Results of M.
Ross’s 2006 Statistical Study of Civil War
Covers 90
civil wars from 1960-1999
“State as
target” model has strong support (I differ with Ross’ interpretation)
-
High
fuel rents increase risk of all CW
o
double
the risk of fuel-less state
o
but
only 40% greater risk if GDP per capita rises too
-
Having
fuel off-shore increases risk of
national and non-ethnic conflicts
-
Mined
diamonds increase of all CW but only slightly (9%)
o
But
can be offset if increase in per capita GDP rises too
o
If
DRC/Zaire cases are dropped, no reliable effect
-
Oddly,
higher non-tax revenue does not increase
CW risk
“Resources
as target model” modest support (I differ with Ross’ interpretation)
-
On-shore
fuel resources & non-fuel rents increases risk CW
§
if
wars in
§
13
of 22 separatist conflicts were based in
oil producing regions
-
we should expect that secessionist
CW would be more likely than national CW if this model is correct
-
only
alluvial diamonds increase risk of secessionist war
o
but
secessionist conflicts not fought in diamond-producing regions
o
only
12 of 90 CWs involved diamond-producing regions
“Rebel
incentives” model has partial support
-
fuel
onshore is correlated with CW
-
but
alluvial diamonds are not correlated in most models
-
others
studies found support for this connection
“Resource
curse” model has partial support
-
fuel
on-shore, off-shore & primary diamonds increases CW risk
-
but
not non-tax revenue
“International
economy” model has partial support
-
tricky
to measure validly
-
oil
shocks (changes in prices of oil) increase CW in half of models
-
negative
shocks only cause separatist wars
-
positive
shocks only cause nationalist wars (which implies “state as target” model is
more accurate
Contraband
exports lengthen civil wars but individual resources do not