Dispensa- 8 -Civil Wars Flashcards
Similarities and differences of inter- and intra-states conflicts
The realist unitary actor assumption applies less well to groups than states:
- rebel groups do not have the resources of a state
- groups may not be well defined (ethnicity, religion, ideological positions overlay).
The liberal theory on positive effects of democracy and capitalism do not apply at the group level.
Bargaining issues for civil conflicts
- Commitment problems: lack of information on who effectively controls the state.
- Asymmetric information: complex to know about rebel groups.
- Ideology and religion matter more for ruling a state than for inter-state relations.
- Multilateral interests are normal in multi-ethnic and fragmented societies.
Facts about Civil Wars
- 1/3 of all countries had a civil war in the last 60 years.
- 1/3 of African countries had a civil war in the 90’s.
Negative Shock Argument
Peace can only sustain if
A’s expected utility from peace >= A’s expected utility from winning
(p - m)θt + c ≥ δ[mV^V - V^P]
if square bracket is positive, bad shock leads to war
Situations increasing the likelihood of a secessionist civil war
- One minority or out of government group is concentrated in a geographical area
- That area has natural resources whose revenues are claimed by the group
- Resource pointiness (diamonds)
Situation increasing the likelihood of a centrist civil war
The majority is not sharing much the resources outside that area with the minority.
Relative strength vs. relative wealth
- on the one hand, pushing power sharing towards making it reflective of the relative strengths (strength proportionality) eliminates the incentives of centrist wars
• secessionist wars could then materialise as the previously-wealthy group might not want to share its riches with the other faction - on the other hand, making power sharing depend on the groups’ relative resource endowments (resource proportionality) eliminates the incentives of secessionist wars
• centrist wars could then materialise as the majority group has an incentive to use its strength to gain more power
conflict is overall less likely when
Resources, groups or both are either absent ore evenly concentrated
Bargaining failure is more likely if
- Winning chances of the minority are much higher for secessionist than for centrist conflict
- Most of the natural resources are located in the region of the minority group
- Minority group is highly concentrated in a relatively homogeneous region
- Poor country and/or war is not very destructive
- Low state capacity (low lambda)
secession over centrist if
- pop is over represented in the region
- natural advantage is better for the region than for the whole country
civil war or conflict definitions
war: violence between organised groups+1000 deaths in 1y
conflict: violence+25 deaths in 1y
model for correlation between low GDP per capita and probability of civil war
Model for causal link between negative shock in wealth and likelihood of civil conflict
- p - share of the national economic wealth for the group A
- 𝜃t - the value of the entire wealth at time t
- m - A’s relative power
- Vv - the continuation value if A prevails in the war
- Vp - the value from the most peaceful continuation equilibrium
- 𝛿 - the discount rate
Peace can only sustain if:
As expected utility from peace outweighs As expected utility from war.
p𝜃t+𝛿Vp> m(𝜃t+𝛿Vv) -c
The inequality above can be rewritten as:
(p-m)𝜃t +c > 𝛿 [mVv- Vp ]
the opportunity cost of not going to war is superior to the opportunity cost of going to war
whenever the square bracket is positive, there may exist a sufficiently bad shock to making the inequality false, and so, leading to war.
if we assume that long-run values are unaffected by shocks to 𝜃, there might be a decrease in national wealth severe enough, to make a group dissatisfied enough with its current economic share, to use violence to get the control of the whole pie
Model for civil wars and resources
basics
In a state, 2 regions, 2 groups.
population sizes Ni1, Nj1, Ni2 and Nj2
resources R1 R2
group i gets alpha resources, chosen by j
Lambda represents the accumulated advantage of the group in power, lambda’ the accumulated advantage of the group in power in region 1 only
Group i can then choose between accepting that arrangement peacefully (action p), start a centrist war (action c), or request a secession (action s)
Group j can also trigger a centrist war (in other words, a preventive repression of the minority group), or wait for i’s choice and react to it.
Model for civil wars and resources
secession or centrist conflict probabilities
payoffs
Assuming that a faction’s military power is roughly proportional to its population size, in case of a centrist conflict, i wins with probability:
pc= Ni / (Ni +Nj + lambda)
in case of a secessionist conflict, i wins with probability
ps= Ni1/ (Ni1 +N j 1 +lambda’)
The minority group has a higher probability of winning a secessionist conflict than a centrist conflict if
Ni1/N1 > Ni/N
and/or if lambda> lambda
the peaceful status quo gives i a payoff alpha(R1 + R2)
i’s expected utility from a centrist conflict is (Ni/(N+lambda) )* (R1 + R2) - dci,
dci is the destruction cost of a centrist war for group i
i’s expected utility from a secession is (N1i/(N1+lambda’))* (R1) - dsi, with dsi < dci
additionally,
dsj < (N1j + lambda’/ N1 +lambda)*R1
Model for civil wars and resources
Peace or not, wrt alpha
On the one hand, group i prefers the status quo to a secessionist conflict if and only if
alpha>=alpha s
alpha s =(N1i)/(N1+lambda’)*(R1/R) -dsi/R
On the other hand, j prefers the status quo as long as
alpha<=alpha barre
alpha barre= (Ni)/(N+lambda)+dsj/R
so peace only if dcj + dsi < N1i/(N1 + λ′)R1 − Ni/(N + λ)R
then there is no value of α that can induce both groups to choose peace, because α s > α barre.
