Discrimination and ethnic conflict: a dyadic analysis of politically-excluded groups in sub-Saharan Africa

Figures & data

Table 1. Descriptive statistics summary table.

Variables	Mean/Odds	Std. Deviation	Min	Max
Conflict	0.04	0.19	0	1
Political configurations of dyads
Excluded/in power	0.23	0.42	0	1
Discriminated/in power	0.08	0.27	0	1
Not discriminated/in power	0.15	0.36	0	1
Both excluded	0.07	0.25	0	1
Both discriminated	0.01	0.11	0	1
Both not discriminated	0.03	0.18	0	1
Discriminated/not discriminated	0.02	0.15	0	1
Both in power	0.70	0.46	0	1
Control factors
Education inequality	1.64	1.69	0	8.90
Average education	5.00	2.56	0.50	11.90
Wealth inequality	8.81	10.54	0	72.90
Average wealth	25.38	13.56	2.15	93.40
Population inequality (%)	10.46	11.92	0.01	86.47
Average population (%)	10.49	7.62	1.00*	43.87
Elite share inequality (%)	5.14	6.80	0	64.70
Average elite share (%)	9.60	7.50	0	52.35
Difference in urbanization (%)	16.94	15.60	0	88.70
Average urbanization (%)	34.09	16.44	3.20	93.15
Conflict occurred 1 year prior (%)	0.04	0.19	0	1
Conflict occurred 2 years prior (%)	0.02	0.13	0	1
Conflict occurred 3 years prior (%)	0.01	0.11	0	1
Conflict occurred 4 years prior (%)	0.01	0.10	0	1
Rule of law (WGI)	−0.64	0.54	−2.13	0.35

*Minimum cutoff value is set at 1%.

Table 2. Multilevel logistic regression coefficients for bivariate associations between selected independent variables and ethnic conflict incidence in sub-Saharan African countries 1990–2013.

Variables	Bivariate Coefficients
Political configurations of dyads
Excluded/in power	2.30*** (0.41)
Discriminated/in power	3.50*** (0.82)
Not discriminated/in power	1.10 (0.23)
Both excluded	0.73 (0.20)
Both discriminated	1.07 (0.53)
Both not discriminated	0.79 (0.33)
Discriminated/not discriminated	0.57 (0.25)
Both in power	0.45*** (0.09)
Control factors
Education inequality	1.17* (0.07)
Average education	0.83* (0.06)
Wealth inequality	0.99 (0.01)
Average wealth	0.98 (0.01)
Population inequality	0.99 (0.01)
Average population	1.03 (0.02)
Elite share inequality	0.99 (0.02)
Average elite share	1.02 (0.02)
Urbanization inequality	0.99 (0.01)
Average urbanization	1.00 (0.01)
Rule of law	0.14*** (0.05)
Conflict occurred 1 year prior	28.01*** (6.35)
Conflict occurred 2 years prior	6.49*** (1.83)
Conflict occurred 3 years prior	4.37*** (1.42)
Conflict occurred 4 years prior	3.20** (1.19)
Observations	9,783
Conflict-year combinations	375

Coefficients are the log odds for the variables in the models and estimates for random-effects parameters, with standard errors in parentheses. *p < .05; **p ≤ .01; ***p ≤ .001. Models were estimated for each variable separately, whereby both components of the socioeconomic control variables (inequality and mean) were estimated simultaneously.

Table 3. Multilevel logistic regression coefficients for multivariate associations between selected independent variables and ethnic conflict incidence in sub-Saharan African countries, 1990–2013.

Variables	1	2	3	4
Constant	0.00*** (0.00)	0.00*** (0.00)	0.00*** (0.00)	0.00*** (0.00)
Political configurations of dyads
Excluded/in power	Reference
Discriminated/in power		Reference	3.22*** (0.83)	3.13*** (0.86)
Not discriminated/in power		0.40** (0.12)	Reference	Reference
Both excluded	0.56 (0.17)
Both discriminated		0.55 (0.32)	Reference	Reference
Both not discriminated		0.31* (0.15)	Reference	Reference
Discriminated/not discriminated		0.23** (0.12)	Reference	Reference
Both in power	0.45*** (0.09)	0.27*** (0.07)	Reference	Reference
Control factors
Education inequality	1.36*** (0.13)	1.37*** (0.13)	1.36** (0.13)	1.30** (0.12)
Average education	0.83 (0.09)	0.83 (0.09)	0.83 (0.09)	0.85 (0.10)
Wealth inequality	0.96 (0.02)	0.96 (0.02)	0.96 (0.02)	0.98 (0.02)
Average wealth	0.99 (0.02)	1.00 (0.02)	1.00 (0.02)	1.00 (0.02)
Population inequality	0.99 (0.01)	0.99 (0.02)	0.99 (0.02)	0.99 (0.02)
Average population	1.06* (0.03)	1.06* (0.03)	1.06* (0.03)	1.06* (0.03)
Elite share inequality	0.96 (0.03)	0.95 (0.03)	0.96 (0.03)	0.95 (0.03)
Average elite share	1.07** (0.02)	1.08*** (0.02)	1.07** (0.02)	1.06* (0.02)
Urbanization inequality	1.00 (0.01)	1.00 (0.01)	1.00 (0.01)	1.00 (0.01)
Average urbanization	1.01 (0.01)	1.01 (0.01)	1.01 (0.01)	1.01 (0.01)
Conflict occurred 1 year prior	14.53*** (3.46)	13.36*** (3.18)	13.15*** (3.13)	12.31*** (2.86)
Conflict occurred 2 years prior	3.94*** (1.15)	3.65*** (1.07)	3.62*** (1.05)	3.47*** (1.00)
Conflict occurred 3 years prior	2.88** (0.98)	2.72** (0.93)	2.67** (0.91)	2.53** (0.86)
Conflict occurred 4 years prior	2.30* (0.88)	2.24* (0.86)	2.22* (0.85)	2.11* (0.81)
Rule of law	0.16*** (0.06)	0.18*** (0.07)	0.17*** (0.07)	0.14*** (0.06)
Interactions
Education inequality * discriminated/in power				0.89 (0.13)
Wealth inequality * discriminated/in power				0.93* (0.03)
Population inequality * discriminated/in power				0.99 (0.01)
Average elite share * discriminated/in power				1.10** (0.04)
Random-intercept parameters
Country level	1.51*** (0.36)	1.43*** (0.36)	1.45*** (0.36)	1.38*** (0.36)
Ethnic dyad level	1.18*** (0.20)	1.26*** (0.21)	1.26*** (0.21)	1.27*** (0.20)
Omnibus test Chi^2 (DF)	78.82 (20)	74.07 (23)	74.05 (19)	76.24 (23)
Observations	9,783	9,783	9,783	9,783
Conflict-year combinations	375	375	375	375

Coefficients are the log odds for the variables in the models and estimates for random-effects parameters, with standard errors in parentheses. *p < .05; **p ≤ .01; ***p ≤ .001.

Figure 1. Interaction effects of wealth inequality and average elite share on the risk of conflict between discriminated and in-power groups (based on standardized logit coefficients).