Transcript a resource curse - UCSB Economics - University of California, Santa

The Political Economy of the Natural Resource
Curse: An Interpretive Survey
Robert T. Deacon
University of California, Santa Barbara
Resources for the Future
June 2, 2010
17th Ulvön Conference on Environmental Economics
Ulvön, Sweden, June 22-24, 2010
1
Former U.S. President William J. Clinton:
“With … [its] vast human and natural
resources, a revitalized Nigeria can be the
economic and political anchor of West
Africa ….”
From remarks on signing of a joint declaration with Nigerian
President Obasanjo, August 26, 2000.
2
Sheik Ahmed Yamani, former Oil Minister of Saudi Arabia:
“All in all, I wish we had discovered water.”
Cited in Ross, Michael. “The political economy of the resource curse.”
World Politics 1/1 (1999) 297-322.
3
Motivation: anecdotes
Nigeria’s per capita GDP in 2000 was 30% lower than in 1965,
despite oil revenue receipts of $350 billion.
Venezuela’s terms of trade grew 13.7% per year during 19701990, but its output per capita fell by 1.4% per year; ran a current
account deficit.
Saudi Arabia’s GDP per capita was lower in 1999 than it was
before the oil price increases of the 1970s.
OPEC: per capita GNP decreased 1.3% per year during 19651998
Average growth in all lower- and middle-income countries
2.2% per year.
4
5
Motivation: case study evidence
Gold: 16th century Spain
Tin & natural gas: Bolivia
Timber: Indonesia & Philippines
Oil: Venezuela and Nigeria
Bird Guano: Peru
6
Motivation: case study evidence
Gold: 16th century Spain
Tin & natural gas: Bolivia
Timber: Malaysia & Philippines
Oil: Venezuela and Nigeria
Bird Guano: Peru
7
Counter examples: resource rich countries
that are, well, rich
Norway, following North Sea oil boom.
Botswana, diamond capital of the world.
U.S., Canada, Australia, Iceland, New Zealand prospered
from natural resource wealth.
Malaysia, Chile, …
8
Motivation: Why care about the resource
curse?
May provide insight on the key question in economics:
Why do some countries grow and become
prosperous, while others remain poor?
May prove a valuable ‘testing ground’ for models of political
economy.
9
Market-based explanations for the RC
Resource wealth and the Dutch disease
Resource wealth diverts attention from education,
entrepreneurship, …
Volatility in resource prices hinders investment, growth.
Resource rich countries experience long run declining terms of
trade.
10
Market-based explanations for the RC
Dutch disease
Resource rents divert attention from education, entrepreneurship, …
Volatility in terms of trade hinders investment, growth.
Long run declining terms of trade for resource intensive countries.
Unexplained patterns in the data
RC most problematic when governance institutions are weak;
Doesn’t operate when institutions are strong.
RC more problematic for concentrated than diffuse resources.
11
Market-based explanations for the RC
Dutch disease
Resource rents divert attention from education, entrepreneurship, …
Volatility in terms of trade hinders investment, growth.
Long run declining terms of trade for resource intensive countries.
Unexplained patterns in the data
RC most problematic when institutions are weak.
RC more problematic for concentrated than diffuse resources.
Consider political economy explanations for RC
12
Rent-seeking models of the RC
The ‘Voracity’ Effect,
Diverted Entrepreneurship and the Resource Curse
Rent-seeking among Competing Groups and Institutional Decline
Rent-seeking, Violent Conflict and Resource Exploitation Policy
Political Economy (Institutional) Theories of the RC
Rent-induced Regime Transitions,
Public Employment at a Commitment Mechanism,
Protection (Resources) for Sale and Political Competition
Concentrated Power, Resource Rents and Public Goods Supply
Political Competition and Entry Barriers
13
Rent-seeking (Institution-free) Models of RC
The ‘Voracity’ Effect
Tornell and Lane, Amer. Econ. Rev., 1999; Tornell and Velasco, Jour. Pplit.
Econ. (1992); Lane and Tornell, J Econ. Growth, 1996; Tornell and Lane, J
Int. Econ. 1998.
Diverted Entrepreneurship and the Resource Curse
Torvik, J Dev. Econ., 2002; Mehlum, Moene and Torvik, Econ. Jour., 2006.
Rent-seeking among Competing Groups and Institutional Decline
Hodler, Eur. Econ. Rev., 2006.
Rent-seeking, Violent Conflict and Resource Exploitation Policy
Van der Ploeg and Rohner, Working paper, University of Oxford, 2010.
14
Political Economy (Institutional) Models of RC
Rent-induced Regime Transitions
Aslaksen and Torvik, Scand. J. Econ., 2006
Public Employment at a Commitment Mechanism
Robinson, Torvik and Verdier, J Dev. Econ., 2006.
Protection (Resources) for Sale and Political Competition
Bulte and Damania, B.E. Jour. Econ. Anal. & Policy, 2008.
Resource Rents, Concentrated Power, and Public Good Supply
Smith, Jour. of Politics, 2008.
Political Competition and Entry Barriers
Tsui, Working paper, Clemson University, 2009.
15
The Political Response to Windfalls: Voracity,
Growth and the Resource Curse
• Politically powerful groups can costlessly transfer highly productive
capital from the formal sector: ‘formal’ capital stock is a common
pool.
• These groups can transfer capital to a less productive, but
invulnerable, ‘informal’ sector.
• Capital flows from formal to informal sector; growth rate and PV
welfare are suboptimal.
• Availability of secure ‘wealth haven’ may or may not improve
welfare.
16
Voracity, … Response to windfall on formal k
stock rate of return.
• Contrary to (my) intuition, an increase in the rate of return to formal
capital causes transfers by elite groups to increase by more than the
productivity gain.
• Consequently, formal sector capital is reduced
• Contrary to intuition, the size of the transfer is relatively large when the
number of powerful groups is small (but > 1)
….economic outcomes are worse when political power is
concentrated.
• Reminder: no institutional barriers to prevent transfers.
• Personal comment: I haven’t worked so hard to understand a paper
since graduate school.
17
Voracity, …
A representative group’s perceived after-transfer rate of return on
capital left in formal sector:
pa  (n  1) x(t )
where p is output price, a is physical productivity of formal capital
stock, n is number of groups, x is one group’s transfer as a share of
formal capital stock
18
Voracity, …
Introduce a second, informal sector. Its productivity is b < a, but it is
immune to transfers. If rates of return equalize, we have
pa  (n  1) x  b
Each group counts only the transfers other (n -1) groups take from the
stock, when calculating the rate of return it perceives on formal capital.
19
Now the magic happens …
Suppose the rate of return on formal capital increases due to price
increase, Dp. Capital naturally flows from formal to informal sector. If
rates of return are again equalized in equilibrium, the change in
transfers per group, Dx, must satisfy
Dpa  (n  1)Dx  0
since the informal sector’s productivity is unchanged. This implies that
each group’s transfer changes as follows:
Dx  Dpa /( n  1)
20
Now the magic happens …
The aggregate transfer, nx, changes by
n
nDx  Dpa
 Dpa
(n  1)
That is, the increase in aggregate transfer exceeds the productivity
gain. Notice that the aggregate transfer is larger when n is small (so
long as n > 1).
21
What’s going on here?
The equilibrium concept here is Markov perfect equilibrium:
•
Each group’s strategy is a function of payoff relevant state
variable, k.
•
Each group internalizes the fact that its transfer will reduce k,
and therefore affect other groups’ demands.
“Don’t kill the goose that lays the golden egg”
•
If the net-of-transfer rate of return on formal capital falls below
the rate of return on informal capital, each group’s best
response is to transfer the entire stock of k to the informal
sector.
•
Risk of “killing the goose that lays the golden egg” disciplines
transfer demands.
•
May also be extreme equilibria, where all k is transferred.
22
How to Interpret?
‘Small n’ economies (concentrated political power) always perform
worse than ‘large n’ economies (diffuse political power):
•
Economy with larger n always always achieves higher growth
rate and PV utility.
•
However, perverse response to windfalls is more dramatic in
large n economies—pushes their performance down toward
what small n economies achieve.
Tornell and Lane’s (1999) interpretation:
“… if the shift to democracy brings with it the destruction of
entrenched interest groups, and power becomes more
diffused, then growth performance and adjustment to windfalls
will improve.”
Similar to political economy principle, but different reasoning.
23
Evidence on Voracity and Resource Curse
Tornell and Lane (1999) informally examine response of 3 oil-rich
economies, Mexico, Nigeria and Venezuela, to oil price shocks of
1970s and early 1980s.
•
Government transfers as share of GDP more than doubled:
Payments to special interests?
•
Growth rates in all three well below rates predicted from
empirical growth model. Growth negative in Nigeria and
Venezuela!
24
Evidence on Voracity, cont.
Lane and Tornell (1996) examine cross-country cross section data:
•
Controls: Absence of institutional barriers to transfers, ICRG
scores.
•
Controls: Concentrated political interests, concentrated
manufacturing sector.
•
Voracity term: Price shock, average terms of trade growth.
•
Results: Positive price shock increases growth in countries
with strong institutions, but no growth occurs when
institutions are weak.
•
Results: Similar effects on investment rates (formal k
growth), but less robust.
25
Voracity: Untested Implications
Tornell and Lane (1999, 1996): Price spike causes increased ‘theft’ of
formal capital by political interests. Formal capital is largely linked to
oil extraction, so theft may be visible in:
•
Greater risk of expropriation, nationalization of oil capital by
government;
•
Greater frequency of bribes, kickbacks to resource
contractors.
•
Voracity predicts that formal capital will actually shrink
following price jump; possibly visible in data on oil production
investments, etc.
Lane and Tornell (1996): Institutional barriers control is arguably a
measure of ‘theft’ activity; thus endogenous and not a valid control.
26
Rent-seeking, Diverted Entrepreneurship and RC
Historical example: Spain’s appropriation of gold from the New World
(Karl, 1997).
•
Most spectacular windfall in history, linked to natural resource
discovery.
•
Resource curse of epic proportions: eight declarations of
bankruptcy between 1557 and 1680.
What happened (Karl, 1997):
27
Rent-seeking, Diverted Entrepreneurship and RC
Historical example: Spain’s appropriation of gold from the New World:
What happened (Karl, 1997):
“[The monarchy] consolidated the loyalty of the lesser
aristocracy through political favoritism, especially by
selling patents of nobility and ecclesiastical
appointments. This practice dramatically expanded
the size of a parasitic noble class . . . while
simultaneously siphoning off the most productive
talent from business and commerce. …..
“The state bought the talents of those who might
have become small entrepreneurs through
awarding of offices …”
28
Rent-seeking, Diverted Entrepreneurship and RC
Torvik (2002), Mehlum, et al, (2006) develop a model that formalizes
this story.
A fixed number of talented individuals can apply their skills in either of
two ways:
i.
Operating a ‘modern,’ increasing returns firm that can
generate profits.
ii.
Rent-seeking at the public trough.
Allocation of talent depends on returns in both activities, returns
equalized in equilibrium.
−
Resource rent accrues to government, encourages rentseeking.
−
Rent reduces talent allocated to modern firms, so less profit
is generated.
29
Rent-seeking, Diverted Entrepreneurship and RC
Profit from modern firms spurs growth throughout the economy, ‘big
push’ logic.
−
Shifting resources away from modern sector to primitive
sector reduces income.
Return to rent-seeking depends on country’s institutional quality and
size of the resource rent.
30
Rent-seeking, Diverted Entrepreneurship and RC
Predicted effect of larger resource rent:
−
With low quality institutions, larger resource rent shifts talent
away from modern sector to primitive sector and reduces
income, an economic resource curse.
−
An economy otherwise immune to rent-seeking may
succumb, if rents are large enough, a political resource
curse.
31
Empirics: Diverted Entrepreneurship, …
Hypothesis: Resource rent is a curse only when institutions cannot
resist rent-seeking.
•
Format: Sachs-Warner cross country growth (1970-1990)
regression equation.
•
Control for Institutional quality with country ratings on
corruption, contract repudiation, expropriation risk, …
•
Resource rent proxy is share of GDP in primary products
sectors.
•
Control for initial income, openness to trade, education, other
factors.
Result: Resource rent reduces growth when institutions are weak, but
not otherwise.
32
Empirics: Diverted Entrepreneurship, …
Untested predictions:
•
Resource rent windfall can increase rent-seeking activity,
rendering institutional quality endogenous.
•
Resource rent windfall should shift economic activity away
from modern sectors toward primitive sectors.
33
Political Economy (Institutional) Models of RC
Rent-induced Regime Transitions
Aslaksen and Torvik, Scand. J. Econ., 2006
Public Employment at a Commitment Mechanism
Robinson, Torvik and Verdier, J Dev. Econ., 2006.
Protection (Resources) for Sale and Political Competition
Bulte and Damania, B.E. Jour. Econ. Anal. & Policy, 2008.
Resource Rents, Concentrated Power, and Public Good Supply
Smith, Jour. of Politics, 2008.
Political Competition and Entry Barriers
Tsui, Working paper, Clemson University, 2009.
34
Political Economy Models ….
Government, or gov’t. leader, is a distinct decision making agent
•
Institutions whereby leader is selected are spelled out.
•
Selection rules generally discipline actions the leader can take.
•
Contrast: rent-seeking models characterize outcomes as
equilibria of effort expended by private agents.
Lindbeck-Weibull ‘probabilistic’ voting framework
•
Specifies very dispersed distribution of political power
•
Can be generalized to represent systems in which power is held
by powerful, organized interests.
35
Rent-induced Regime Transitions
Historical accounts: competition to control resource wealth can erupt
in violence
•
Examples: civil wars in Angola, Nigeria, Sierra Leone, Zaire,
fought to control diamonds, oil and metallic minerals.
•
Criterion for controlling government shifts away from an electoral
process, toward violent struggle.
36
Rent-induced Regime Transitions ….
Aslaksen and Torvik, Scand. J. Econ., 2006 combine political
economy and rent-seeking models to examine transitions.
•
Candidates seek office to control a resource rent.
•
Adapt Lindbeck-Weibull voting model to characterize outcomes
under democracy.
•
Characterize welfare of election’s loser as function of resource
rent , preference parameters, …
•
Adapt rent-seeking model to characterize each candidate’s
expected welfare under violent competition.
•
If election loser’s expected welfare under democracy is less than
loser’s expected welfare under violent competition, violence
erupts.
37
Rent-induced Regime Transitions ….
General setup
•
Rival candidates (politicians, factions, parties) play repeated game.
•
Reward is resource rent controlled by government, net of any
transfers to supporters in democratic regimes and ‘fighting’ costs in
non-democratic regimes
•
Strategies depend on history of play;. Once violence erupts, it
continues indefinitely.
•
Focus on (‘best’) Nash equilibria when agents play Nash reversion
(trigger) strategies.
•
Acquiescing to democracy = cooperating;
Resorting to violence = defecting.
38
Rent-induced Regime Transitions ….
Objectives
•
Characterize conditions under which democracy can persist
•
Determine how resource rent affects the viability of sustained
democracy.
39
Rent-induced Regime Transitions …
Probabilistic voting:
•
Citizens get utility from income, which depends on government
policies, and from candidates’ attributes
•
Citizen i’s utility if candidate B wins is:
Bi  WBi   i  
Utility from
income (policy)
•
Ideological
preference for B
relative to A
Relative popularity
of B versus A
i
Let W A be citizen i’s utility from income if A wins. Then i votes
for candidate B iff
 i  W Ai  WBi  
40
Rent-induced Regime Transitions …
Role of non-income based preferences:
, ideological preference for B vs. A, is uniformly distributed with
density f.
, relative popularity of B vs. A, is uniformly dist. with density y.
•
Candidates know density of , but value is not realized
until after election.
•
The smaller is y the more willing voters are to trade off
•
ideology for money.
Candidates will exploit this and capture greater shares of
rent in equilibrium when
y is small.
41
Rent-induced Regime Transitions …
Candidate A’s objective function:
max Pr( N A  12 )  ( R  A' s Transfers )
Vote share for A,
depends on A’s
transfers.
Resource rent,
captured by
winner
Transfers promised
to supporters
B’s objective function is symmetric.
Given symmetry, both candidates choose the same policy (Hotelling)
and outcome is determined by realization of popularity term, .
42
Rent-induced Regime Transitions …
Equilibrium rent retained by winning candidate is:
Citizen’s
wage income
w R
~
XI 
2y  1
The more important is the non-monetary component of
preference, i.e., the smaller is y, the greater is the rent retained
by winner.
Winner’s rent also depends positively on R.
43
Rent-induced Regime Transitions …
Candidates’ payoffs under perpetual democracy:
VI 
1
w R

1  b 2(2y  1)
Discount
factor
Loser’s expected PV payoff under continued democracy:
b
w R
V 

1  b 2(2y  1)
I
44
Rent-induced Regime Transitions …
In conflict regime, candidates compete by spending effort
fighting. Prize is the resource rent, R. Opportunity cost of effort is
foregone consumption. Initiating conflict has a fixed cost, F units
of effort.
Once initiated, conflict lasts forever. Expected PV payoff from
perpetual conflict is:
1
1 b
R


wF
 4

45
Rent-induced Regime Transitions …
An outcome under which each candidate continues to submit to elections,
rather than initiating violence, is feasible only if
b
w R
1

1  b 2(2y  1) 1  b

R


wF
4



This cannot be satisfied (democracy cannot be sustained) unless
b  y  12
Future cannot be discounted too heavily
Non-monetary preferences must be sufficiently large (small y).
(If non-monetary preferences are important, candidates capture large
share of rent under democracy.)
46
Rent-induced Regime Transitions …
Empirical implications?
•
Transition from democracy to conflict is more likely when R is large,
or after a resource rent windfall arrives.
•
Democracy more likely to be sustained when ideology is important,
future not discounted too heavily.
Possibility of transition from conflict to democracy?
47
Public Employment at a Commitment Mechanism
Robinson, Torvik and Verdier, J Dev. Econ., 2006.
In political systems where politicians need popular support to gain
office, candidates and citizen-supporters face commitment problems.
•
Candidates’ promises not fulfilled until after election. What
guarantee promises will be kept?
•
Citizens’ support, at individual level, cannot be verified by
candidate.
48
Public Employment at a Commitment Mechanism
Public employment as a political commitment mechanism:
•
Incumbent offers jobs to like-minded supporters; public jobs pay
more than private sector jobs.
•
Incumbent finds it costly to fire public employees after election,
so she cannot renege; if challenger wins, he can fire employees
hired by incumbent.
•
Workers know they will benefit if incumbent wins, so they carry
through with promised support.
49
Public Employment, …. so, what about the RC?
Candidates seek to gain office, in order to control a resource rent:
•
Rent can be taken as private income, or spent on public
employment to enhance probability of winning election.
•
Greater rent, i.e., a windfall, increases desire to retain office,
and expands public employment.
Public employment is less productive than private sector employment:
•
Hiring more public employees to retain office lowers income.
•
Connecting the dots, higher resource rent leads to lower
income—a resource curse.
•
RC mechanism is political competition to stay in office.
50
Public Employment, ….
Empirical implications:
•
A resource windfall will reduce income, RC.
•
A resource windfall will expand public employment, the RC
mechanism.
•
A politician’s benevolence toward his own group has the
following effects:
-It increases public sector employment, raising the
incumbent’s probability of winning.
-It leads to a more severe RC as a consequence.
-It diminishes incentive to over-extract in pre-election period
(because probability of staying in office is greater)
51
Conclusions, Implications for testing, etc.
Why some countries grow while others stagnate is arguably the most
important question in economics.
Several political economy models predict the ‘resource curse’:
•
So, verifying this link between resource windfalls and slow
growth doesn’t provide useful information at this point.
•
Need to test all implications of a political economy model, in
attempt to falsify.
52
Conclusions, Implications for testing, etc.
Opportunities for empirical research:
1.
Identifying appropriate institutional controls:
• Rent-seeking activity is endogenous, inappropriate.
• Susceptibility to rent-seeking is more appropriate:
- Possibly depends on cultural, religious, historical
factors, fractionalization, etc.
1.
Research designs that examine windfalls and exploit timing,
as well as untreated observations, are relatively convincing.
•
Windfalls, e.g., price changes, can be identified in time
and measured easily.
•
Resource abundance measures create ambiguities r.e.
timing, reflect sectoral composition—are less convincing.
53
Conclusions, Implications for testing, etc.
3.
Diversion of entrepreneurship deserves more careful attention:
• Prominent in some historical and case study accounts.
• May be observable in data on sectoral shifts following
windfalls.
4.
Public employment argument deserves closer attention.
5.
•
Should be easily tested.
•
Seems consistent with casual empiricism.
•
Benevolence effect possibly testable?
Increased likelihood of transition from democracy to conflict
following resource windfall.
54
Conclusions, Implications for testing, etc.
6.
Reduction in resource extraction capital following windfall: my
interpretation of Tornell and Lane’s formal capital.
7.
Specify identity of ‘groups’ in political economy models in ways
that are consistent with theory.
•
Players in non-cooperative games.
•
A group’s members adhere to a common strategy.
•
A group acts independently from other groups.
55
That’s it---finally!
Thanks for your patience!
56