The economics of civil wars: A survey
Free interpretative arrangements; mostly from: C. Blattman and E. Miguel (2010), Civil wars, Journal of Economic Literature, 48(1), 3-57.
UCDP Armed Conflict Database
Incidence of civil war/conflict over time: 1960-2006
Incidence of civil war/conflict by duration: 1960-2006
Incidence of civil war/conflict by per capita GDP: 1960-2006
Modeling: methodology
Group rationality and strategic behavior of insurgents/counterinsurgents/state Appropriation vs production - contest success function:
We (economists) do not do this! Ohhh no!
Modeling: Individual participation
Individual choice: 1. opportunity costs of fighting and expected benefits theoretical implications for per capita income are uncertain 2. individuals place inherent value on retaliation against an unjust state Principal agent: 1. rebel leaders use material incentives/ethnic appeals to motivate citizens incentives and methods of recruitment vary with distancegeographic or socialbetween the leader and the recruit 2. threats and punishments as selective incentives, in the absence of a shared social basis for mobilizing support - rationale for using coercion on low productivity recruits (especially children) Coalition formation: 1. Along racial/ethnic groups to commit not to defect to winner 2. Along pre-existing networks for efficient information transmission/coordination of violent enforcement
Modeling: War as rational conflict resolution mechanism
Correlations with civil war incidence: Cross-country evidence
Correlation: slow current growth per capita income Income shocks natural resources/total exports low levels of secondary school attainment rough terrain population size Uncorrelated after conditioning: ethnic fractionalization income inequality democracy
Correlation/causation caveat
Example: civil wars ---------- per capita income 1. low per capita income civil war (via small opportunity cost of fighting) 1. civil war low per capita income (via destruction of physical, human, and social capital) 1. ethnic fragmentation low per capita income civil war
Natural experiments, IVs, better measurement
rainfall as IV for income shocks (yes! but possibly: via govt resources and opportunity costs of farmers) import-export price shocks as IV for income shocks (no) more precise measures of natural resources: oil, diamonds (yes!) better theoretically grounded measures: income polarization for inequality (yes!), ethnic/religious fractionalization (?) political institutions and governance from case studies: enclave production, weakness of democratic institutions from polity IV, new vs established democracies, inclusiveness of political institutions, geographic concentration of power, state control over the geographic periphery, direct taxation for state efficiency (unclear!) related lhs variables: length (correlated with ethnic fractionalization), termination
Descriptive factors from case studies
external financing to sustain insurgencies earlier state repression persistent ethnic or elite class dominance peripheral regions with weak central government Institutional/historical transformations: the nineteenth century incorporation of most of Africa and Asia into European empires, and mid-twentieth century formation of nation-states in those regions contagion from neighboring countries refugee flows foreign aid Cold War interventions
Missing (?)
models of spread of rebellion, social (global/local) interactions, networks, contagion
(across individuals, not countries and I am not talking about Twitter and Facebook)
micro-data appropriate to test and estimate these models
Off-topic: qualms between academic disciplines
Why didnt anybody complain about the inability of political scientists to accurately anticipate the Arab spring and they did about economists and the 2008 financial crisis? Dedicato a Giovanni Sartori e ai suoi bravi
