rates and organised crime on the one hand and civil war on the other. This study represents a first exploration of the impact of migration on urban social conflict in Sub- Saharan Africa. The relationship uncovered suggests that governments should seek to encourage positive between-group relations and reduce inequalities along ethnic and religious divides as a means
- f lowering the frequency of urban conflict. In addition, curbing urban unemployment, creating
- pportunities for migrants and non-migrants alike as well as generally encouraging economic growth
in cities may lead to fewer cases of social unrest.
SLIDE 19
19 REFERENCES Awumbila, M., Owusu, G., & Teye, J. K. (2014). Can Rural-Urban Migration into Slums Reduce Poverty ? Evidence from Ghana (No. 13). Migrating Out of Poverty. Bhavnani, R., & Miodownik, D. (2008). Ethnic Polarization, Ethnic Salience, and Civil War. Journal of Conflict Resolution, 53(1), 30–49. doi:10.1177/0022002708325945 Black, D., & Henderson, V. (2011). A Theory of Urban Growth. Journal of Political Economy, 107(2), 252–284. doi:10.1086/250060 Bocquier, P., & Maupeu, H. (2006). Analysing Low Intensity Conflict in Africa Using Press Reports. In H. Brunborg, E. Tabeau, & H. Urdal (Eds.), The Demography of Armed Conflict (internatio., pp. 279– 302). Springer. Buhaug, H., & Gates, S. T. (2002). The Geography of Civil War *. Journal of Peace Research, 39(4), 417– 433. Buhaug, H., & Rød, J. K. (2006). Local determinants of African civil wars, 1970–2001. Political Geography, 25(3), 315–335. doi:10.1016/j.polgeo.2006.02.005 Buhaug, H., & Urdal, H. (2013). An urbanization bomb? Population growth and social disorder in cities. Global Environmental Change, 23(1), 1–10. doi:10.1016/j.gloenvcha.2012.10.016 Centre for Systemic Peace. (2015). Polity IV Annual Time Series, 1800-2015. http://www.systemicpeace.org/inscrdata.html Cincotta, R., Engelman, R., & Anastasion, D. (2003). The security demographic: Population and civil conflict after the Cold War. Washington D.C. Collier, P., & Hoeffler, A. (2004). Greed and grievance in civil war. Oxford Economic Papers, 56, 563–
- 595. doi:10.1111/j.1468-2346.2012.01100.x
Cook, S. (2015). Rural-urban migration and social exclusion among Cambodian youth: Discourses and narratives from Phnom Penh. Utrecht. Cramer, C. (2011). Unemployment and Participation in Violence. de Brauw, A., Mueller, V., & Lee, H. L. (2014). The role of rural-urban migration in the structural transformation of Sub-Saharan Africa. World Development, 63, 33–42. doi:10.1016/j.worlddev.2013.10.013 den Boer, A., & Hudson, V. M. (2004). The Security Threat of Asia’s Sex Ratios. SAIS Review, 24(2), 27–
- 43. doi:10.1353/sais.2004.0028
Dyson, T. (2012). Causes and Consequences of Skewed Sex Ratios. Annual Review of Sociology, 38(1), 443–461. doi:10.1146/annurev-soc-071811-145429 Eaton, J., & Eckstein, Z. (1997). Cities and growth: Theory and evidence from France and Japan. Regional Science and Urban Economics, 27(4–5), 443–474. doi:10.1016/S0166-0462(97)80005-1 Esteban, J., Mayoral, L., & Ray, D. (2012). Ethnicity and conflict: An empirical study. American Economic Review, 102(4), 1310–1342. doi:10.1257/aer.102.4.1310 Fearon, J. D., & Laitin, D. D. (2003). Ethnicity, Insurgency, and Civil War. American Political Science
SLIDE 20
20 Review, 97(1), 75–90. doi:10.1017/CBO9781107415324.004 Fearon, J. D., & Laitin, D. D. (2011). Sons of the Soil, Migrants, and Civil War. World Development, 39(2), 199–211. doi:10.1016/j.worlddev.2009.11.031 Fink, G., Günther, I., & Hill, K. (2014). Slum Residence and Child Health in Developing Countries. Demography, 51(4), 1175–1197. doi:10.1007/s13524-014-0302-0 Fox, S., & Beall, J. (2012). Mitigating conflict and violence in African cities. Environment and Planning C: Government and Policy, 30(6), 968–981. doi:10.1068/c11333j Gizewski, P., & Homer-Dixon, T. (1995). Urban gowth and violence: Will the futute resemble the past? Washington D.C. Gleditsch, N. P., & Urdal, H. (2002). Ecoviolence? Links between population growth, environmental scarcity and violent conflict in Thomas Homer-Dixon’s work. Journal of International Affairs, 56(1), 283. doi:10.1017/CBO9781107415324.004 Goldstone, J. A. (2002). Population and Security: How Demographic Change Can Lead to Violent
- Conflict. Journal of International Affairs, 56(1), 3–22.
Hamilton, C. . H., & Henderson, F. M. (1944). Use of the Survival Rate Method in Measuring Net
- Migration. Journal of the American Statistical Association, 39(226), 197–206.
Hannerz, H. (2001). Manhood trails and the law of mortality. Demographic Research, 4(7), 185–202. doi:10.4054/DemRes.2001.4.7 Hegre, H., Ellingsen, T., Gates, S., & Gleditsch, N. P. (2001). Toward a democratic civil peace? Democracy, political change, and civil war, 1816–1992. American Political Science Review, 95(1), 33–48. doi:10.1017/CBO9781107415324.004 Higashijima, M., & Houle, C. (2017). Ethnic Inequality and the Strength of Ethnic Identities in Sub- Saharan Africa. Political Behavior, 1–24. Hvistendahl, M. (2011). Young and restless can be a volatile mix. Science (New York, N.Y.), 333(6042), 552–4. doi:10.1126/science.333.6042.552 International Monetary Fund. (2014). World Economic Outlook Database. https://www.imf.org/external/pubs/ft/weo/2014/02/weodata/index.aspx Keyfitz, N. (1980). Do Cities Grow by Natural Increase or by Migration? Geographical Analysis, 12(2), 142–156. Macunovich, D. J. (2000). Relative cohort size: Source of a unifying theory of global fertility transition. Population and development review, 26(2), 235–61. McPhail, C., & Wohlstein, R. T. (1983). Individual and Collective Behaviors Within Gatherings, Demonstrations, and Riots. Annual Review of Sociology, 9(1), 579–600. doi:10.1146/annurev.so.09.080183.003051 Menashe-Oren, A., & Stecklov, G. (n.d.). Rural-Urban Population Age and Sex Composition in Sub- Saharan Africa. Population and development review, 1–39. Mesquida, C. G., & Wiener, N. I. (1999). Male Age Composition and Severity of Conflicts. Politics and the Life Sciences, 18(2), 181–189.
SLIDE 21
21 Moultrie, T., Dorrington, R., Hill, A., Hill, K., Timaeus, I. M., & Zaba, B. (2013). Tools for Demographic
- Estimation. (T. Moultrie, R. Dorrington, A. Hill, K. Hill, I. M. Timaeus, & B. Zaba, Eds.). Paris:
International Union for the Scientific Study of Population. Oucho, J. O. (2014). Changing perspectives of internal migration in Eastern Africa. Pinard, M. (2011). Motivational Dimensions in Social Movements and Contentious Collective Action. McGill-Queen’s University Press. Preston, S. H. (1979). Urban Growth in Developing Countries : A Demographic Reappraisal. Population and Development Review, 5(2), 195–215. Rogers, A. (1982). Sources of Urban Population Growth and Urbanization, 1950-2000: A Demographic
- Accounting. Economic Development and Cultural Change, 30(3), 483–506.
Sahn, D. E., & Stifel, D. C. (2002). Urban-Rural Inequality in Africa. Cornell University. Sahn, D. E., & Stifel, D. C. (2003). Urban-Rural Inequality in Living Standards in Africa. Journal of African Economies, 12, 564–597. doi:10.1093/jae/12.4.564 Salehyan, I., Hendrix, C. S., Hamner, J., Case, C., Linebarger, C., Stull, E., & Williams, J. (2012). Social Conflict in Africa: A New Database. International Interactions, 38(4), 503–511. doi:10.1080/03050629.2012.697426 Sommers, M. (2010). Urban youth in Africa. Environment and Urbanization, 22(2), 317–332. doi:10.1177/0956247810377964 Staveteig, S. (2005). The Young and the Restless: Population Age Structure and Civil War. ECSP Report, (11), 12–19. Stewart, F. (2010). Horizonal Inequalities as a Cause for Conflict: A Review of CRISE Findings. World Development Report 2011 Background Paper, (January), 1–9. United Nations. (2014a). Urban and Rural Population by Age and Sex, 1980-2015. Department of Economic and Social Affairs, Population Division. http://www.un.org/en/development/desa/population/publications/dataset/urban/urbanAndRur alPopulationByAgeAndSex.shtml United Nations. (2014b). Methodological note: Estimates of the urban and rural population by age and sex, 1980-2015. Urdal, H. (2008). Population, Resources, and Political Violence A Subnational Study of India, 1956–
- 2002. Journal of Conflict Resolution, 52(4), 590–617.
Urdal, H., & Hoelscher, K. (2009). Urban Youth Bulges and Social Disorder An Empirical Study of Asian and Sub-Saharan African Cities (No. 5110). World Bank Policy Research Working Paper. Weber, H. (2013). Demography and democracy: the impact of youth cohort size on democratic stability in the world. Democratization, 20(2), 335–357. doi:10.1080/13510347.2011.650916 Wilson, M., & Daly, M. (1985). Competitiveness, risk taking, and violence: the young male syndrome. Ethology and Sociobiology, 6(1), 59–73. doi:10.1016/0162-3095(85)90041-X Wirth, L. (1938). Urbanism as a Way of Life. American Journal of Sociology, 44(1), 1–24. World Bank. (2016). World Development Indicators. http://data.worldbank.org/data-catalog/world-
SLIDE 22