"Power Struggles and the Natural Resource Curse" Francisco Caselli (LSE) April 2006 Francisco Caselli (LSE) () Power Struggles April 2006 1 / 14
The Resource curse Example: Nigeria , ! started generating signi…cant oil revenues in 1965 , ! per capita growth since then has been close to zero More general observations: , ! countries with large natural resource endowments are not generally rich , ! extremely resource–abundant countries have not experienced sustained growth Puzzle: one can imagine a country squandering its resource wealth. But how can it make them “worse o¤”? Francisco Caselli (LSE) () Power Struggles April 2006 2 / 14
Chart 3: Cumulative Revenues from Oil, 1965-2000 (at 1995 Prices) 350 3500 300 3000 cumulative revenues per 250 2500 capita, right scale 200 2000 cumulative revenues in US$ billions, left scale 150 1500 GDP per capita in 100 1000 constant US$, right scale 50 500 0 0 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999
Basic idea of this paper Natural resources are more easily appropriated by governing elites than other sources of wealth , ! countries with more natural resources experience stronger competition for power , ! this reduces the e¤ective rate of return to incumbents from investing in a country’s development Model is consistent with observations that countries with more natural resources (1) adopt worse policies and (2) experience more civil con‡ict Francisco Caselli (LSE) () Power Struggles April 2006 3 / 14
Model Assumptions Two period economy: t = 1 , 2 All agent’s risk neutral and no discounting Two activities , ! exploitation of natural resources: generates per-period ‡ow α per capita , ! “industry” generates per capita output y t = ρ h t where h t = human capital per capita Period 2 human capital: h 2 = h 1 + I where h 1 is exogenous and I = government investment (period 1) Francisco Caselli (LSE) () Power Struggles April 2006 4 / 14
Two sources of government revenue , ! direct recipient of income from resources , ! tax on industrial output, τ < 1 Allocation of political power , ! exogenous in period 1 , ! in period 2 a “leader” emerges who may stage a coup If a coup takes place, it succeeds with probability γ , ! if it succeeds the leader forms the new government , ! if it fails, the leader incurs a punishment cost, D Francisco Caselli (LSE) () Power Struggles April 2006 5 / 14
Succeed N Coup Fail L Invest in h No coup Succeed N G Coup Fail Don't Invest L No coup Figure: Game Tree Francisco Caselli (LSE) () Power Struggles April 2006 6 / 14
Analysis Solve by backwards induction. Conditional on h 2 , should the “leader” stage a coup? Expected payo¤ of leader if she stages a coup: γ [ α + τρ h 2 ] � ( 1 � γ ) D Expected payo¤ if she doesn’t ( 1 � τ ) ρ h 2 A coup takes place if and only if γα � ( 1 � τ � γτ ) ρ h 2 > ( 1 � γ ) D Francisco Caselli (LSE) () Power Struggles April 2006 7 / 14
Assuming 1 � τ > γτ then a coup takes place if and only if h 2 < h � = γα � ( 1 � γ ) D ( 1 � τ � γτ ) ρ , ! threat of coup gives current government incentive to invest in human capital ! the required level h � is increasing in α , , ! governments with more natural resources must invest more to avoid challenges to their power , ! if α 2 � ( 1 � γ ) D / γ , a coup never takes place Francisco Caselli (LSE) () Power Struggles April 2006 8 / 14
Incumbent government’s problem in period 1 Choose I to maximize expected income α + τρ h 1 + Z ( α + τρ h 2 ) � I where � 1 if h 2 � h � Z = if h 2 < h � 1 � γ subject to h 2 = h 1 + I I � ρτ h 1 + α , ! assumes no government debt Francisco Caselli (LSE) () Power Struggles April 2006 9 / 14
Assume τρ > 1, but ( 1 � γ ) τρ < 1 , ! then, if the government has enough period 1 income to ensure h 2 � h � , it will invest to the max: I � = ρτ h 1 + α and no coup will occur ! but if h 1 + I � < h � it will not invest at all and a coup may occur , So, no investment if h 1 + ρτ h 1 + α < γα � ( 1 � γ ) D ( 1 � τ � γτ ) ρ Francisco Caselli (LSE) () Power Struggles April 2006 10 / 14
This implies that if α > α � = ( 1 � γ ) D + ( 1 + ρτ ) ( 1 � τ � γτ ) ρ h 1 γ + τρ + γτρ � ρ the period 1 government makes no investment and faces a coup in period 2. Otherwise, it uses all its resources to develop the country and no coup occurs ) countries with a lot of natural resources fail to industrialize and experience political instability Francisco Caselli (LSE) () Power Struggles April 2006 11 / 14
Implications for GDP GDP in period 2 is given by � α + ρ h 1 if α > α � GDP = if α < α � α + ρ [( 1 + τρ ) h 1 + α ] ) non–monotonic relationship , ! GDP rises with α within the two groups of countries, but falls at the critical value α � GDP growth between periods 1 and 2 ( if α > α � 1 GDP in period 2 � � GDP in period 1 = α + τρ h 1 if α < α � 1 + ρ α + ρ h 1 ! growth increases with α for α < α � , but falls to zero when α � α � , Francisco Caselli (LSE) () Power Struggles April 2006 12 / 14
GDP Resource α ∗ Endow ment Figure: Implied relationship between Resources and Per capita GDP Francisco Caselli (LSE) () Power Struggles April 2006 13 / 14
Policy Implications There are gains from reducing the amount of royalties that rule elites can appropriate for their own bene…t , ! reduces incentives fo potential chanllengers to stage a coup d’etat , ! increases incumbent’s chances of staying in power , ! raises rate of return on development investments Re‡ected in recent policy initiatives , ! UK’s Extractive Industries Development Initiative (EITI) , ! World Bank conditions for …nancing Chad-Cameroon oil pipeline , ! IMF requires borrowers to be more open about use of resource revenues Francisco Caselli (LSE) () Power Struggles April 2006 14 / 14
Recommend
More recommend
