1. travel
2. tourist destinations
3. africa

Fiscal Policy and Civil Conflict
Alvaro Aguirre∗
C´esar Calder´on†
Central Bank of Chile
The World Bank
May 18, 2015
[PRELIMINARY DRAFT]
Abstract
We explore empirically the effect of fiscal policy responses to economic shocks on the likelihood of conflict. Our main finding is that a countercyclical fiscal response to these shocks
lowers the likelihood of conflict, particularly in Africa and when considering shocks to the price
of mineral commodities as the triggers to macroeconomic cycles. These results are stronger when
considering more recent periods and only non-military spending. Although macroeconomic income shocks are behind conflict onset in our framework, the effect of the fiscal response seems
to be associated to mechanisms that go beyond macroeconomic stabilization.
∗
†
[email protected]. Click here for the latest version of this paper.
[email protected]
1
1
Introduction
The disastrous effects of civil wars and their prevalence in poor countries have spawned a research
agenda that tries to understand their occurrence and persistence.1 Two different theoretical perspectives frame this analysis (Rule, 1989). The first one emphasizes relative deprivation as the
motivation for political action. The second stresses that, regardless of grievances, conflict is determined by the ability to mobilize resources to create and sustain rebel groups. In this context of
greed versus grievances, as put forth by Collier and Hoeffler (2004), the role of income inequality,
lack of political rights, and ethnic and religious divisions, may be downplayed by geography and
external support.
Perhaps since it relates to both greed and grievances hypotheses, economic growth has been one
of the most studied determinants of civil conflict (See for instance Miguel et al., 2004; Hegre and
Sambanis, 2006; Br¨
uckner and Ciccone, 2010; Dube and Vargas, 2013; Bazzi and Blattman, 2014).
Under the greed hypothesis it influences the opportunity costs of fighting (Chassang and Padroi-Miquel, 2009), the gains from state appropriation (Besley and Persson, 2011), and the capacity
of the state to bargain or fight insurgencies (Fearon and Laitin, 2003; Bazzi and Blattman, 2014).
In term of grievances, the larger availability of resources may be used to alleviate the problems of
marginalized groups, reducing the likelihood of conflict, or, alternatively, to increase it if resources
are appropriated by political elites.
But despite the abundant literature on the effect of economic shocks on conflict, the role of
stabilization policies has been neglected. The macroeconomic framework, particularly the conduct
of fiscal policy, may influence both the impact of shocks on economic growth, and the effects of
growth on conflict. The first, although studied in different contexts, is well known. The second has
not been explored. In the case of poor countries fiscal management may downplay the link between
growth and the income of the poor, providing social insurance and guaranteing the provision of
public goods and investment programs.2 Hence different macroeconomic policies could generate an
heterogeneous relationship between economic shocks and conflicts, both across countries and time.
The aim of this paper is to fill this gap, and explore the different strategies on the macroeconomic
front, and specifically fiscal management, to lower the vulnerability of countries to civil conflict.3 We
modify previous empirical work introducing interactions of growth shocks, specifically commodity
1
See Sambanis (2002), Collier and Hoeffler (2007), and Blattman and Miguel (2010) for detailed reviews of this
literature.
2
Calder´
on and Levy-Yeyati (2009) show that the adverse effects of aggregate volatility on inequality and poverty
are mitigated by public expenditure. Their focus is on the size of the public sector, but ours is on the short-term
response of fiscal spending. This feature differentiates our paper from others who explore the relationship between
fiscal variables and conflict (Savun and Tirone, 2012; Singh et al., 2014). See the literature review for a comparative
analysis of the two approaches.
3
We focus on fiscal policy because it is more likely to influence the relationship between economic activity and
the risk of conflict. Additional macroeconomic policies, such as monetary and exchange rate policies, have a clearer
role smoothing shocks to output.
2
prices shocks, and indices capturing fiscal responses. Since these and other economic policies
depend on deeper, slow-moving country characteristics, that are also deep determinants of conflict
(i.e. political institutions), we focus on time, instead of cross-country variation, to identify the
effects of policies. We are aware however that, still in this framework, the effect of policies on
conflict is dependent on these features, and hence the improvement of policies is not likely to
change the long-run incidence of conflict in countries that do not improve institutions at the same
time. In our view this last step, although difficult, may be facilitated by temporarily reducing the
incidence of violent conflict through a better macroeconomic management.
We find a significant effect of commodity price shocks on conflict onset. This effect varies across
countries depending on institutional and geographical features, and it is stronger when considering
not only full blown civil conflicts but small armed conflicts as well. Fiscal spending growth does
not seem to affect the likelihood of conflict when not interacted with commodity price shocks. The
novel and main finding is that the sensitivity of violent conflict to economic shocks depends on
the fiscal response to them. Although we consider a large cross-section of countries, these results
are mostly found in the sample of African countries. This is the region with the highest incidence
of conflict, and the one where the growth-conflict nexus seems to be the strongest (Br¨
uckner and
Ciccone, 2010; Bazzi and Blattman, 2014). The interaction between commodity price shocks and
the change in fiscal spending is positive and significant, meaning that a countercyclical response to
these shocks lowers the likelihood of conflict. The facts that more spending increases the likelihood
of conflict when facing a good price shock, and that when it is not interacted with shocks spending
does not affect conflict, reduce the likelihood that our results are capturing reverse causality from
conflict to spending. Results are mostly driven by minerals and persist when we differentiate
between positive and negative shocks. Although weaker, we also find some conditionality to fiscal
responses in the case of large civil wars. In this case minerals are also important, and increases in
spending, both after positive and negative shocks, reduce risks. When exploring possible changes
in the relationship between conflict, economic shocks, and fiscal policy we find that this hasn’t
been stable. The response of fiscal spending has been much more relevant since the 1990s. Finally,
using a smaller sample, we confirm that our results are not driven by changes in military spending.
When considering this type of spending we find a positive relationship with conflict risk, probably
due to reverse causality.
Additional work is needed to understand the mechanisms behind our findings. What we find
when considering the effect of our explanatory variables on growth is that, although the unconditional effect of commodity shocks is significant, validating growth as the main channel in that case,
the response of fiscal policy to shocks is not. Hence results suggest a role for fiscal policy beyond
its effects on macro stability, which may be related to social insurance or productive investments.
After a brief literature review, section 2 presents the empirical strategy and describes the data.
Section 3 shows the econometric results and section 4 concludes.
3
1.1
Literature Review
There is a large empirical literature on the effects of growth on civil conflict. Hegre and Sambanis (2006), after an extensive empirical exploration of the literature on civil war determinants,
conclude that the relationship between growth and conflict is particularly robust. More recently
the literature has focused on finding a causal relationship between growth and civil war trough
the use of instrumental variables and panel regressions. Miguel et al. (2004) use rainfall growth
as an instrument for economic stagnation to explain, successfully, the onset of civil wars in subSaharan Africa. Br¨
uckner and Ciccone (2010) use variations in international commodity prices
as the exogenous determinant. They find that, between 1981 and 2006, a 20% drop in countries
export price indices raised the probability of civil war outbreak by around 2.8 percentage points
in Sub-Saharan Africa. They show that commodity prices affect the likelihood of conflict because
of their effect on economic growth. Bazzi and Blattman (2014) find weaker results for a larger
sample of countries, but they still find a significant effect of commodity price shocks, particularly
on the intensity of fighting. Introducing policies may help to explain the apparent instability in
the relationship between shocks and conflict.4 Dube and Vargas (2013) focus on Colombia and find
differential effects depending on the type of commodities, with those easily-captured commodities
raising violence, and the rest, basically agricultural and labor intensive commodities, lowering it.
Our empirical framework is based on this papers. We use commodity price shocks as an exogenous
determinant of growth to see how stabilization policies may reduce the vulnerability to conflict,
and we explore heterogenous effects depending on the type of commodities.
The role of macroeconomic policies intermediating the link between shocks and conflict has not
been studied before. Singh et al. (2014) explore if public spending has a mitigating effect on conflict.
But they study a different dimension of fiscal policy. They focus on how differences across countries
in terms of the size of public spending to GDP, and its allocation, affect the incidence of conflict in
natural-resource abundant countries. We focus on variations over time of total spending growth,
without considering cross-section variability. Hence our results are associated to the short-run
response of fiscal policy to shocks (i.e. fiscal policy cyclicality), while theirs is associated to state
capacities and conflict (on this see also Fjelde and De Soysa, 2009; Thies, 2010). Moreover they
focus on the link between natural resources and conflict –oil reserves are the main driver of conflict
in their framework–, while we focus on the link between growth and conflict, and use commodity
prices only as an exogenous determinant of growth following the literature just described. Relatedly,
Savun and Tirone (2012) explore the interaction of shocks with foreign aid in generating civil wars,
but they results are influenced by cross-section differences as well, they don’t explore how fiscal
4
For instance the relationship between rainfall shocks and civil conflict found by Miguel et al. (2004) for the period
before 2000 in sub-Saharan Africa disappears afterwards (Ciccone, 2011), apparently due to the weakening of it effect
on GDP growth (Miguel and Satyanath, 2011). Similarly the different strength in the findings by Br¨
uckner and
Ciccone (2010) and Bazzi and Blattman (2014) are in part due to different samples of countries and periods.
4
policy responds to shocks, and their focus is on agricultural commodities only.5
2
Empirical Strategy
We extend Br¨
uckner and Ciccone (2010) framework to investigate the effect of fiscal policy on the
impact of commodity prices on conflict onset. Our baseline specification, for county i and year t,
is the following,
coi,t = αi + δt + β0 dlnpi,t + β1 dlnpi,t−1:2 + β2 dlngi,t−1:2 + β3 dlnpi,t−1:2 × dlngi,t−1:2 + i,t
where αi and δt are fixed and time effects respectively, and the subindex t − 1 : 2 denotes the
two-years average lag of the corresponding variable.
The dependent variable is coi,t , a conflict onset dummy. It takes a value of 1 if there is conflict
in period t but not in period t − 1, and a value of 0 if there is no conflict in period t. Otherwise, if
in both periods t and t − 1 there is conflict, the country-year pair for t is dropped from the sample.
Hence the specification includes all peace years and (the first year of) onset. Neither persistence
nor conflict ending are included. We do this because violent conflict constraints macroeconomic
management, the focus of our study, and because during years of conflict data on fiscal spending
is less reliable.6
The explanatory variables are the percentage change in commodity prices, dlnp, and in public
spending, dlng. For the last variable we use its lagged value to reduce the likelihood of capturing
reverse causality. Additionally we use the two-years average to minimize the effects of measurement
problems. Hence we split the effect of commodity prices in a contemporary effect, and in a twoyears lagged effect. Although we drop from the sample large commodity exporters, separating
contemporary and lagged effects of commodity shocks also helps in minimizing the effect of reverse
causality in our estimations.7 We explore heterogenous effects decomposing the commodity price
index by type of commodity, and by the sign of its change. We also differentiate public spending
between military and non-military spending, although in a smaller sample.
Behind the vulnerability to conflicts there are deep determinants, mainly associated with geography and political institutions. These, on the other hand, may be important determinants of
macroeconomic policies. Since most of the variation of these deep determinants is observed across
countries, we exploit only time variation in our specification to try to isolate the effect of macro
5
Our paper is also related to the literature that studies the effects of fiscal consolidation and social unrest (See
for instance Morrisson et al., 1994; Haggard et al., 1995; Ponticelli and Voth, 2011; Voth, 2013). These papers focus
on fiscal adjustments and not on the response of fiscal policy to shocks, as we do.
6
Bazzi and Blattman (2014) implement this strategy to avoid the bias when the dynamics of the dependent variable
is ignored. They also include a separate regression for episodes when conflicts end.
7
African countries dropped from the sample are Ivory Coast (35% of cocoabeans’ World exports) and Senegal
(47% of ground nut oil).
5
policies on conflict onset. In addition to use the change in spending instead of its level as the explanatory variable, which allows us to focus on the stabilization role of fiscal policy, we adjust the
variation in public spending by country means, so the interaction does not capture cross-section differences in public spending behavior. Additionally we control for interactions between commodity
price shocks and variables capturing deeper determinants of conflict and policies.
Note that the elasticity of conflict onset to commodity price shocks in our framework is defined
as,
∂coi,t
= β1 + β3 dlngi,t−1:2 .
∂dlnpi,t−1:2
We expect from the results in previous papers that β1 < 0. Under a positive value for β3 a
procyclical fiscal policy (i.e. positively correlated with the shock) lowers the drop in the probability
of conflict after a positive price shock, and increases the raise in the probability when facing a
negative price shock. A countercyclical policy has the opposite effects and hence it would be the
type of policy that minimizes the probability of conflict onset under β3 > 0.
2.1
Data
We collect data on civil conflict, commodity prices, and public spending for 105 countries, of which
42 are from Africa, from 1960 to 2013. The list of the countries included is presented in Table 1.8
In the following paragraph we describe the sources and present basic statistics.
Civil Conflict: The source is the UCDP/PRIO Armed Conflict Dataset (Harbom et al., 2008).
UCDP/PRIO defines armed conflict as “a contested incompatibility that concerns government
and/or territory where the use of armed force between two parties, of which at least one is the
government of a state, results in at least 25 battle-related deaths in a year.” In particular internal
armed conflict occurs between the government of a state and one or more internal opposition
group(s) regardless of intervention from other states. This data allows us to distinguish between
small and large conflicts, an exercise we implement below. In particular we distinguish between
civil conflicts, an episode with more than 25 battle-related deaths in a year, and civil wars, where
the threshold raises to 100 deaths.
Table 2 describes the conflict data in our sample of 105 countries. We show the probability of
conflict onset conditional on observing peace, which is our dependent variable, and the unconditional probability of conflict onset (i.e. considering both peace and conflict years in the denominator). The first are 4.3% and 1.1% for civil conflicts and wars respectively. These are almost
8
For most of our specifications we present results for three groups of countries; African countries (42), which
include both North and Sub-Saharan Africa; Developing Countries (89), which include the group of African countries
and the other developing countries listed in Table 1; and World (105), including the group of developing countries
and the group of industrial countries listed in Table 1. We also consider a smaller sample of African countries that
are resource rich. They are marked with an asterisk in Table 1. The sample is obtained after dropping observations
for which data on the variables considered is not available, dropping countries with less than 5 observations, and, as
mentioned above, dropping large exporters (see footnote 7).
6
completely concentrated in developing countries, particularly in Africa where the probabilities rise
to 5.8% and 1.6%, respectively. Among these countries, those that are resource rich are significantly more prone to conflict, as these probabilities raise to 8.4% and 2.4%, respectively. When
considering the unconditional probability conflict onset occurs in 2.7% and 0.9% of our sample in
the case of civil conflicts and wars, respectively. Table 2 also shows how prevalent is conflict in
the countries analyzed. While a 12% of the time the average country is in conflict, a 2.5% is in
civil war. These numbers rise to more than 14% and 3.5% in the case of African countries, and to
18% and 5% when taking into account only resource-rich economies. It is clear from these numbers
that civil wars are rare events, and perhaps because of this it is more difficult to find systematic
determinants of them, as our results will show.
Figure 1 shows the temporal pattern of conflict. The left-hand side panel shows conflict prevalence, for each type of conflict, and for Africa and the rest of the world separately. In the last
group we can see an inverse U-shaped prevalence of conflict, with a peak around 1990. For Africa,
although weaker, we observe a similar shape but with a different peak, closer to the year 2000. In
the right-hand side panel we show conflict onset, which doesn’t seem to show a pattern as clear as
in the case of prevalence.
Commodity prices: We construct a commodity price index following the methodology proposed
by Collier and Goderis (2012), which is based on Deaton and Miller (1995) and Dehn (2000).
We collect yearly commodity price indices for 50 different commodities, which is as many as data
availability allows. For each country we first construct a weighted average of commodity prices,
considering only commodities in which the country is a net exporter. The weight for each commodity
is its participation on total net exports in 1990 of the commodities considered.9 Additionally we
construct subindexes for minerals, agricultural commodities, and oil.10
Table 3 shows summary statistics of the commodity indexes. Panel A presents statistics for the
aggregate index, and panels B, C, and D for minerals, agricultural, and oil subindexes, respectively.
We denote by wi the fraction of net commodity exports in GDP in the base year, which is 1990.
In our sample of 42 African countries the average fraction of net commodity exports is close to
10%.11 A similar number is observed in other developing countries, but not in developed countries,
which show a smaller fraction. In terms of volatility the standard deviations are similar across
groups, with Africa showing a slightly higher number. In the case of subindexes most countries
9
As in Collier and Goderis (2012), prices are deflated by an annual world average of export unit values. We also
followed the adjustments proposed by these authors to treat gaps and missing values. See their paper for details.
10
The 50 commodities are aluminium, natural gas, phosphatrock, uranium, urea, coal, ironore, nickel, silver,
copper, lead, tin, and zinc, which we classify as minerals; bananas, cotton, oliveoil, pulp, sugar, barley, fish, oranges,
rice, sunfloweroil, butter, groundnutoil, palmkerneloil, rubber, swinemeat, cocoabeans, groundnuts, palmoil, sisal,
tea, coconutoil, hides, pepper, sorghum, tobacco, coffee, jutes, plywood, soybeanoil, wheat, copra, maize, poultry,
soybeans, and wool, classified as agricultural commodities; and gasoline and oil, which we classify as oil.
11
We don’t show the medians to save space but they are generally lower than the means. In the case of Africa the
median wi is 5.4%.
7
have positive net agricultural exports, while in the case of minerals and oil only two thirds and
one sixth show positive numbers. This last two fractions are lower in Africa, although conditioning
on being positive, the mean and median values are in general higher than in the other groups.
Volatilities are similar across groups and subindexes, although the price of commodities related to
oil seems to be more volatile when considering developing countries in and out of Africa.
Public Spending: The third main variable is the change in public spending. We use data from
the World Bank’s WDI. We obtain a measure of real public spending multiplying GDP in LCU and
the public spending to GDP ratio. This increases the sample size considerably relative to using
public spending in LCU, and the correlation with this last variable is close to 1. The data shows
extreme observations so we drop the highest and lowest 2.5% changes in spending of the entire
sample.12 On average spending grows at a rate of 4.3% per year in our sample, with developed
countries showing a slightly lower rate, around 3.9%. The standard deviation is much higher
in developing countries. In Africa the average standard deviation is 10.8%, while in developing
countries out of Africa it is 8.6%. We also consider differences between military and non-military
public spending in our regressions. The World Bank’s WDI publishes data on military spending as
a fraction of GDP only since 1988, so the sample size is reduced importantly. The average growth
rate of military spending in the 87 countries for which we have data is 2.4%, and the standard
deviation is 13.7%. The highest numbers are observed in Africa, with an average growth of 3.1%,
and a standard deviation of 19%.
Before showing regression results we explore the relationship between our three main variables of
interest. To do this we first identify all the episodes of conflict onset for which there was no conflict
in the previous three years. Then we compute the median value of the changes in commodity prices
and public spending across all the episodes, before and after the onset of a conflict. Figure 2 shows
the results, both for African countries and for the rest of the world. It also presents the median
values for the two variables of interest for all the years and countries in the sample. On the lefthand side panel civil conflict onset episodes are considered. The clearest feature is the large drop,
from a relatively high value, in commodity prices during the two years before the onset of conflict
in African countries. This is not observed in the rest of the countries, where the drop occurs in
the first year after the onset. We also find a more pronounced change in fiscal spending in Africa,
which growth rate starts rising two years before the onset, from a null change to a rate close to 6%
in the year the conflict begins. In the case of civil wars we observe a similar pattern, but with a
relatively smaller change in commodity prices and a relatively larger fall in fiscal spending growth.
12
The fact that we use two-years average also helps with this problem. The standard deviation of the percentage
change in public spending is reduced from 8.8% to 6.6% after averaging it for two years.
8
3
Regression Results
In this section we present the regression results. Before considering public spending we first explore
in detail the relationship between commodity price changes and conflict onset. Table 4 shows the
results. The first 9 columns show the results using civil conflicts as the dependent variable, and the
remaining columns use civil wars. First we restrict the sample to African countries. In column (1)
we include the contemporary change in the commodity index and its two lags. While negative in all
cases, only the first lag is significant. We include the two-lags average in column (2), our preferred
specification for the reasons indicated above, and this is significant at the 10% confidence level.
When we include the average of the three years in column (3) the coefficient becomes not significant.
We repeat the same for developing countries (including Africa), and for the entire sample. In none
of the cases we find a significant effect of commodity prices. Something similar occurs in the case
of civil wars (columns 10-18), although in this case commodity prices are not significant in Africa,
and have a positive, and sometimes significant effect, when adding other developing countries.
As already mentioned we control for interactions of commodity prices with deeper determinants
of conflict and policies. After testing different variables we selected the ones that were more significant. One is associated to geography, the wheat-sugar ratio, which is a measure of land suitability
for wheat versus sugarcane, used by Easterly (2007) as an instrument for income inequality. The
second is associated to political institutions, the index of executive constraints from the Polity IV
database, for which we use its average value from 1960 to 2013.13 Results are shown in Table 5,
where these interactions are included in our preferred specification with the average of two lags of
commodity prices. Both the wheat-sugar ratio and executive constraints are evaluated at the mean
of each group. The interactions between lagged commodity prices and both of these variables are
significant for all the groups of countries in the case of civil conflicts (left panel). This means that a
higher wheat-sugar ratio is associated with a higher sensitivity of civil conflict onset to commodity
price changes. The opposite is observed in the case of average executive constraints, meaning that
countries with better institutions are less sensitive to commodity price shocks. In the case of civil
wars (right panel), only the interaction with executive constraints is significant, and the sign is
negative. This means that better institutions make the conflict elasticity to price shocks higher.
This may be explained by a possible nonmonotonicity in the relationship between institutions and
conflict (Hegre et al., 2001). These results show that the relationship between commodity prices
and civil conflict is heterogenous across countries. But it is important to notice that since we
exploit only time variation in spending our results are not influenced by these interactions, as these
variables are fixed over time.14 However, when interpreting the results, magnitudes are valid for
13
Additionally we test the significance of mountainous terrain, ethnolinguistic fractionalization, polarization, and
the Gini-Greenberg index proposed by Esteban et al. (2012), the share of agricultural land occupied by family farms
in 1858, constructed by Vanhanen (2003), a measure of rain variability, the ICRG index, average urbanization and
population density, the fraction of agricultural land and fertile soil, latitude, and a dummy for tropical climate.
14
We try time-varying variables as well, but none of them where significant. These were the time-varying component
9
the average country in each group, since we evaluate the interaction with fixed effects at that level.
3.1
Baseline Results
Table 6 shows the results from estimating the baseline specification, for each type of conflict and for
African, developing (including Africa), and all the countries separately. We show results with and
without public spending using our preferred specification, which is separating the contemporary
effect from the average two-years lagged effects.
As already shown in the case of civil conflicts and Africa we find a significant effect of commodity
prices (column 1). In column (2) we include public spending, and although it is not significant by
itself, its interaction with commodity prices is. To have a better idea of the magnitudes involved
we adjust the change in public spending by its mean for African countries, which is 4,4%. Hence,
for a country with that level of spending growth, the coefficient rises from -0.09 to -0.15, and
becomes significant at a 1% level of significance. The interaction effect means that the likelihood
of civil conflict becomes more sensitive to changes in commodity prices as the growth rate of public
spending goes down. To see this we plot in Figure 3 the effect on the probability of civil conflict to
an average 10% decrease in commodity prices for two years, as a function of public spending growth.
The unconditional elasticity, reported in column 1 of Table 6, is very close to the elasticity estimated
by Br¨
uckner and Ciccone (2010) (in green), if we assume in this last case that the average drop of
10% occurs for three years. In black it is shown the unconditional elasticity of civil conflict onset
in the region, which is close to 6% as reported in Table 2. The estimated conditional elasticity (in
red) means that an increase in spending larger than 10%, which is less than one standard deviation
above the mean for African countries, eliminates the effect from the drop in commodity prices.
It also implies that negative growth rates in spending make the response significantly larger than
the unconditional response. For instance, if public spending remains constant, the increase in the
probability of conflict onset is twice as large than in the unconditional case, and one-third of the
unconditional likelihood.
Going back to Table 6 we can see that these effects are only present in the case of Africa. When
we include other developing countries in columns (3) and (4), and developed countries in columns
(5) and (6), we don’t find any significant coefficient. The same happens when we consider civil
wars for every group of countries in columns (7) to (12).
To further illustrate the implications of our baseline estimations we explore actual events of
large drops in the price of commodities in Africa. In particular we compute the change in the
predicted probability of conflict using the coefficients in column 2 of Table 6 when, instead of using
the observed change in spending, we use the sample average to evaluate this predicted probability.
Table 7 shows the results. We select episodes where the commodity index fell more than 10%, and
pick the 10% with the largest increase in the predicted probability, and the 10% with the largest
of executive constraints, democracy (both from Polity), the ICRG index, urbanization, and population density.
10
reduction. In the upper panel are presented the episodes where an expansionary response of fiscal
policy generated the largest reduction in the probability of conflict onset. The table shows the
country and the year of the episode, the change in the commodity price index and in government
spending, and the change in the fitted value of the probability of conflict (dco).
ˆ Despite the large
drops in commodity prices, in none of these episodes a conflict actually started. Botswana shows
the three largest episodes, with reductions in the predicted likelihood of conflict larger than 7%.
Most of these episodes are concentrated in the 1970s and 1980s. The latest events are Nigeria in
1999 and Mali in 2000, when drops of roughly 15% in the commodity indexes were accompanied by
increases of more than 20% in fiscal spending, implying reductions in the likelihood of conflict of 4%
in both cases. In the lower panel we show the largest increases in the predicted probability. In two
out of these 19 episodes a conflict actually started; Chad in 1976 and Sudan in 1983. According to
our estimates the fiscal contraction, of 2.8% and 16.4%, increased the probability of conflict onset
in 2.5 and 3.6% respectively. The latest event is Botswana 2005, when a 5.2% fall in fiscal spending
increased the fitted probability in 2.4%.
3.2
Heterogeneous Effects
Not all commodities may have the same effect on conflict risk. In Table 8 we explore heterogeneity across different commodity types for our sample of African economies, again considering civil
conflicts and wars. For comparisons, in columns (1) and (5) we present the baseline results, which
correspond to those in columns (2) and (8) of Table 6. In columns (2) to (4) we split the commodity
index in agricultural, minerals, and oil subindexes, and show the results using civil conflict onset
as the dependent variable. We include all of the subindexes in a single regression, so the results
in columns (2) to (4) are estimated jointly. Although the signs of the variables of interest are the
same than in column (1), significance is only found in the case of minerals (column 3). Although
agricultural commodities show a similar pattern, it seems that minerals are the ones generating
the non-linear effect of commodity prices on conflict. In the case of civil wars (columns 6-8), the
coefficients are not significant for every subindex. For larger samples, which we don’t show to save
space, results are similar. In developing countries and civil conflicts the interaction of minerals and
spending is significant, although close to one half of the coefficient found for Africa.
Results so far indicate that public spending reduces the sensitivity of conflict onset to commodity
price shocks, particularly in the case of minerals. Note that, although we interpret the results
considering a drop in commodity prices, the same would happen in the case of a positive shock.
Higher spending growth would reduce the fall in the likelihood of conflict in this case. Then, a
countercyclical fiscal policy during booms would be desirable not only to generate resources to
spend in bad times, but also because of a direct effect in the risk of conflict. A lower rate of
spending growth may make the country less vulnerable to other type of shocks or saving resources
may reduce the incentives to fight over them. To see if spending acts in this way both after a fall
11
and rise in prices, we split commodity shocks by their sign. Results are shown in Table 9. In column
(1) we consider the index including all of the commodities. The interaction is positive for positive
as well as negative shocks, although the coefficient is larger, and significant, in the first case. To
better understand the result we plot in the left-hand side panel of Figure 4 the response to a 10%
positive and negative shock, with the corresponding 10% confidence intervals. Although the slope
is small in the case of negative shocks, the prediction on the desirability of a countercyclical fiscal
policy is validated: to reduce the likelihood of a civil conflict not only a country need to spend
more when facing a bad shock, but even more efficiently, it has to spend less when receiving a good
shock.
To see if this depends on the type of commodity we present the results from estimating a
single equation including the three commodity prices subindexes in columns (2) to (4) of Table 9.
Significance is against observed for minerals, but unlike in column (1), the effect is stronger in the
case of a negative shock. In the center panel of Figure 4 we present the corresponding elasticities
and confidence intervals. Hence, although not significant, the difference in terms of magnitudes
obtained in column (1) seems to come from agricultural goods, for which the interaction in the case
of positive shocks is larger.
On the right-hand side panel of Table 9 we show results considering civil wars. Unlike in the
other cases this variable is used, when differentiating between negative and positive shocks we do
find significant coefficients. And again these correspond to minerals. In the case of positive shocks
the sign is the opposite than in the case of civil conflicts. As shown in the right-hand side panel of
Figure 4, higher spending lowers the likelihood of conflict, in a statistically significant magnitude,
when facing both good and bad shocks. It is worth to notice that the coefficient on public spending,
not shown in Table 9 to save space, is not significant. Note also that the change in sign is obtained
for agricultural goods as well when moving from civil conflicts to wars (columns 2 and 6). Finally,
when we include other developing countries results are very similar, and even the interaction with
positive shocks in column (6) becomes significant at the 5% significance level. Although we cannot
test the mechanism behind this change with respect to smaller conflicts with our simple framework,
it support the idea that political elites need to signal that they won’t steal the resources (see e.g.
Alesina et al., 2008, for a related mechanism).
3.3
Resource-Rich Countries
Another dimension in which heterogeneity may play a role is the dependency of the country in
natural resource rents. To explore this possibility we split the sample of African countries into
those where rents from natural resources are above and below 10% of GDP according to the WDI.15
15
We use natural resource rents excluding forests. We include Botswana, Liberia, Namibia, Sierra Leone, and
Niger in the group of resource-rich countries despite having rents below the threshold, because of their possession of
diamonds (the first three), and uranium (Niger).
12
These are the countries marked with an asterisk in Table 1. Note that we still drop big exporters of
a particular commodity, to reduce the likelihood of measuring causality from conflict to commodity
prices. Results for civil conflicts are presented in Table ??. For comparisons, the first three columns
show the results for the whole group of African countries. The three columns in the middle present
the results when only the 19 resource-rich countries are included, and the last three columns do
that for the rest of the African countries in the sample. First we can see that the unconditional
effect is larger in resource-rich countries (columns 4 and 7). The interaction, introduced in columns
5 and 8, is also greater in the first group. It is not significant though, probably because of the
smaller size of the sample. When differentiating between positive and negative shocks we observe
significant differences between the two groups. While for resource-rich countries the interaction is
positive and significant in the case of positive shocks only (column 6), it is positive and significant
only in the case of negative shocks for countries that are not rich in natural resources. Therefore
the results imply that resource-rich countries reduce the probability of civil conflict after reducing
spending facing positive shocks, while non-resource-rich countries do the same increasing spending
when facing negative shocks.16
3.4
Temporal Changes
As shown in Figure 1 the pattern of conflict incidence in Africa since 1960 differs from the one
observed in the rest of the developing world. To see if the nature of the relationship between
conflict, commodity prices, and fiscal policy has changed over time as well we estimate the baseline
regressions for Africa, those in columns (2) and (8) of Table 6, in windows of 20 years, from 1960
to 2013. The resulting series for the interaction between commodity prices and fiscal spending, and
its 10% confidence interval, are presented in Figure 5. On the left-hand side panel we can see that
the interaction is more or less stable until de mid eighties, when the point estimate goes from 1.15,
a similar level than the one estimated for the entire sample, to close to 4 in the samples starting
around 1990. In the case of civil wars, shown in the right-hand side panel, we find an increase in
the coefficient as well, although sooner (around 1985), and the coefficient remains not significant
as in the entire sample.
To explore more in detail the change in the interaction for civil conflicts we present in Table 11
the results from estimating our baseline specifications, including differences by signs of the shocks,
for the years before and after 1990. The first thing to notice is that the unconditional effect of
commodity prices (columns 1 and 4) drops to one half and becomes not significant in the second
period. Then, as expected from the rolling regressions just analyzed, the interaction term with
16
We don’t present results for civil wars to save space. Results in this case show that the unconditional effect
of commodity prices becomes significant in both groups. The sign differs though. Commodity price shocks have a
positive effect on the probability of civil war onset in the case of resource-rich countries, but a negative effect, in line
with the result for civil conflicts, in resource-poor countries.
13
public spending has the opposite variation. It rises from 0.85 to 3.05 (columns 2 and 5). This
means that the lower sensitivity to commodity prices is due to the fiscal response to shocks. To
better understand these results in columns (3) and (6) we differentiate the interaction term by the
sign of the shock. We can see that before 1990 a positive spending response to a drop in commodity
prices could reduce their effect on conflict, while a response to an increase in prices didn’t change
the response of conflict probability. After 1990 the pattern is the opposite. A negative spending
response to a positive shock make even less likely the onset of a conflict, while in the case of
a negative price shock spending does not have a significant effect moderating the increase in the
probability. Splitting the commodity index by type of commodity increases importantly the number
of regressors, and since the sample size is smaller, results are very unprecise. Because of this we
don’t show the results, which seem to indicate that most of the changes come from the effects of
agricultural commodities.
3.5
Military Spending
Results may be influenced by military spending. If that were the case then the mechanisms behind
our results would not be related to the stabilization or insurance roles of fiscal policy. Moreover it
would be more likely that the results were capturing reverse causality. To explore this possibility
we differentiate between military and non-military spending. However, as already mentioned, data
on military spending is available for a shorter period of time. The WDI publishes data on this
category only from 1988 onwards, which roughly coincides with the first year of the second subsample considered in the previous sub-section. Regression results for the three groups of countries,
considering separately military and non-military spending, are presented in Table ??. For each
group the first column present the baseline results for comparisons, using total spending as before,
but adjusting the sample to countries and periods where military spending is available. Column
(1) shows that the interaction is still positive in the case of Africa, and similar to the coefficient
estimated for the second sub-sample in Table 11. However, due to the smaller sample, the coefficient
is not significant at the 10% confidence level. In column (2) we split public spending into military
(M ) and non-military (N M ) spending. The coefficient on non-military spending is positive, large,
and highly significant. Military spending on the other hand is negative and not significant. This
would imply that it is non-military spending which is behind the results obtained so far in the
paper. In the left-hand side panel of Figure 6 we show the corresponding elasticity and confidence
interval for the case of a drop in commodity prices.
In columns (3) and (4) we explore heterogeneous effects with respect to the sign of the shock.
Again for comparisons, column (3) shows the results with total spending. Although results are still
similar to those for the second sub-sample in Table 11, the coefficients are still not significant. But
in column (4), where we distinguish again by the type of spending, results are significant for each
type of spending, and for each sign of the shock. In the case of non-military spending results are
14
in line with what we have found in the previous specifications; a countercyclical fiscal policy lowers
the likelihood of conflict when facing a commodity shock, irrespectively of its sign. Results are even
stronger than the ones found in the second sub-sample using total spending (column 6 of Table 11).
The conditional elasticities and their respective confidence intervals are shown in the center panel
of Figure 6. Results for military spending differ. As shown in the right-hand side panel of Figure 6,
lower increases in this type of spending after commodity shocks are always associated with a lower
probability of conflict onset. Hence, it is very likely in this case causality is running from this last
variable to spending changes.
We repeat the estimations for the group of developing countries (columns 5 to 8) and for the
entire sample (columns 9 to 12). As before results become weaker. The exception is the effect of a
response of non-military spending to a positive price shock, which is large and significant in both
groups. The pattern is similar to what we found for Africa when considering the second sub-sample
(column 6 of Table 11). It implies that the response of conflict probability to negative price shocks
is unconditional on spending. However, in the case of a positive price shock, a countercyclical
response reduce the probability of conflict risk in a statistically significant magnitude.
3.6
Growth Effects
Br¨
uckner and Ciccone (2010) show that the significant effects of commodity price shocks on conflict
onset are due to the effects on growth. To explore the same question but regarding the interaction
between these shocks and public spending we run the baseline regressions but using GDP growth
as the dependent variable. For comparisons we present again the results when civil conflict onset
is the dependent variable in columns (1) to (4). In column (5) we can see that the commodity
price index do have a significant effect on GDP growth, both contemporarily and with lags. This is
consistent with the results presented by Br¨
uckner and Ciccone (2010). But note that in the case of
the interaction with public spending, unlike when explaining civil conflict in column (1), the effect
is not significant. When decomposing the index (columns 6 to 8) we can see that only in the case
of agricultural commodities the effect of commodity prices is significant, and more importantly for
our results, even in the case of minerals, the interactions are not significant. Hence it is likely
that the response of fiscal spending to commodity shocks influence the likelihood of conflict trough
channels that are not necessarily related to economic growth. The effect of fiscal policy on the
relationship between growth and conflict, and not between commodity shocks and growth, seems
to be the relevant dimension behind our results.
4
Conclusions
The results confirm the negative effect of commodity shocks on the likelihood of conflict onset
in Africa (for the average country), and economic growth as the main channel behind this. More
15
importantly for our analysis is that the effect is conditional on public spending: drops in commodity
prices that coincides with an increase in public spending, and hikes that coincides with low spending,
reduce the risk of conflict. And this doesn’t seem to be driven by military spending or the effect of
fiscal policy on growth.
16
References
Alesina, A., F. R. Campante, and G. Tabellini (2008): “Why Is Fiscal Policy Often Procyclical?,” Journal of the European Economic Association, 6(5), 1006–1036.
Bazzi, S., and C. Blattman (2014): “Economic shocks and conflict: the Evidence from commodity prices,” American Economic Journal: Macroeconomics, Forthcoming.
Besley, T. J., and T. Persson (2011): “The Logic of Political Violence,” Quarterly Journal of
Economics, 78(1), 1–34.
Blattman, C., and E. Miguel (2010): “Civil War,” Journal of Economic Literature, 48(1),
3–57.
¨ ckner, M., and A. Ciccone (2010): “International Commodity Prices, Growth and the
Bru
Outbreak of Civil War in Sub-Saharan Africa*,” The Economic Journal, 120(544), 519–534.
´ n, C., and E. Levy-Yeyati (2009): “Zooming in: From Aggregate Volatility to Income
Caldero
Distribution,” World Bank Policy Research Working Paper Series 4895.
Chassang, S., and G. Padro-i-Miquel (2009): “Economic shocks and civil war,” Quarterly
Journal of Political Science, 4(3), 211–228.
Ciccone, A. (2011): “Economic shocks and civil conflict: A comment,” American Economic
Journal: Applied Economics, 3(4), 215–227.
Collier, P., and B. Goderis (2012): “Commodity prices and growth: An empirical investigation,” European Economic Review, 56(6), 1241–1260.
Collier, P., and A. Hoeffler (2004): “Greed and Grievance in Civil War,” Oxford Economic
Papers, 56(4), 563–595.
(2007): “Civil War,” Handbook of Defense Economics, 2, 711–739.
Deaton, A., and R. I. Miller (1995): International Commodity Prices, Macroeconomic Performance, And Politics In Sub-Saharan Africa. International Finance Section, Department of
Economics, Princeton University.
Dehn, J. (2000): “Commodity Price Uncertainty In Developing Countries,” World Bank Policy
Research Working Paper N.2426.
Dube, O., and J. F. Vargas (2013): “Commodity price shocks and civil conflict: Evidence from
Colombia,” The Review of Economic Studies, 80(4), 1384–1421.
17
Easterly, W. (2007): “Inequality does cause underdevelopment: Insights from a new instrument,”
Journal of Development Economics, 84(2), 755 – 776.
Esteban, J., L. Mayoral, and D. Ray (2012): “Ethnicity and Conflict: An Empirical Study,”
American Economic Review, 102(4), 1310–42.
Fearon, J. D., and D. D. Laitin (2003): “Ethnicity, Insurgency, and Civil War,” American
Political Science Review, 97(1), 75–90.
Fjelde, H., and I. De Soysa (2009): “Coercion, Co-optation, or Cooperation? State Capacity
and the Risk of Civil War, 1961-2004,” Conflict Management and Peace Science, 26(1), 5–25.
Haggard, S., J.-D. Lafay, and C. Morrisson (1995): The political feasibility of adjustment in
developing countries. Development Centre of the Organisation for Economic Co-operation and
Development.
Harbom, L., E. Melander, and P. Wallensteen (2008): “Dyadic Dimensions of Armed
Conflict, 1946-2007,” Journal of Peace Research, 45(5), 697–710.
Hegre, H., T. Ellingsen, S. Gates, and N. P. Gleditsch (2001): “Toward A Democratic
Civil Peace? Democracy, Political Change, And Civil War, 1816–1992,” American Political
Science Association, 95(01), 33–48.
Hegre, H., and N. Sambanis (2006): “Sensitivity Analysis of Empirical Results on Civil War
Onset,” Journal of Conflict Resolution, 50(4), 508–535.
Miguel, E., and S. Satyanath (2011): “Re-examining economic shocks and civil conflict,”
American Economic Journal: Applied Economics, 3(4), 228–232.
Miguel, E., S. Satyanath, and E. Sergenti (2004): “Economic Shocks and Civil Conflicts:
An Instrumental Variable Approach,” Journal of Political Economy, 112(4), 725–753.
Morrisson, C., J.-D. Lafay, and S. Dessus (1994): Adjustment programmes and politicoeconomic interactions in developing countries: lessons from an empirical analysis of Africa in
the 1980s. Royal African Society.
Ponticelli, J., and J. Voth (2011): “Austerity and anarchy: Budget cuts and social unrest in
Europe, 1919-2008,” .
Rule, J. B. (1989): Theories of civil violence. Univ of California Press.
Sambanis, N. (2002): “A review of recent advances and future directions in the quantitative
literature on civil war,” Defence and Peace Economics, 13(3), 215–243.
18
Savun, B., and D. C. Tirone (2012): “Exogenous Shocks, Foreign Aid, and Civil War.,” International Organization, 66(3).
Singh, R. J., C. Bodea, and M. Higashijima (2014): “Oil and civil conflict: can public spending
have a mitigation effect?,” World Bank Policy Research Working Paper, (7100).
Thies, C. G. (2010): “Of rulers, rebels, and revenue: State capacity, civil war onset, and primary
commodities,” Journal of Peace Research, 47(3), 321–332.
Vanhanen, T. (2003): Democratization: A Comparative Analysis of 170 Countries, Routledge
Research in Comparative Politics, 7. Taylor & Francis Group.
Voth, H.-J. (2013): “Tightening Tensions: Fiscal Policy and Civil Unrest in Eleven South American Countries, 1937-1995,” in Fiscal Policy and Macroeconomic Performance, ed. by L. F.
C´espedes, and J. Gali, pp. 59–92. Central bank of Chile.
19
Africa
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Developing
Zambia
Somalia
Gabon
Swaziland
Namibia
Liberia
Kenya
Angola
Djibouti
Mauritania
Congo, Rep.
Burkina Faso
Equatorial Guinea
Nigeria
Sierra Leone
Guinea-Bissau
Malawi
Rwanda
Madagascar
Morocco
Lesotho
Guinea
Mali
Niger
Central African Republic
Libya
Egypt, Arab Rep.
Chad
Sudan
Togo
South Africa
Cameroon
Mozambique
Ghana
Burundi
Botswana
Congo, Dem. Rep.
Gambia, The
Benin
Algeria
Tanzania
Zimbabwe
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
Yemen, Rep.
Nicaragua
Paraguay
Sri Lanka
Iran, Islamic Rep.
Papua New Guinea
Saudi Arabia
Guyana
Mexico
Syrian Arab Republic
Mongolia
Jamaica
Honduras
Peru
Uruguay
Guatemala
Jordan
Bolivia
Dominican Republic
Vietnam
Tunisia
Cambodia
Kuwait
India
Hungary
Costa Rica
Suriname
Fiji
Turkey
Panama
Pakistan
Lao PDR
Ecuador
Thailand
Nepal
Lebanon
Czech Republic
Bulgaria
Oman
Albania
Qatar
Poland
Romania
Haiti
El Salvador
Venezuela, RB
United Arab Emirates
World
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
Norway
Finland
New Zealand
Switzerland
Greece
Japan
Korea, Rep.
United Kingdom
Denmark
Ireland
Portugal
Sweden
France
Austria
Italy
Belgium
Notes: In the tables showing regression results Developing includes African and Other
Developing countries, and World includes African, Other Developing, and Industrial countries. Asterisks denote African countries classified as resource-rich.
Table 1: Countries in the Sample
20
Conflict Onset
Conditional on peace
Africa
Africa, Resource Rich
Other Developing
Developed
Total
Conflict Prevalence
Unconditional
Countries
Civil Conflict
Civil War
Civil Conflict
Civil War
Civil Conflict
Civil War
42
19
47
16
105
5.79
8.39
4.16
0.43
4.25
1.58
2.40
0.99
0.00
1.07
3.59
4.27
2.67
0.25
2.67
1.30
1.84
0.85
0.00
0.90
14.27
17.78
13.24
2.70
12.05
3.47
4.77
2.50
0.00
2.51
Table 2: Civil Conflict: Summary Statistics
Africa
Other Developing
Developed
Total
42
9.9
17.4
47
9.8
16.8
16
2.0
16.3
105
8.6
17.0
42
4.0
16.4
47
4.3
15.8
15
1.0
16.2
104
3.7
16.1
21
3.0
16.0
33
1.2
18.4
16
0.4
19.0
70
1.6
17.8
9
20.7
20.9
15
14.6
21.3
6
1.8
14.5
30
13.9
19.8
A. All commodities
wi > 0
Mean wi
Mean sd(∆Pi,t )
B. Agricultural
wi > 0
Mean wi
Mean sd(∆Pi,t )
C. Minerals
wi > 0
Mean wi
Mean sd(∆Pi,t )
D. Oil
wi > 0
Mean wi
Mean sd(∆Pi,t )
Notes: wi is net exports of the commodities considered in each index as a fraction of GDP in 1990. Its
mean a median values, and the corresponding statistics for the volatility of the index, are computed
across all the countries in the group for which its value is positive.
Table 3: Commodity Prices: Summary Statistics
21
22
0.14
1260
0.14
1260
0.11
3127
0.11
3127
0.04
0.14
1260
0.01
(6)
0.07
0.03
−0.01
0.02
0.03
(5)
−0.09
0.11
3127
0.02
0.04
0.05
−0.01
−0.01
−0.08
∗
0.03
−0.00
0.02
0.03
(4)
0.04
−0.07
0.03
0.03
∗
−0.01
−0.00
(3)
Developing
0.11
3887
0.02
−0.01
0.02
−0.01
0.02
0.02
(7)
0.11
3887
0.03
−0.02
0.02
0.02
(8)
World
0.11
3887
0.04
0.00
(9)
0.12
1434
0.02
0.03
0.02
0.00
0.02
0.00
(10)
0.12
1434
0.03
0.04
0.02
−0.00
(11)
Africa
0.12
1434
0.04
0.03
(12)
0.08
3516
0.01
0.08
3516
0.02
0.08
3516
0.02
0.08
4298
0.01
0.02∗
0.01
0.01
−0.00
(16)
0.02∗
0.03
(15)
0.01
0.03
0.01
−0.00
(14)
0.01
0.01
0.01
−0.00
(13)
Developing
Civil Wars
0.08
4298
0.02
0.03∗
0.01
−0.00
(17)
World
0.08
4298
0.02
0.02
(18)
Table 4: Commodity Prices and Civil Conflict
Notes: The dependent variable is a conflict onset dummy. The subindexes t − 1 : 2 and t : 3 indicate the average of the first two-lags, and these and the current value,
of the explanatory variable respectively. Developing includes African and Other Developing countries, and World includes African, Other Developing, and Industrial
countries. Fixed and time effects are included in all the regressions. Standard errors are clustered by country.
R2
Observations
dlnPt:3
dlnPt−1:2
dlnPt−2
dlnPt−1
dlnPt
(2)
(1)
Africa
Civil Conflict
Civil Conflict
dlnPt
Africa
(1)
Developing
(2)
World
(3)
Africa
(4)
Developing
(5)
World
(6)
−0.01
0.03
0.03
−0.00
−0.01
−0.00
0.04
0.03
0.02
0.02
0.01
0.01
−0.01
−0.02
0.03
0.03
0.03∗
0.05
0.04
0.03
0.03
0.02
0.02
−0.05
0.05
0.03
0.07
0.01
0.02
0.12
0.11
0.08
0.06
0.04
0.03
−0.09
dlnPt−1:2
dlnPt ×
wheat-sugar
dlnPt ×
executive constraints
dlnPt−1:2 ×
wheat-sugar
dlnPt−1:2 ×
executive constraints
R2
Observations
Civil Wars
∗
∗
−0.04
−0.06
−0.07
−0.11
−0.02
−0.01
0.16
0.09
0.07
0.06
0.03
0.02
−0.50∗∗
−0.40∗∗
−0.34∗∗∗
−0.03
−0.04
0.00
0.22
0.17
0.13
0.10
0.07
0.04
0.32∗∗
0.19
0.16∗∗
−0.27∗∗
−0.11∗∗
−0.07∗
0.16
0.12
0.08
0.10
0.05
0.04
0.14
1225
0.11
2591
0.11
3352
0.12
1400
0.09
2980
0.08
3763
Notes: The dependent variable is a conflict onset dummy. The subindex t − 1 : 2 indicates the average of the first two lags
of the explanatory variable. Developing includes African and Other Developing countries, and World includes African, Other
Developing, and Industrial countries. Executive constraints and the wheat-sugar ratio are adjusted by the mean of each group.
Fixed and time effects are included in all the regressions. Standard errors are clustered by country.
Table 5: Commodity Prices and Civil Conflict, Interactions with Country-Fixed Variables
Civil Conflict
Africa
dlnPt
dlnPt−1:2
Africa
Developing
World
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
−0.01
−0.01
0.03
0.03
0.03
0.03
−0.00
−0.00
−0.01
−0.01
−0.00
−0.00
0.04
0.04
0.03
0.02
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.01
−0.09∗
−0.15∗∗∗
−0.01
−0.02
−0.02
−0.03
0.03
0.04
0.03
0.04
0.03∗
0.04∗
0.05
0.06
0.04
0.04
0.03
0.04
0.03
0.04
0.02
0.03
0.02
0.02
dlnPt−1:2 × dlnGt−1:2
R
Observations
World
(1)
dlnGt−1:2
2
Developing
Civil Wars
0.14
1225
−0.04
0.04
0.02
−0.03
−0.01
−0.01
0.08
0.07
0.06
0.06
0.04
0.04
1.15∗∗
0.24
0.23
−0.20
−0.21
−0.22
0.48
0.36
0.33
0.33
0.20
0.19
0.14
1225
0.11
2591
0.11
2591
0.11
3352
0.11
3352
0.12
1400
0.12
1400
0.09
2980
0.09
2980
0.08
3763
0.08
3763
Notes: The dependent variable is a conflict onset dummy. The subindex t − 1 : 2 indicates the average of the first two lags
of the explanatory variable. Fixed and time effects, and interactions of commodity prices with institutions (average executive
constraints from polity) and the wheat-sugar ratio, are included as additional regressors (both variables are adjusted by each
group’s mean). The change in fiscal expenditure is adjusted by country means, and normalized by the mean for African
countries. Developing includes African and Other Developing countries, and World includes African, Other Developing, and
Industrial countries. Standard errors are clustered by country.
Table 6: Commodity Prices, Fiscal Spending, and Civil Conflict
23
Country
Year
dlnPt−1:2
dlnGt−1:2
dco
ˆ
Botswana
Botswana
Botswana
Swaziland
Lesotho
Congo, Dem. Rep.
Gabon
Niger
Cameroon
Cameroon
Mali
Ghana
Kenya
Rwanda
Mauritania
Nigeria
Gambia, The
Lesotho
Burundi
1975
1976
1982
1977
1976
1989
1987
1976
1987
1988
2000
1987
1973
1976
1973
1999
1988
1981
1988
-42.2
-38.9
-35.5
-39.2
-38.9
-21.8
-37.1
-31.4
-26.2
-20.4
-15.6
-10.3
-13.9
-15.9
-14.3
-14.8
-14.6
-10.9
-19.0
25.9
19.7
20.4
18.0
18.1
26.5
18.0
15.9
17.7
19.5
22.7
29.2
23.6
20.9
22.4
21.4
21.4
24.9
17.2
-11.5
-7.6
-7.3
-6.8
-6.8
-6.5
-6.4
-4.7
-4.6
-4.2
-4.1
-4.0
-3.9
-3.8
-3.8
-3.7
-3.6
-3.5
-3.4
Botswana
Mali
Togo
Niger
Rwanda
Lesotho
Burkina Faso
Egypt, Arab Rep.
Botswana
Benin
Congo, Dem. Rep.
Chad
Namibia
Congo, Dem. Rep.
Guinea
Nigeria
Sudan
Zambia
Liberia
1969
1983
1983
1982
1980
1965
2000
1988
2005
1976
1980
1976
1990
1973
1993
1987
1983
2000
1976
-22.5
-11.0
-10.4
-20.7
-18.5
-18.3
-15.1
-14.1
-18.4
-27.1
-19.4
-27.4
-22.0
-14.9
-16.2
-37.0
-11.4
-14.3
-47.7
-1.9
-6.7
-7.8
-2.9
-4.1
-4.6
-6.2
-6.8
-5.2
-2.8
-5.2
-2.8
-4.6
-9.2
-9.5
-3.0
-16.4
-15.9
-6.2
1.8
1.8
2.0
2.0
2.1
2.2
2.3
2.3
2.4
2.5
2.5
2.5
2.6
2.9
3.1
3.4
3.6
4.2
6.1
Table 7: Fiscal Spending and Conflict Onset: Mayor Events
24
Civil Conflict
Equation 1:
Baseline
dlnPt
dlnPt−1:2
dlnPt−1:2 × dlnGt−1:2
2
R
Observations
Civil Wars
Equation 1:
Baseline
Equation 2: by type
Equation 2: by type
(1)
Agro
(2)
Minerals
(3)
Oil
(4)
(5)
Agro
(6)
Minerals
(7)
Oil
(8)
−0.01
−0.01
−0.03
0.03
−0.00
−0.01
0.02
−0.02
0.04
0.03
0.07
0.05
0.02
0.02
0.03
0.05
−0.15
∗∗∗
−0.08
−0.15
−0.18
0.04
0.01
0.01
−0.04
0.06
0.06
0.10
0.15
0.04
0.05
0.07
0.04
1.15∗∗
0.76
1.28∗∗
0.12
−0.20
−0.25
0.04
−0.08
0.48
0.58
0.57
1.34
0.33
0.40
0.29
0.19
0.14
1225
0.15
1225
0.12
1400
0.12
1399
Notes: The dependent variable is a conflict onset dummy. The subindex t − 1 : 2 indicates the average of the first two lags
of the explanatory variable. Fixed and time effects, the change in public spending, and interactions of commodity prices with
institutions (average executive constraints from polity) and the wheat-sugar ratio, are included as additional regressors (the last
variables are adjusted by their African mean). The change in fiscal expenditure is adjusted by country means, and normalized
by the mean for African countries. Standard errors are clustered by country.
Table 8: Commodity Prices and Civil Conflict in Africa, by Type of Commodity
Civil Conflict
Equation 1:
Baseline
dlnPt
dlnPt−1:2
dlnPt−1:2 > 0 × dlnGt−1:2
dlnPt−1:2 < 0 × dlnGt−1:2
2
R
Observations
Civil Wars
Equation 2: by type
Equation 1:
Baseline
Equation 2: by type
(1)
Agro
(2)
Minerals
(3)
Oil
(4)
(5)
Agro
(6)
Minerals
(7)
Oil
(8)
−0.01
−0.00
−0.02
0.03
−0.00
−0.01
0.03
−0.02
0.04
0.03
0.07
0.05
0.02
0.02
0.03
0.05
∗∗∗
−0.15
−0.08
−0.15
−0.18
0.03
0.01
0.02
−0.05
0.06
0.06
0.10
0.14
0.04
0.05
0.06
0.04
1.70∗
1.00
0.88
0.37
−0.98
−0.79
−1.23∗∗
0.06
0.92
1.02
0.59
2.75
0.65
0.71
0.52
0.47
0.66
0.53
1.65∗
−0.11
0.53
0.23
1.12∗
−0.22
0.57
0.59
0.90
0.47
0.40
0.42
0.65
0.47
0.15
1225
0.15
1225
0.13
1400
0.13
1399
Notes: The dependent variable is a conflict onset dummy. The subindex t − 1 : 2 indicates the average of the first two lags
of the explanatory variable. Fixed and time effects, the change in public spending, and interactions of commodity prices with
institutions (average executive constraints from polity) and the wheat-sugar ratio, are included as additional regressors (the last
variables are adjusted by their African mean). The change in fiscal expenditure is adjusted by country means, and normalized
by the mean for African countries. Standard errors are clustered by country.
Table 9: Commodity Prices and Civil Conflict in Africa, by Type of Commodity and Sign of Shock
25
All Africa
dlnPt
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
−0.01
−0.01
−0.01
−0.03
−0.03
−0.04
0.01
0.02
0.02
0.04
0.04
0.04
0.08
0.08
0.08
0.06
0.06
0.06
−0.11
−0.15
−0.16
0.13
0.13
0.13
∗
0.05
dlnPt−1:2 × dlnGt−1:2
−0.15
∗∗∗
∗∗∗
∗
−0.15
0.06
−0.15
0.06
0.08
−0.21
∗∗
0.10
1.38
0.48
0.98
1.70
0.14
1225
0.14
1225
0.72
0.45
∗
dlnPt−1:2 < 0 × dlnGt−1:2
−0.22
∗∗
0.10
1.15∗∗
dlnPt−1:2 > 0 × dlnGt−1:2
R2
Observations
Rest
(1)
−0.09
dlnPt−1:2
Resource-Rich
3.89
∗∗
0.47
0.92
1.98
0.77
0.66
−0.56
1.00∗
0.57
1.10
0.60
0.15
1225
0.28
486
0.28
486
0.29
486
0.11
727
0.11
727
0.11
727
Notes: The dependent variable is the civil conflict onset dummy. The subindex t − 1 : 2 indicates the average of the first two
lags of the explanatory variable. Fixed and time effects, the change in public spending, and interactions of commodity prices
with institutions (average executive constraints from polity) and the wheat-sugar ratio, are included as additional regressors
(the last variables are adjusted by each group’s mean). The change in fiscal expenditure is adjusted by country means, and
normalized by the mean for African countries. See Table 1 for African countries classified as resource rich. Standard errors are
clustered by country.
Table 10: Commodity Prices and Civil Conflict in Africa, Resource-Rich Countries
1960-1989
dlnPt
dlnPt−1:2
1990-2013
(1)
(2)
(3)
(4)
(5)
(6)
−0.05
−0.05
−0.05
0.04
0.05
0.04
0.05
0.05
0.05
0.06
0.06
0.06
−0.13∗
−0.19∗∗
−0.20∗∗
−0.07
−0.20∗∗
−0.19∗∗
0.07
0.08
0.08
0.09
0.10
0.10
∗∗
dlnPt−1:2 × dlnGt−1:2
∗∗
0.90
3.18
0.44
dlnPt−1:2 > 0 × dlnGt−1:2
dlnPt−1:2 < 0 × dlnGt−1:2
2
R
Observations
0.21
622
0.21
622
1.60
0.03
4.55∗∗
0.88
2.24
1.60∗∗
0.72
0.81
2.79
0.21
622
0.23
603
0.24
603
0.24
603
Notes: The dependent variable is the civil conflict onset dummy. The subindex t − 1 : 2 indicates the average of the first two
lags of the explanatory variable. Fixed and time effects, the change in public spending, and interactions of commodity prices
with institutions (average executive constraints from polity) and the wheat-sugar ratio, are included as additional regressors
(the last variables are adjusted by their African mean). The change in fiscal expenditure is adjusted by country means in each
subsample, and normalized by the mean for African countries. Standard errors are clustered by country.
Table 11: Commodity Prices and Civil Conflict in Africa, Subsamples
26
Africa
dlnPt
dlnPt−1:2
dlnGt−1:2
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
−0.02
−0.04
−0.03
−0.02
0.01
0.00
0.01
0.00
0.01
0.00
0.01
0.01
0.09
0.08
0.09
0.08
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.02
−0.30∗∗
−0.25∗
−0.29∗
−0.20
−0.06
−0.10
−0.05
−0.06
−0.06
−0.10
−0.06
−0.06
0.15
0.14
0.15
0.13
0.07
0.08
0.07
0.07
0.06
0.06
0.06
0.06
−0.23
−0.34
0.03
−0.05
0.02
−0.05
0.18
0.23
0.12
0.16
0.11
0.14
dlnMt−1:2
−0.27∗
−0.18
−0.02
−0.10
−0.03
−0.10
0.15
0.15
0.09
0.08
0.08
0.08
0.06
−0.24∗∗
0.07
−0.07
0.07
−0.07
0.08
0.11
0.06
0.07
0.05
0.07
2.78
0.64
0.56
1.72
0.78
0.73
dlnPt−1:2 × dlnN Mt−1:2
dlnPt−1:2 × dlnMt−1:2
2.66∗∗∗
1.70∗∗
1.66∗∗
0.33
0.72
0.74
−0.94
−0.10
−0.14
1.23
0.69
0.66
dlnPt−1:2 > 0 × dlnGt−1:2
dlnPt−1:2 < 0 × dlnGt−1:2
3.96
1.41
1.32
2.62
1.30
1.19
0.85
−0.52
−0.54
2.83
1.16
0.99
dlnPt−1:2 > 0 × dlnN Mt−1:2
dlnPt−1:2 < 0 × dlnN Mt−1:2
dlnPt−1:2 > 0 × dlnMt−1:2
R
Observations
1.97∗∗∗
2.06∗∗∗
2.01∗∗∗
0.33
0.58
0.60
5.06∗∗
−0.08
−0.02
2.31
1.37
1.25
2.65
dlnPt−1:2 < 0 × dlnMt−1:2
2
World
(1)
dlnN Mt−1:2
dlnPt−1:2 × dlnGt−1:2
Developing
0.25
447
0.30
447
0.25
447
∗∗
1.30
∗∗
1.23∗∗
1.22
0.63
0.60
−5.17∗∗
−2.08
−2.03
2.51
1.83
1.77
0.33
447
0.24
1058
0.26
1058
0.24
1058
0.26
1058
0.23
1421
0.25
1421
0.23
1421
0.26
1421
Notes: The dependent variable is a conflict onset dummy. The subindex t − 1 : 2 indicates the average of the first two lags of
the explanatory variable. dlnN M and dlnM are the change in non-military and military public spending, respectively. Fixed
and time effects, and interactions of commodity prices with institutions (average executive constraints from polity) and the
wheat-sugar ratio, are included as additional regressors (both variables are adjusted by each group’s mean). The change in
fiscal expenditure is adjusted by country means, and normalized by the mean for African countries. Developing includes African
and Other Developing countries, and World includes African, Other Developing, and Industrial countries. Standard errors are
clustered by country.
Table 12: Commodity Prices and Civil Conflict, Military and Non-military Spending
27
Civil Conflict
Equation 1:
Baseline
dlnPt
dlnPt−1:2
dlnPt−1:2 × dlnGt−1:2
R2
Observations
Output Growth
Equation 2: by type
Equation 1:
Baseline
Equation 2: by type
(1)
Agro
(2)
Minerals
(3)
Oil
(4)
(5)
Agro
(6)
Minerals
(7)
Oil
(8)
−0.01
−0.01
−0.03
0.03
0.03∗∗
0.01
0.03∗∗
0.05∗
0.04
0.03
0.07
0.05
0.01
0.02
0.02
0.02
−0.15∗∗∗
−0.08
−0.15
−0.18
0.07∗∗∗
0.08∗∗∗
0.02
0.06
0.06
0.06
0.10
0.15
0.02
0.03
0.02
0.04
0.12
0.11
−0.05
−0.17
0.31
1.34
0.20
0.20
0.26
0.49
∗∗
1.15
0.76
0.48
0.58
0.14
1225
0.15
1225
∗∗
1.28
0.57
0.17
1224
0.18
1224
Notes: The dependent variable in columns 1-4 is the civil conflict onset dummy. The subindex t − 1 : 2 indicates the average
of the first two lags of the explanatory variable. Fixed and time effects, the change in public spending, and interactions
of commodity prices with institutions (average executive constraints from polity) and the wheat-sugar ratio, are included as
additional regressors (the last variables are adjusted by their African mean). The change in fiscal expenditure is adjusted by
country means, and normalized by the mean for African countries. Standard errors are clustered by country.
Table 13: Commodity Prices, Civil Conflict and GDP growth in Africa
28
Notes:
Figure 1: Civil Conflict, 1960-2013
Notes:
Figure 2: Commodity Prices and Fiscal Spending around Civil Conflict Episodes
29
6.0%
Unconditional Likelihood
5.0%
4.0%
Conditional elasticity
3.0%
2.0%
Unconditional elasticity
1.0%
Bruckner & Ciccone
0.0%
-10% -8% -6% -4% -2%
0%
2%
4%
6%
8% 10% 12% 14%
dlnG
Notes: response of the probability of civil conflict to an average 10% decrease in commodity prices for two years.
dlnG is the two-years average change in public spending. The standard deviation for this variable in the sample of
African countries is 6.6%, and its average 4.4%. Elasticities are evaluated at the African mean value of executive
constraints and the wheat-sugar ratio. Dotted lines are 90% confidence intervals. For Br¨
uckner and Ciccone (2010),
whose estimations are for the period 1961-2006, three-years average changes in P and g are assumed to obtain the
elasticity.
Figure 3: The Impact of a Commodity Price Shock in Africa
6.0%
6.0%
3.0%
Positive price shock
Negative price shock
4.0%
Negative price shock
4.0%
2.0%
2.0%
2.0%
1.0%
0.0%
0.0%
0.0%
-2.0%
-1.0%
-2.0%
Positive price shock
-4.0%
-4.0%
-6.0%
-6.0%
-10% -8% -6% -4% -2%
0%
2%
4%
6%
8% 10% 12% 14%
Negative price shock
-2.0%
Positive price shock
-3.0%
-10% -8% -6% -4% -2%
dlnG
0%
2%
4%
6%
8% 10% 12% 14%
-10% -8% -6% -4% -2%
dlnG
0%
2%
4%
6%
8% 10% 12% 14%
dlnG
Notes: response of the probability of civil conflict (left and center) and civil war (right) to an average 10% change
in commodity prices (left) and mineral prices (center and right), for two years. dlnG is the two-years average
change in public spending. The standard deviation for this variable in the sample of African countries is 6.6%, and
its average 4.4%. Elasticities are evaluated at the African mean value of executive constraints and the wheat-sugar
ratio. Dotted lines are 90% confidence intervals.
Figure 4: The Impact of a Commodity Price Shock in Africa, by Sign of Shock
8
3
7
3
6
2
2
5
1
4
1
3
0
2
-1
1960-1980
1
-11960-1980
1965-1985
1970-1990
1975-1995
1980-2000
1985-2005
1990-2010
-1
0
-2
1965-1985
1970-1990
1975-1995
1980-2000
1985-2005
-2
1990-2010
Notes: interaction term between the growth in commodity price index and public spending estimated by rolling
regressions for 20-years windows from 1960 to 2013. The dependent variables are civil conflict (left) and civil war
(right) onset. Spending is adjusted by the country mean in each subsample. Dotted lines are 90% confidence
intervals.
Figure 5: Interaction between Commodity Price Shocks and Public Spending in Africa, Rolling
Estimations
30
9.0%
8.0%
Negative price shock
7.0%
7.0%
5.0%
4.0%
3.0%
3.0%
Negative price shock
6.0%
5.0%
2.0%
1.0%
0.0%
-1.0%
1.0%
-2.0%
-3.0%
-1.0%
-4.0%
-5.0%
-7.0%
Positive price shock
-6.0%
4%
-8.0% -10% -8% -6% -4% -2%
Positive price shock
-3.0%
-10% -8% -6% -4% -2%
0%
2%
4%
dlnNM
6%
8% 10% 12% 14%
-9.0% -10% -8% -6% -4% -2%
0%
2%
dlnNM
6%
8% 10% 12% 14%
0%
2%
4%
6%
8% 10% 12% 14%
dlnM
Notes: response of the probability of civil conflict to an average 10% change in commodity prices, for two years.
dlnN M and dlnM are the two-years average change in non-military and military public spending, respectively.
The standard deviations for these variables in the sample of African countries are 8.3 and 13%, and their averages
5 and 2.6%, respectively. Elasticities are evaluated at the African mean value of executive constraints and the
wheat-sugar ratio. Dotted lines are 90% confidence intervals.
Figure 6: The Impact of a Commodity Price Shock in Africa, Military and Non-Military Spending
31