Week 05: Civil Wars

# Week 05: Civil Wars
## Ending Civil Wars
<html>
<div style="float:left">

</div>
<hr color='#EB811B' size=1px width=800px>
</html>
### Danilo Freire
### 08 October 2019

---

<style>

.remark-slide-number {
  position: inherit;
}

.remark-slide-number .progress-bar-container {
  position: absolute;
  bottom: 0;
  height: 6px;
  display: block;
  left: 0;
  right: 0;
}

.remark-slide-number .progress-bar {
  height: 100%;
  background-color: #EB811B;
}

.orange {
  color: #EB811B;
}
</style>

# Last week we saw that...

* Violence prevents civilian defector and elicits information about enemies

* Violence is unlikely in areas fully controled by the rebels/gov't forces

* In contested areas, it's hard for groups to gather information about the enemies and defection is easy

* Thus, _selective violence against civilians are more prevalent in areas over which the groups have dominant but incomplete control_
]
---

# Last week we saw that...

.font150[
* Humphreys and Weinstein highlight another factor that explains violence against civilians in internal wars: _rebel group structure_

* Less cohesive groups tend to commit more violence: less control over rogue soldiers

* They also find that material incentives also increase violence in civil wars

* Note: no distinction between selective and indiscriminate violence here
]

---

# Last week we saw that...

.font130[
* Balcells focuses her attention on *conventional civil wars*, where insurgents and the government have similar resources

* She affirms that Kalyvas and H&W focused too much on *irregular wars*, or guerrillas

* While guerrillas fight for territorial control, in conventional civil wars rebel groups control well-delineated areas

* Violence occurs due to rivalry and revenge:
	- Rivalry: When parties have parity, small victories can be decisive, so groups are more likely to fight 
	- Revenge: Higher levels of violence by an armed group at `\(t_1\)` leads to retaliation by the rival group in `\(t_2\)`
]

---

# Last week we saw that...

* Ethnic rivalries, greed, and gender norms are conventional explanations

* Rape is a socialisation tool: increases rebel group cohesion and boosts rebels' self-esteem

* Cohen finds that rape is associated with state collapse and rebel group characteristics 
]
---

# How do civil wars end?

<html><div style='float:left'></div><hr color='#EB811B' size=1px width=720px></html> 
---

# Why do some civil wars last so long?

* Military victory does happen in civil war, but it is rare
]
--
.font150[
* **Question**: why do you think that's the case?]
--
.font150[
* Wars last longer when parties perceive the war as less costly relative to peace

* And when combatants are overly optimistic about winning the war
]
---

# Why do some civil wars last so long?

* Government or rebels are unable to trust the commitments

* More combatants increase coordination costs]
---

# Fearon (2002)

# Washington Post, 2013

# Washington Post, 2013

# Walter (1997)

# Negotiated settlements are rare

.font130[
* While most international wars end in agreements, negotiations are unfrequent in civil wars

* "Most internal wars ended with the extermination, expulsion, or capitulation of the losing side"

* Groups almost always choose to fight instead of pursuing an agreements

* Puzzle: if civil wars are costly for both parties, why do conflicts often last so long? ]
--
.font130[
* Negotiations fail because civil war opponents are asked to do what they consider unthinkable: *lay down their weapons and integrate into the national army*
]
---

# Why don't negotiations work?

* But peace talks often fail

* Problem: lack of neutral police force and government enforcement
]
---

# Cooperation under anarchy

.font150[
* Countries can use international organisations, military alliances, or trade agreements as incentives

* Domestic factions don't have the same options

* Settlements make parties vulnerable and compliance is risky

* No credible democratic institutions
]
---

# Third-party guarantees

* Three conditions:
	- Guarantor must have an interest in peace: colonial ties, economic interests, etc
	- Must be willing to use force if needed
	- Must be able to signal resolve: promises have to be credible
]
---

# Five hypotheses

.font140[
* H1: Credible guarantees on the terms of the settlement are almost impossible to arrange by the combatants themselves
	- Need for a third party

* H2: Successful settlements should be more likely as the expected costs of winning a war increases
  - Losers have an incentive to start peace talks

* H3: Adversaries are more likely to settle when stakes are easier to divide
	- Financial arragements, political participation
]
---

# Five hypotheses

.font150[
* H4: The greater the skill of the mediation, the greater the likelihood of a peace deal
	- Credible commitments

* H5: Ethnic/cultural conflicts are harder to solve
	- Groups cannot find common ground

* **Question**: how is H5 different from H3?
]
---

# Results

# Results

# Discussion

* Longer, larger civil wars are more likely to be mediated

* Military stalemates increase the chance of successful negotiations

* Issue divisibility and ethnic motivations are unrelated with the outcomes (decisive victory or negotiated settlements)
]
---

# Conclusion

* A third party is necessary because groups are unable to compromise

* The findings hold even for ethnic civil wars

* Implications:
	- Peacekeeping operations should be willing to use force
	- Mediators might even be biased, as long as they show resolve
	- Guarantors should stay during the whole process
]
---

# Conclusion

.font150[
* Remember: Walter stresses that security guarantees are a necessary, *but not sufficient*, condition for peace

* **Question**: Does her theory sound somewhat familiar to you? 
]
---

# Leviathan

# Questions?

<html><div style='float:left'></div><hr color='#EB811B' size=1px width=720px></html> 
---

# Fortna (2004)

# Do peacekeeping operations work?

* Peacekeeping operations provide third-party law and police enforcement

* In a sense, Fortna puts Walter's theory to test (at least partially)
]
---

# Do peacekeeping operations work?

* Fortna accuses them of two biases:
	- Pessimism bias
	- Selection bias

* **Question**: what does selection bias mean in this case?
]
---

# Do peacekeeping operations work?

.font150[
* Peacekeeping decisions *are not random*
	- Forces are deployed only to the very worst conflicts
	- E.g.: conflicts with more than two parties or when victory is unlikely

* To have an accurate assessment of the effect of peacekeeping operations, we should compare similar cases only 
]
---

# Where do peacekeepers go?

# What is the effect of peacekeeping?

# What is the effect of peacekeeping?

# Types of peacekeeping missions

# Discussion

* ... it has a negative effect on civil war duration after 1991

* Evidence that peacekeeping operations are getting more effective

* All types of peacekeeping operations seem to reduce civil war duration 
]
--
.font150[
* *Yes, peacekeeping does keep peace*
]
---

# Questions?

<html><div style='float:left'></div><hr color='#EB811B' size=1px width=720px></html> 
---

# Howard and Stark (2018)

# Howard and Stark (2018)

* The trend seems to have reversed

* Post-9/11: stabilisation over democratisation and non-negotiation with terrorist groups

* Neorealist explanation: the nature of the international system explains civil war duration 
]
---

# Unipolarity

* In the post-Cold War years, the US doesn't not need to intervene so strongly
	- No other superpower to compete for influence

* Colour Revolutions, Arab Spring have also shown that democratisation is not always feasible

* Solution: stabilisation instead of regime change
]
---

# Three periods

# Three periods

# Text analysis

---

# New norms

.font150[
* Since 9/11 the US and other powers applied new norms of anti-terrorism in civil wars where Islamists were involved
	- Afghanistan, Central African Republic, Iraq, Syria, etc

* No negotiations with terrorist groups

* Result: civil wars are less likely to end 
]
---

# Questions

.font150[
* Will these norms change with the rise of China? What would be the impact of a new bipolar world on civil war duration?

* Why is there no push for democratisation from European countries?

* Should the US promote democracy abroad? 
]
---

# Questions?

<html><div style='float:left'></div><hr color='#EB811B' size=1px width=720px></html> 
---

# Findley & Young (2015)

# Findley & Young (2015)

* Extend civil war research beyond rebels versus governments

* *Spoilers*: individuals or small groups that derail a peace process

* *De facto* "veto players" in negotiations
]
---

# Terrorism in civil conflicts

.font150[
* Groups can use terrorism in three different ways during the peace process:
	- Before, during, and after the peace process

* **Question**: why would they do so?]
--

* During the process, to increase their bargaining power

* After the process, to alter or derail the original agreement
]
---

# Hypotheses

* H2. As terrorism occurs more frequently, the duration of postwar peace should decrease.

* Mechanisms: erodes trust among parties and create incentives for individuals to fight

* They proxy spoiling by a measure of terrorist attacks directly connected to civil wars
	]
---

# Results

# Results

# Discussion

.font150[
* The results confirm both hypotheses:
	- Terrorism increases the duration of civil wars
	- And it also increases the chance a new conflict will erupot

* **Question**: Kalyvas states that indiscriminate violence is counter-productive. Under what conditions his statement is true? And when is it false?
]

---

# Questions?

<html><div style='float:left'></div><hr color='#EB811B' size=1px width=720px></html> 
---

---

# Toft (2010)

# Two views on civil war termination

.font150[
* First, negotiated agreements are always the best option 
	- The sooner the conflict ends, the better

* "Give war a chance"
	- Decisive victories lead to more stable states

* What are the pros and cons of each approach?]
--

* (Yes, "giving war a chance" is a serious argument!)
]
---

# Some background information

.font150[
* Civil wars than ended in victory are twice as likely to remain settled when compared to cease-fires or negotiated agreements

* Wars ended through negotiated settlement were twice as likely to reignite as those ending in victory

* Rebel victories were more stable than government victories
]
--
.font150[
* *Should we then hope for a rebel victory?*
]
---

# Results

# Results

# Discussion

* Negotiated settlements, in contrast, increase them (model 2)

* Government victory *does not* seem to decrease civil war duration

* The effect comes from rebel victories

* No difference between wars than ended in victories or settlements in any measure except duration
]
---

# Post-civil war politics

* Polity IV: -10 (most authoritarian) to +10 (most democratic)
]
---

# Results

# Discussion

.font150[
* Negotiated settlements are associated with greater levels of authoritarianism as the postwar period is extended

* Incumbent governments seem to sink precipitously into authoritarianism

* Cease-fires/stalemates do not appear to have an impact on democratisation

* *Levels of autocracy decreased after rebel victories*]
---

# Post-civil war prosperity

# Discussion

* Rebel victory does not seem to negatively impact economic growth
]
---

# Mechanisms

.font150[
* **Question**: why do insurgents promote better governance when they win the war? Do they? 
]
--
.font150[
* Rebels have to address popular grievances

* They do not have international recognition

* Face little opposition to implement reforms
]
---

# Case study 1: Uganda

# Case study 1: Uganda

# Case study 1: Uganda

# Case study 2: Somaliland

# Case study 2: Somaliland

# Case study 2: Somaliland

# See you next week!