lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/lecture7-8_090527.pdf · lectures...
TRANSCRIPT
![Page 1: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/1.jpg)
Lectures 7-8_090520.pdf
Political Economics IIISpring 2009
IV. State Capacity, Conflict and DevelopmentLectures 7-8
Torsten Persson, IIES
1
![Page 2: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/2.jpg)
Discuss roles of state building and conflict in development
develop basic framework and extend ina couple of directions
new theory and clear hints at how to approach the data
A. The origins of state capacity
B. State capacity and the genius of taxation
C. State capacity and the strategy of conflict
2
![Page 3: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/3.jpg)
A. The origins of state capacityMotivations — Figure 1
State capacity in political and economic history
power to raise revenues important for militarysuccess and state development, more generally
Ignored by economists, yet implicit in economic theory
state capacity to raise (any) taxes basically assumedin public finance, political economics, development...
as is capacity to enforce contracts (support markets)
Empirical observations of development
over wide range of countries (financial) market developmentand tax capacity positively correlated with each other and GDP
3
![Page 4: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/4.jpg)
010
2030
40S
hare
of I
ncom
e T
axes
in G
DP
0 50 100 150 200private credit to GDP
Above median income Below median incomeFitted values
Figure 1
![Page 5: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/5.jpg)
Purpose of analysis
Study investment in two forms of state capacity
fiscal capacity: constrains tax policy, which raisesrevenue for common-interest spending or redistribution
legal capacity; constrains regulatory policy, which protectsproperty rights (in potentially discriminating way)
What are the main determinants of state capacity?
model which economic, political and institutional featurespromote investments, under uncertainty, in state capacity
4
![Page 6: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/6.jpg)
Central result
Complementarity between fiscal and legal capacity
investments in the two types of state capacityare mutually reinforcing
we should indeed see market development andtaxation go hand in hand
complementarity has a number of — analytical,substantial and empirical — implications
5
![Page 7: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/7.jpg)
Related work
Economic and political history
body of work on fiscal capacity and its relation to militarycapacity and the fighting of wars
but no links to legal capacity (support of markets)
Law, politics, and financial development
recent work on legal, or political, origins of financial developmentbut no links to taxation and fiscal development: "why dopowerful groups maldesign financial system to protect rentsrather than support markets and redistribute to themselves?"
6
![Page 8: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/8.jpg)
Road map
1. A basic model
2. Equilibrium policy
3. Equilibrium investments in state capacity
4. A preliminary look at the data
5. Final remarks
7
![Page 9: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/9.jpg)
1. A basic model
Main building blocks
Private sector
earns income in private markets and consumes
Government
regulates (legally protects capital markets)taxes and spends on transfers or public goodsinvests in legal and fiscal capacity
8
![Page 10: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/10.jpg)
Private groups and types
Two time periods, s ∈ {1, 2}
Two groups of individuals, J ∈ {A,B}each has share 12 of populationgroups defined by some observable featureevery member has income wJ
s , no savingsutility is linear in consumption (risk neutrality)
Incomes
depend on regulation pJs
wJs = w(pJs)
think about "legal protection of property rights"microfoundations: credit market model with partialenforcement of collateralized debt contracts
9
![Page 11: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/11.jpg)
Regulation
Government has discretion over current regulation
policy pJs can be group-specificconstrained by existing legal capacity, i.e., pJs ≤ πs .
Investment in legal capacity
courts, judges, credit and property registries.initial stock π1 is given, but can be augmented in s = 1by (positive) investment π2 − π1, at (convex) cost L(π2 − π1),where Lπ(0) = 0
10
![Page 12: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/12.jpg)
Taxation
Government has discretion over current taxation
government taxes output at rates tJs
these can be negative, i.e., redistributive transferspolicy can be group-specificconstrained by existing fiscal capacity, i.e., tJs ≤ τsmicrofoundations: an individual can earn a share(1− τs) of her income in the informal sector)
Investment in fiscal capacity
tax authority, infrastructure to enforce income taxor impose a value added tax
initial stock is given τ1 but can be augmented in s = 1by (positive) investment τ2 − τ1, at (convex) cost F (τ2 − τ1),where Fτ (0) = 0
11
![Page 13: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/13.jpg)
Public goods
Common-interest (non-redistributive) motive for taxation
public goods have same linear payoff αsGs to everybodythink about as “threat of external conflict” and “defense”
αs has two-point distribution αs ∈ {αL, αH ]}, where αH > 2αL < 1, and Prob[αs = αH] = φ
shocks to α iid over timerealization of αs known when {tJs} set
12
![Page 14: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/14.jpg)
Private utility and government budget
Utility of each group
linear payoffs allow us to write
vJ = αsGs + (1− tJs)w(pJs)
Government budgetsXJs
tJsw(pJs)
2+R1 = G1 + [L(π2 − π1) + F (τ2 − τ1)]
in period 1, and XJs
tJsw(pJs)
2+R2 = G2
in period 2
Rs is additional revenue source, Rs ∈ {RL,RH}, whereProb[Rs = RH] = ρ — interpret as natural resource rents.
13
![Page 15: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/15.jpg)
Government decision weights and turnover
Representative governments
at policy stage in each s one group Is holdsthe government, the other is the opposition Os
Decision weights in policy
the incumbent Is gives weights 1− θ to itselfand θ to Os, where θ ≤ 1
2
Government turnover
γ ∈ [0, 1] is probability of transition of powerin s = 2 from I1 to O1
Interpret crudely as political institutions
θ inclusiveness, minority protectionγ political instability
14
![Page 16: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/16.jpg)
Timing in period s
1. Start out with state capacity {πs, τs}and incumbent group Is−1
2. Nature determines αs, and Rs
3. Group Is−1 remains in power withprobability 1− γ
4. New incumbent Is chooses policy {tIs, tOs, pIs, pOs,Gs}and invests in legal and fiscal capacity (only at s = 1)s t government budget constraint and state capacity
5. Payoffs are realized and agents consume
15
![Page 17: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/17.jpg)
2. Equilibrium policy
Can study policy and investments separately
incumbent government maximizes
V Is = (1− θ)w³pIs´(1− tIs) + θw
³pOs´(1− tOs) +
αs
⎡⎢⎣tIsw³pIs´+ tOsw
³pOs´
2+ zs
⎤⎥⎦ ,
where we have eliminated Gs and defined
zs = Rs −½L(π2 − π1)− F (τ2 − τ1) if s = 1
0 if s = 2 .
s t constraints Gs ≥ 0, tJs ≤ τs and pJs ≤ πs
16
![Page 18: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/18.jpg)
Regulation (legal protection) — Proposition 1(a)
For s ∈ {1, 2} any government Js, and any αs and Rs,optimal regulation fully utilizes all legal capacity, pIs = pOs = πs
“obvious” result, cf. Diamond-Mirrlees production efficiencyresult breaks down if there are rents — explain in B
does not imply property rights well protected everywhere
17
![Page 19: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/19.jpg)
Taxation and public goods — Proposition 1 (b)
Common-interest state, αs = αH ≥ 2(1− θ)
where tIs = tOs = τs, and Gs = τsw(πs) + zsvalue of public goods worth more than transfers to incumbentall available revenue used or public goods
Redistributive state, αs = αL < 2(1− θ)
where tOs = τs,Gs = 0,−tIs = τsw(πs)+2zsw(πs)
all available revenue transferred to incumbent group
In more general model: undersupply of public goods
if α continuous, public goods provided in too few states ofworld, compared to Utilitarian optimum with θ = 1
2where condition is αs ≥ 1
18
![Page 20: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/20.jpg)
3. Equilibrium investments in state capacity
Objective
Incumbent I1 maximizes expected utility
expected weighted payoff at s = 2,under uncertainty about α2, R2 and I2,less investments costs at s = 1
first-order conditions are
wp(π2){1 + τ2[(E(λ2)− 1]} 0 λ1Lπ (π2 − π1) (1)c.s. π2 − π1 > 0
w(π2)[(E(λ2)− 1] 0 λ1Fτ (τ2 − τ1) (2)c.s. τ2 − τ1 > 0
19
![Page 21: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/21.jpg)
Key public finance parameters
E(λ2) = φαH + (1− φ)2[(1− θ − γ(1− 2θ)] .expected value of future public funds, and
λ1 = Max [α1, 2(1− θ)]
realized period-1 value of public funds
20
![Page 22: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/22.jpg)
Central condition
E(λ2)− 1 ≥ 0holds if common-interest state probable enough,or political instability is low enough
Sufficient condition for positive investmentsof both types — assume this holds throughout A
Sufficient condition for complementarity
higher π raises incentives to invest in τ and v.v.moreover, payoff function is supermodular: we canexploit results on monotone comparative statics
simple to derive effects of most parameter shifts
21
![Page 23: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/23.jpg)
Economic development/real wages — Proposition 2 (a)
Countries with higher real wages choosegreater investment in fiscal and legal capacity.
upward multiplicative shift of w(·) raises w(π2) and wp(π2)for given π2, for given E(R2) this implies higher GDP w(π2)+R2
consistent with empirical observation (Figure 1)
22
![Page 24: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/24.jpg)
Share of resource rents in GDP — Proposition 2 (b)
Countries with higher share of non-resource income choosegreater investment in fiscal and legal capacity.
for given GDP, higher w(·) raises the non-resource share inGDP w(π2)/(w(π2) +R2)
prospective explanation for weak African states
23
![Page 25: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/25.jpg)
Value of public goods — Proposition 2 (c)
Higher expected demand for public goods raisesinvestment in state capacity.
∂E(λ2)∂φ = αH − 2 [1− θ − γ (1− 2θ)] > 0
common interests make fiscal capacity more valuableprediction: external conflict does promote fiscal capacityas argued in economic and political history literature
by complementarity, auxiliary prediction for legal capacity
24
![Page 26: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/26.jpg)
Political instability and inclusiveness — Proposition 2 (d)
Investment in fiscal and legal capacity are promoted by lowerpolitical instability, and by more inclusive political institutionsif political instability is high enough
∂E(λ2)∂γ = − (1− φ) 2(1− 2θ) < 0 and ∂E(λ2)
∂θ = (1− φ) 2(2γ − 1)
the ruling group sees that fiscal capacity will be usedto redistribute away from its own group less often
additional prediction: interaction term — political stabilitymore important, the more non-inclusive are political institutions
minority protection good for investment, when this lowers theprospective losses from redistribution by the opponent group
prediction: more consensual political institutions, e.g.,parliamentary democracy, may raise investments in state capacity
25
![Page 27: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/27.jpg)
Costs of investments — Proposition 2 (e)
Lower costs of either legal or fiscal capacity raisesinvestments in legal and fiscal capacity
downward multiplicative shift of L(·) or F (·) cutsRHS of (1) or (2) for given π2 and τ2
theoretical rationale for "legal origins" hypothesis,but with auxiliary prediction for fiscal capacity
26
![Page 28: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/28.jpg)
4. ImplicationsGrowth and development
"Endogenous" growth
due to investments in economic institutionsY2 − Y1
Y1=w(π2)− w(π1) +R2 −R1
w(π1) +R1
growth due to institutional improvement leading tomore efficient private markets
due to complementarity, government size is growing tooempirical results: hard find negative corr. taxation — growth,easy find positive corr. financial development — growth
Can add private accumulation of capital
additional positive effect on growth of legal capacity,but standard negative effect on growth of fiscal capacity
27
![Page 29: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/29.jpg)
A first look at the data
General prediction about determinants
determinants of legal and fiscal capacity should be common
Specific predictions about determinants
positive effect of demand for common-interest public goodsincidence of external wars in past (COW),1816/independence—1975
positive effect of more consensual political institutionsincidence of democracy, parl. democracy in the past (Polity IVand Persson-Tabellini), 1800/independence—1975
different effects of legal origins (Shleifer et al) — may expectBritish vs. French legal origin to affect both legal and fiscalcapacity in the same directions
28
![Page 30: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/30.jpg)
Legal capacity — Table 1
Alternative outcome measures
private credit to GDP (as in King-Levine), avg 1975—contract enforcement (Doing business project), circa 2005government anti-diversion policy (ICRG), avg 1982-97
Consider correlations with suggested determinants
29
![Page 31: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/31.jpg)
Table 1 Economic and Political Determinants of State Capacity across Countries
Legal capacity
(1) (2) (3) Private credit
to GDP (1975-2000)
Contract enforcement (circa 2005)
GADP (1982-1997)
Incidence External Conflict up to 1975
0.509*** (0.143)
1.010*** (0.277)
0.576*** (0.170)
Incidence Democracy up to 1975
0.095 (0.059)
0.044 (0.078)
0.125** (0.050)
Incidence parliamentary Democracy up to 1975
0.000 (0.062)
0.040 (0.087)
0.111* (0.061)
English Legal Origin – 0.008
(0.033) 0.126** (0.059)
– 0.007 (0.040)
Socialist Legal Origin — 0.185**
(0.079) 0.096*** (0.034)
German Legal Origin 0.406***
(0.120) 0.466*** (0.062)
0.248*** (0.053)
Scandinavian Legal Origin
0.112*** (0.041)
0.551*** (0.081)
0.254*** (0.055)
Observations 93 122 115 Adjusted R-squared 0.524 0.235 0.596
Robust standard errors in parentheses: * significant at 10%; ** significant at 5%; *** significant at 1%. Socialist legal origin dropped in Col 1, since data on private credit not available in that category.
![Page 32: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/32.jpg)
Fiscal capacity — Table 1 continued
Alternative outcome measures
[1 — informal sector GDP share] (World Bank), circa 2005income taxes as share of total taxes (IMF), avg 1975—total taxes as share of GDP (ditto)
Consider correlations with suggested determinants
30
![Page 33: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/33.jpg)
Table 1 Economic and Political Determinants of State Capacity across Countries
Legal capacity Fiscal capacity
(1) (2) (3) (4) (5) (6) Private credit
to GDP (1975-2000)
Contract enforcement (circa 2005)
GADP (1982-1997)
1 – share of informal sector (circa 2005)
Income taxes in total taxes (1975-2000)
Total taxes in GDP
(1975-2000)
Incidence External Conflict up to 1975
0.509*** (0.143)
1.010*** (0.277)
0.576*** (0.170)
0.478*** (0.137)
0.405* (0.221)
0.555*** (0.162)
Incidence Democracy up to 1975
0.095 (0.059)
0.044 (0.078)
0.125** (0.050)
0.128* (0.072)
0.072 (0.083)
0.088 (0.059)
Incidence parliamentary Democracy up to 1975
0.000 (0.062)
0.040 (0.087)
0.111* (0.061)
0.051 (0.075)
0.157* (0.089)
0.160** (0.067)
English Legal Origin – 0.008
(0.033) 0.126** (0.059)
– 0.007 (0.040)
– 0.011 (0.053)
0.002 (0.047)
– 0.015 (0.042)
Socialist Legal Origin — 0.185**
(0.079) 0.096*** (0.034)
0.159*** (0.054)
– 0.130*** (0.032)
– 0.119** (0.031)
German Legal Origin 0.406***
(0.120) 0.466*** (0.062)
0.248*** (0.053)
0.234*** (0.046
0.206** (0.092)
0.010 (0.083)
Scandinavian Legal Origin
0.112*** (0.041)
0.551*** (0.081)
0.254*** (0.055)
0.127*** (0.052)
0.098 (0.098)
0.292*** (0.087)
Observations 93 122 115 87 103 104 Adjusted R-squared 0.524 0.235 0.596 0.383 0.371 0.639
Robust standard errors in parentheses: * significant at 10%; ** significant at 5%; *** significant at 1%. Socialist legal origin dropped in Col 1, since data on private credit not available in that category.
![Page 34: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/34.jpg)
4. Final remarks
Theoretical framework
useful point of departure to think about development ofstate capacity
central insight: legal and fiscal capacity complements
Can extend theory in several directions
e.g., introduce income inequality — drives legal and fiscalcapacity in different directions
look at interaction with private accumulation — way tounderstand persistent effects of institutions on development?
Serious empirical work remains
31
![Page 35: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/35.jpg)
B. State capacity and the genius of taxation
Growth of government
Expansion of government activity and taxation
one of most salient facts of 20th centurysubject and puzzle in several literaturesincluding modern work on political economics
state capacity approach provides new angle
32
![Page 36: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/36.jpg)
Debate on large government
Is large government costly?
benevolent government (left-wing) view; growth ofgovernment reflects that government does things well
malevolent government (right-wing) view: growth reflectsabuse of power, rent seeking and so on
Both views, politically naive
left-wing: fails to acknowledge important role of(vested) interests in policymaking
right-wing: fails to acknowledge that suppressing suchinterests is not feasible in political context
33
![Page 37: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/37.jpg)
Main idea
Evidence of effects of big government mixed
hard to find evidence for large effects on incomeboth in macro and micro data
Illustrate why this may be the case
the positive economics of Diamond and Mirrleesdevelop simple example where large governmentmay not be so costly as simple extension of state capacityframework from part A
34
![Page 38: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/38.jpg)
Road map
1. A simple two factor economy
2. Equilibrium policy and state capacity
3. Implications
35
![Page 39: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/39.jpg)
1. A simple two-factor economyAlternative production structure
Two factors of production
w(pJs) now group-specific “capital”,productivity depends on legal protection
raw labor, traded at economy-wide market
Ownership of factors
a share, σ, of each group holds 1 unit of capital,remaining share, 1− σ, holds 1 unit of labor
Technology
well-behaved Cobb-Duoglas technology on intensive form lJs(kJs)η,
where kJs = w(pJs)lJs
is the capital-labor ratio
36
![Page 40: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/40.jpg)
Equilibrium wages
Economy-wide capital-labor ratio
k(pIs, pOs) =σ[w(pIs) + w(pOs)]
(1− σ)2
Factor market equilibrium
derive equilibrium wage ω(pIs, pOs) = (1− η)k(pIs, pOs)η
with
∂ω
∂pJs= (1− η) η(k(pIs, pOs))η−1
σwp
³pJs´
2(1− σ)> 0
legal protection of any group raises common wageas more productive capital raises labor demand
37
![Page 41: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/41.jpg)
2. Equilibrium policy and state capacity
Analyze policy in modified model
yJs(pIs, pOs) = (1− σ)ω(pIs, pOs) + σlJs[(kJs)η − ω(pIs, pOs)]
income of each group depends on legal protectionof both groups, through equilibrium wage
replace w(pJs) by yJs(pIs, pOs)
Main consequences
if σ large, Is may prefer low ω(pIs, pOs) to raise rentscan keep down wage by cutting pOs, blockinglegal protection for other group’s capital owners
Go through similar steps as in A
same tax policy, but potentially different legal protection38
![Page 42: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/42.jpg)
Efficiency or not — Proposition 3
If θ = 12 or τs = 1, legal capacity is always fully utilized for
both groups: pOs = pIs = πs.There exists bτ (θ, α), such that legalprotection of opposition minimal, pOs = 0, for all τ < bτ (θ, α),and where bτθ(θ, α) < 0, and bτα(θ, α) < 0 if α = αH.
always efficient legal protection in utilitarian case,or when fiscal capacity is high enough
but with low fiscal capacity, politically motivated Ismay prefer to redistribute by distorting marketsrather than by taxes and transfers at maximum income
critical bound bτ(θ, α) depends on institutions andcircumstances, decreasing in θ and αH
39
![Page 43: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/43.jpg)
Endogenize state capacity
Low fiscal capacity may trigger inefficient redistribution
argument known e.g., from Acemoglu (2006)
However, argument is incomplete
it takes τs and πs as given — why τs so low?key to low investments already in Proposition 2
Is there a possibility of a non-investment trap?
i.e., can inefficient policies persist over time if statecapacity is chosen in same way as in A. above
40
![Page 44: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/44.jpg)
Non-investment trap — Proposition 4
Suppose that τ1 < τ̂ (θ, αL) .Then, for φ close to 0, there is a rangeof γ > 1/2, where τ2 = τ1 and investment in legal capacity π2 islower than it would be if τ1 > τ̂ (θ, αL)
if τ2 = τ1 < bτ opposition gets minimal legal protection,pOs = 0, in each period, whichever group holds power
incumbent in period 1 fears redistribution via taxes and doesnot invest in fiscal capacity to bring economy out of this trap
investment in legal capacity also lower (by complementarity)
41
![Page 45: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/45.jpg)
3. ImplicationsLarge vs. small states
Two economies, S and L, where Proposition 4 applies
τS1 < bτ < τL1 and πS1 = πL1 = π1
non-resource GDP Y (pIs, pOs) = [yIs(pIs, pOs) + yOs(pIs, pOs)]/2
Compare income levels
Y L1 − Y S
1 = Y (π1, π1)− Y (π1, 0) > 0
Y L2 − Y S
2 = Y (πL2 , πL2 )− Y (πS2 , 0) > Y L
1 − Y S1
large state has higher GDP in both periods, andincome gap to small state grows over time (as πL2 > πS2 )
in A, low fiscal capacity (and taxation) related to low income(co-determined by other factors, or low income causes small state)here, small state may actually cause low income
42
![Page 46: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/46.jpg)
Political incentives to distort markets
E.g., literature on the politics of financial development
important to not only consider incentives forcreating or preserving rents
some results on the politics of underdevelopedmarkets implicitly assume limited fiscal capacity
have to pose “political Coase-theorem question” explicitlyanalysis here shows how that can be done
43
![Page 47: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/47.jpg)
Argument more general than simple example
Might similar mechanism explain other production distortions?
take e.g., tariff protection or red-tape regulationsalient instances of policy-induced production distortions
casual evidence suggest such distortions are common in developingcountries with weak states
similar joint determination due to political instability and otherfactors hampering investments in state capacity
44
![Page 48: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/48.jpg)
C. State capacity and the strategy of conflictBasic facts about civil war — Figure 2
Internal conflict widespread phenomenon
civil war has plagued many nations in postwar periodprevalence over all countries and years since 1950 is 7%,cumulated death toll exceeds 15 million
Big fact # 1
prevalence varies greatly over years,peaks in early 1900s above 12%
Big fact # 2
prevalence varies greatly over countries,civil war and poverty (low GDP/capita) strongly correlated
45
![Page 49: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/49.jpg)
Figure 2 Incidence of Civil War
![Page 50: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/50.jpg)
Repression — violation of human rights
Government repression also widespread phenomenon
many governments use violent means to raise their probabilityof staying in power without civil war breaking out
such one-sided repression shows up in violations of humanrights: executions, political murders, imprisonments, brutality, ...
Prevalence
by strict measure (purges) 6% of country-years, 1962-2005by wider measure (human-rights infringements) 32%, 1976-2006
46
![Page 51: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/51.jpg)
Main ideas — Table 2
Domestic violence and development jointly determined
same economic and political factors may explainprevalence of civil war and low income:
e.g., two parallel literatures on the "resource curse"
How does civil war interact with state capacity?
by analysis in A, external conflict risk may promote state capacity,by enhancing common, rather than redistributive, interests
domestic violence: manifestation of stark redistributive conflictrisk of domestic conflict may have opposite effects on incentivesto invest in state capacity, compared to risk of external conflict
simple correlations consistent with that idea
47
![Page 52: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/52.jpg)
Table 2: Fiscal capacity and different types of war
(1)
(2)
(3)
(4)
(5)
(6)
Taxes as a percentage of
GDP
Taxes as a percentage of
GDP
Income taxes as a percentage of total taxes
Income taxes as a percentage of total taxes
Average percentage inflation rate
Average percentage inflation rate
Any internal war – 6.993*** – 2.723** – 7.953** – 3.094 14.346*** 12.551** during 1945-1997 (1.843) (1.310) (3.960) (3.574) (4.790) (5.567) Any external war 6.911*** 1.604 16.157*** 6.653** 3.586 6.676 during 1945-1997 (2.267) (1.569) (3.713) (3.317) (5.863) (6.628) Share of years in 5.028** 1.902 – 3.979 democracy during 1945-1997 (2.368) (7.020) (11.157) Share of years in parliamentary 6.529*** 7.553 – 5.403 democracy during 1945-1997 (1.939) (6.401) (4.395) Mean (log of) income during 3.302*** 7.936*** 0.974 1945-1997 (1.067) (2.231) (3.934) Observations 125 105 125 105 116 100 R-squared
0.489
0.739 0.390 0.598 0.307 0.370
Robust standard errors in parentheses (* significant at 10%; ** significant at 5%; *** significant at 1%) In addition to the variables shown, all specifications include a set of dummies for (eight) regions and (five) legal origins.
![Page 53: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/53.jpg)
How approach these issues?
Step-wise approach
set up model allowing for conflict and peacestudy the risk of conflict (domestic violence) as function ofeconomic, political. institutional determinants, for givenlevels of state capacity — this will be a bit of a detour
revisit state capacity decisions, for different risks of (endogenous)domestic conflict
Benefits of explicit theory
guidance on how to approach the data
48
![Page 54: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/54.jpg)
Road map
1. Conflict and violent takeover
2. From theory to evidence
3. Investments in state capacity revisited
4. Summing up: A-C
49
![Page 55: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/55.jpg)
1. Conflict and violent takeover
Existing literature — theory of conflict
Many classical models of conflict and contests
what happens when parties commit resources to fighting?Grossman, Skaperdas, etc.
Not closely integrated with institutions and data
little concern for forces of economic developmentlack of tangible institutionsnot well integrated with observations and evidence
50
![Page 56: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/56.jpg)
Existing literature — evidence on civil war
Substantial literature in political science and economics
determinants of onset and duration of civil war
Suggested explanations for concentration to poor countries
Collier-Hoeffler — low opportunity cost of fightingFearon-Laitin — low state capacity to counter insurgencyboth arguments take income as given
Ongoing debate about other proposed determinants
natural resources, ethnic tensions, institutions...
Methodology
not closely integrated with theorywith a few exceptions, based on cross-sectional inference
51
![Page 57: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/57.jpg)
Existing literature — violation of human rights
Little explicit theory
Some empirical work
more violations of human rights in poor and undemocratic countriesfew attempts to connect incidence of repression and civil war
52
![Page 58: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/58.jpg)
Desirable features of theory
Use simple “canonical model” of internal conflict
illuminate interplay between economic, politicaland institutional factors — connect theory and evidence
To what use?
interpret known factsstay simple enough to allow further extensionsin particular, endogenize state capacity
generate testable predictions over observables
Which testable predictions?
make precise in which way resource rents, wages and politicalinstitutions may drive conflict
uncover a set of common determinants for civil war and repression53
![Page 59: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/59.jpg)
Alternative model for transfer of power
Both incumbent and opposition group can mount armies
δIs−1 ∈ {0, AI
s} where 0 < AIs < 1 is fixed number,
subject to political decision, financed out of public purse
δOs−1 ∈ {0, AO
s }, 0 < AOs < 1 decided and paid within opposition
continuous, rather than discrete, army choices easy to introduce
No conscription
soldiers just paid period s, wage ws — first take as given
Probability that group Os−1 becomes Is
γ(δOs−1, δIs−1), which is increasing in 1st argumentand decreasing in 2nd argument
54
![Page 60: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/60.jpg)
Two changes of the model in A.
New form of government budget constraint
0 ≤X
Js∈{Is,Os}
tJsws
2−Gs + zs − wsδ
Is−1 .
Stage 3 in the timing described in A. is replaced by
3 a. Group Os−1 chooses δOs−1 whether to mount insurgency AO
s .
3 b. Incumbent Is−1 chooses whether to use violence δIs−1
= AIs.
3 c. Group Is−1 remains in office with probability 1− γ³δOs−1, δIs−1
´.
55
![Page 61: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/61.jpg)
Policy outcomes and objectives in common-interest states
Policy decisions at stage 4
for incumbent Is, policy decisions arethe same as in A. (given available budget)
Suppose αs = αH > 2
objectives of Is−1 and Os−1 at stage 3,given optimal policy at stage 5
E[V Is−1(πs, τs) | αs = αH ] =
αH [τsw (πs) + zs − w (πs) δIs−1] + w (πs) (1− τs)
and
E[V Os−1 | αs = αH ] = αH [τsw (πs) + zs − w (πs) δIs−1]
+w (πs) (1− τs − δOs−1) .
56
![Page 62: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/62.jpg)
Peace in common-interest state - Proposition 5
Common-interest states, αs = αH, are always peacefulwith δOs−1 = δIs−1 = 0
objectives are strictly decreasing in δOs−1 and δIs−1
because all residual spending is on public goods,there is nothing to fight over
this will not be the case in redistributive states
57
![Page 63: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/63.jpg)
Objectives in redistributive states
When αs = αL < 1 < 2 (1− θ)
objectives of Is−1 and Os−1 at stage 3, given optimalpolicy at stage 5, now involve clear trade-offs betweenresources and probability of controlling policy
E[V Is−1(πs, τs) | αs = αL] = w (πs) (1− τs) +h(1− θ)− γ
³δOs−1, δIs−1
´(1− 2θ)
i·
2[τsw (πs) + zs − w (πs) δIs−1]
E[V Os−1 | αs = αL] = w (πs) (1− τs − δOs−1) +
γ³δOs−1, δIs−1
´2[τsw (πs) + zs − w (πs) δ
Is−1]
58
![Page 64: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/64.jpg)
Preliminaries
Assumptions(a) The technology for conflict is: γ
³δO, δI
´= μ
hδO − δI
i+ γO
(b) μAIs ≤ γO ≤ 1− μAO
s — probability always between 0 and 1(c) 12 + γO − μAI
s ≥ 1−θ1−2θ — see below
Define government revenue relative to non-resource GDP
Zs =τsw (πs) + zs
w (πs)
and two bounds for this variable
Zs =(1− θ)− γO (1− 2θ)
(1− 2θ)μ +AIs and Zs =
1
2μ+AI
s
with (c) Zs > Zs
59
![Page 65: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/65.jpg)
Potential conflict in redistributive states — Proposition 6
Redistributive states, αs = αL, have three possible outcomes
1. If Zs < Zs, outcome is peaceful, bδOs−1= bδIs−1 = 0.
2. If Zs ∈£Zs, Zs
¤, no insurgency bδOs−1
= 0, but incumbent
uses repression against opposition with bδIs−1 = AIs:
3. If Zs > Zs, civil war where opposition and incumbentmount armies AO
s and AIs
proof uses the reaction functions of two parties
60
![Page 66: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/66.jpg)
Anatomy of three regimes
Peace: Zs < Zs
when w(πs) high, Rs low, or θ near 12too expensive to fight (opportunity cost argument),or not enough to fight over because pie, or sharef it, small
Repression: Zs ∈£Zs, Zs
¤resource rents higher or wages lower, so more at stakeAssumption (c) means government has lower threshold
Civil war: Zs > Zs
even more at stake, so both parties choose to fight
61
![Page 67: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/67.jpg)
2. From theory to evidenceConditions for observing violent outcomes
Consider country c at year s
treat wages wc,s as time-varying but given, andtreat investment in state capacity, Lc as constant and given
condition for civil war at year s is
Zc,s = τc,s +Rc,s − Lc
wc,s> Zc,s =
1
2μc+AI
c,s
while condition for repression is
Zc,s > Zc,s > Zc,s =(1− θc)
(1− 2θc)− γOc
μc+AI
c,s
62
![Page 68: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/68.jpg)
What is observable in the data?
Observable determinants of violence
can measure (or find decent proxies for) Rc,t, wc,t and θc
Unobservable determinants
or very hard to measure μc, γOc , Lc, A
Ic,s − τc,s, and αc,s
Unobserved randomness in Zc,s − Zc,s and Zc,s − Zc,s
i.e., random realizations of conditions for violencedefine
εc,s = AIc,s − τc,s
with c.d.f. Xc(ε)
Incidence of violence
do observe if there is civil war, or repression, in c at t
63
![Page 69: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/69.jpg)
Lesson #1 for an econometrician
Conditional probability of observing civil war
in country c at date s is
(1− φc) ·XcµRc,s − Lc
wc,s− 1
2μc
¶where 1st term is the probability of observing αc,s = αLand 2nd is the probability of observing Zc,s − Zc,s > 0
Basic prediction
higher Rc,s or lower wc,s raises probability of observing civil warin richer model, Xc increasing in Rc,t and decreasing in wc,t
only as long as θc below some, unobserved upper bound θc < 12
can test this with time-varying measures of Rc,s and wc,s (and θc)
64
![Page 70: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/70.jpg)
Lesson #2 for an econometrician
Conditional probability of observing repression
but not civil war, in country c at date t
(1−φc)·"Xc
ÃRc,s − Lc
wc,s− (1− θc)
(1− 2θc)+γOcμc
!−Xc
µRc,s − Lc
wc,s− 1
2μc
¶
Basic prediction
higher Rc,t, lower wc,t, and lower θc raise probabilityof observing repression as well as civil war
these two states are ordered in the three determinantsby Proposition 6
can test this by estimating an ordered logit (or probit)
65
![Page 71: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/71.jpg)
#1. Incidence of Civil War?Preliminaries
Clean inference from unobserved determinants
estimate panel regressions with fixed country effectsis equivalent to estimating
(1− φc) ·hXc³Rc,s−Lcwc,s
− 12μc
´−E{Xc
³Rc,s−Lcwc,s
− 12μc
´}i
Heterogeneity in incidence of civil war over time
mainly driven by time variation in Rc,t and wc,t
rid of influence by unobserved factors (like μc, γOc , and Lc)
if add fixed year effects, we allow for world-wide,non-parametric trend in civil war (recall Figure 2), andexploit only country-specific time variation in Rc,t and wc,t
66
![Page 72: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/72.jpg)
Use variation in commodity prices
Source of exogenous variation
world market prices not likely to be influenced bycivil war in a single country
Can build simple open-economy model
increased export prices raise resource rents Rc,tand hence gain from winning a civil conflict
increased import prices decrease wages wc,t(and thereby indirectly raise resource rents), whichcuts the cost (and boosts the gain) of civil conflict
thus both increased export prices and increasedimport prices raise the critical difference Zc,s − Zc,s
67
![Page 73: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/73.jpg)
Econometric specification
First, estimate linear panel regression
civc,t = ac + at + byc,t + κc,t
where ac and at are country and year fixed effectsand yc,t a vector of time-varying regressors, includingexport and import prices for primary commodities
Test for heterogeneity with regard to θc
estimate in different samples defined by politicalinstitutions in place
Linearity vs. non-linearity
fraction of countries in civil war is relatively low,which may bias linear probability estimates
also estimate by conditional (fixed effect) logit
68
![Page 74: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/74.jpg)
Data
Civil war
binary indicator(s), Corrrelates of War (alt. Uppsala/PRIO)
Independent variables
country-specific commodity export-price indexes, world pricesof about 45 commodities — shocks toRc,t — UNCTAD, NBER/UN
commodity import price indexes — shocks to wc,t — same sourcesdisaster count (floods & heat-waves) — shocks to wc,t — EM-DAT5-year lag GDP/capita, although some risk of simultaneityin model, what matters is share of resource rents in GDP
Split sample by indicators
parliamentary democracy, or high executive constraints — thecondition for θc R θc — Polity IV and Persson-Tabellini
69
![Page 75: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/75.jpg)
Basic results — Table 3
Effects of export and import prices?
both raise the probability of civil war,except oil export prices
Quantitatively significant?
yes! one standard deviation increase in price indexesraise the probability of civil war by 7-15%
Evidence for heterogeneity?
yes! no (or opposite) effects in parliamentary democracy
Other regressors?
more disasters do raise probability of observing civil war
70
![Page 76: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/76.jpg)
Table 3: Within-country determinants of civil war – Basic results
(1) (2) (3) (5) Civil war in year Civil war in year Civil war in year Civil war in year
Export price index 0.035*** 0.039*** -0.046** 0.094*** (0.013) (0.014) (0.022) (0.031) Import price index 0.293*** 0.193*** -1.675*** 0.586*** (0.083) (0.070) (0.373) (0.205) Oil Export Prices -0.001 -0.001 0.015 -0.001 (0.002) (0.002) (0.050) (0.002) Oil Import Prices -0.025*** -0.020*** -0.101 0.056*** (0.006) (0.005) (0.105) (0.022) 5-year lag, log(GDP per cap) -0.046*** -0.106*** -0.076** -0.037* (0.014) (0.015) (0.032) (0.020) Weathershock 0.013** (0.006) Sample All Non-parliamentary
democracies Parliamentary democracies
All
Observations 4648 3531 1117 3814 Number of Countries 124 102 49 117 R-squared 0.045 0.036 0.105 0.055
Notes: Robust standard errors in parentheses; * significant at 10%; ** significant at 5%; *** significant at 1%. All specifications include fixed country and year effects
![Page 77: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/77.jpg)
Robust to functional form — Table 4
Conditional logit results
qualitative results hold upwhen re-estimate OLS specification on restricted sampleearlier quantitative results hold up, as well
71
![Page 78: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/78.jpg)
Table 4: Within-country determinants of civil war – Robustness to specification and sample
(1) Civil war in year
(2) Civil war in year
(3) Civil war in year
(4) Civil war in year
Export price index 1.802*** 0.382*** 2.103*** 0.408*** (0.626) (0.099) (0.671) (0.107) Import price index 9.685*** 1.709*** 7.569** 1.376*** (3.546) (0.357) (3.746) (0.537) Oil Export Prices -0.019 -0.002 -0.015 -0.002 (0.013) (0.002) (0.013) (0.003) Oil Import Prices 3.092 0.382 1.843 0.051 (2.064) (0.257) (2.823) (0.344) 5-year lag, log(GDP per cap) 0.953** 0.018 0.505 -0.005 (0.423) (0.060) (0.469) (0.070) Weathershock 0.060 0.008 0.032 0.001 (0.094) (0.014) (0.117) (0.015) Estimation method Conditional logit OLS Conditional logit OLS Sample Civil war countries
Civil war countries Civil war non-parl.
democracies Civil war non-parl.
democracies Observations 1282 1282 1067 1067 Number of countries 38 38 34 34 R-squared 0.124 0.092
Notes: Standard errors in parentheses; * significant at 10%; ** significant at 5%; *** significant at 1%. All specifications include fixed country and year effects
![Page 79: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/79.jpg)
Robust to measurement — Table 5
Alternative data sources
replace COWmeasure with UCDP/PRIO measure ofcivil war incidence or # battle deaths (intensity of conflict)— point estimates maintain their sign, though much less precise
split sample by high constraints on executive, rather thanparliamentary democracy — results hold up
replace incidence of civil war by onset of civil war— results do not hold up
72
![Page 80: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/80.jpg)
Table 5: Within-country determinants of civil war – Alternative measurement
(1) (2) (3) (4) (5) Civil War
Incidence (UCDP/PRIO)
Civil War Onset
(UCDP/PRIO)
Battle Deaths in thousands
(UCDP/PRIO)
Civil War Incidence (COW)
Civil War Incidence (COW)
Export price index 0.024 -0.014 0.576** 0.101*** -0.049* (0.033) (0.010) (0.311) (0.034) (0.029) Import price index 0.112 0.035 7.594 0.760*** -1.541*** (0.190) (0.053) (5.022) (0.221) (0.331) Oil Export Prices -0.001 -0.000 0.010 -0.001 -0.049 (0.001) (0.001) (0.022) (0.002) (0.045) Oil Import Prices 0.018* -0.002 -0.682 0.044** 0.010 (0.010) (0.004) (2.659) (0.021) (0.089) 5-year lag, log (GDP/cap) -0.001 -0.002 1.095*** - 0.060*** -0.003 (0.016) (0.007) (0.334) (0.023) (0.040) Weathershock 0.021*** -0.001 - 0.131 0.004 0.013 (0.005) (0.002) (0.967) (0.007) (0.010) Sample All All All Low Exec.
constraints High Exec. constraints
Observations Countries
4559 118
4559 118
2599 83
2797 98
1017 56
R-squared 0.047 0.011 0.043 0.046 0.088
Notes: Robust standard errors in brackets; * significant at 10%; ** significant at 5%; *** significant at 1%. All columns include fixed country and year effects
![Page 81: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/81.jpg)
#2 Repression or Civil War?
Preliminaries
Recall Lesson # 2 for an econometrician
higher Rc,t, lower wc,t, and lower θc raise the probabilityof observing civil war or repression, rather than peace
repression and civil war are ordered statesgauge by estimating ordered logit
A few specifications
cross-sectional specifications, including fixed year effectshard to include fixed country effects
73
![Page 82: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/82.jpg)
Two repression measures
Gibney et al (2007) — annual data, 1976-2006
from US State department and Amnesty Internationalpolitical terror scale 1− 5: set binary indicator 1 if 3 or above
Banks (2005) — annual data, 1962-2005
more restrictive measure on number of purges:set binary indicator 1 whenever purges positive
Construct ordered dependent variable
combine each repression measure with Correlates of War datapeace = 0, repression/but not civil war = 1, civil war = 2
Basic correlations
GDP/capita: peace — $6,500, repression — $3,200, civil war — $2,000in parliam. dem: peace — 35%, repression — 16%, civil war — 9%
74
![Page 83: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/83.jpg)
Basic results — Table 6
Effect of key determinants
indicators of resource rents, income and parliamentary democracytend to raise probability of repression and civil war
especially so for commodity prices and the Banks measure
Note on reporting of results
coefficients reported as odds ratios: number above (below) unitycorresponds to a positive (negative) coefficient
significance levels refers to “H0: not different from unity”
75
![Page 84: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/84.jpg)
Table 6: Economic and Political Determinants of Repression and Civil War
(1) (2) (3) (4) (5) (6)
Log GDP 0.668*** 0.652*** 0.660*** 0.631*** 0.630*** 0.626*** (7.85) (7.56) (6.81) (8.37) (8.24) (7.97) Parliamentary Democracy 0.401*** 0.345*** 0.316*** 0.578*** 0.554*** 0.580*** (6.98) (7.84) (7.88) (3.36) (3.72) (3.39) Large Oil Exporter 1.102 1.269 1.081 1.200 1.314* 1.205 (0.63) (1.43) (0.41) (1.13) (1.67) (1.06) Large Primary Exporter 0.644*** 0.572*** 0.377*** 0.284*** 0.284*** 0.195*** (3.97) (4.66) (6.80) (7.30) (7.30) (7.26) Weathershock 1.186*** 1.420*** 1.399*** 1.124*** 1.250*** 1.275*** (3.88) (8.32) (7.66) (2.78) (4.69) (4.93) Commodity Export Price Index 1.106*** 1.172*** (3.24) (3.83) Commodity Import Price Index 0.206** 1.413 (2.52) (0.82) Oil Export Prices 1.008 1.030*** (0.46) (3.33) Oil Import Prices 1.394*** 1.198*** (7.68) (2.59) Year Dummy Variables No Yes Yes No Yes Yes Observations 1993 1993 1878 3549 3549 3394
Notes: In columns 1-3, the dependent variable is constructed from the COW and the repression data in Gibney et al (2007). In columns 4-6, the dependent variable is instead constructed using the purges data in Banks (2005). All columns are estimated using an ordered logit. The reported coefficients are odds ratios with robust z-statistics in parentheses: the significance levels refer to differences from unity (* significant at 10%; ** significant at 5%; *** significant at 1%).
![Page 85: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/85.jpg)
3. Investments in state capacity revisited
Preliminaries
How does civil-war risk affect investment incentives?
when civil-war risk absent, A.3 applies with γO = γ
for simplicity, consider the situation when 1st periodincumbent knows for sure that α2 = αL
Remaining uncertainties
future resource rents and R2 and political control I2consider increased risk of conflict triggered by changesin ρ, making high resource rents more likely
76
![Page 86: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/86.jpg)
Two new effects on state capacity investments
a. Effects via discount factor for 1st-period incumbent
by Prop. 6, probability of turnover in period 2 is
Γ (Z (R2;π2, τ2)) =
⎧⎪⎨⎪⎩γO + μ
hAO −AI
iif Z (R2;π2, τ2) > Z
γO − μAO if Z (R2;π2, τ2) ∈£Z,Z
¤γO if Z (R2;π2, τ2) < Z
depends exogenously on R2, and endogenously on π2 and τ2instability smallest with repression, larger or smallerwith conflict than with peace depending on AO R AI
b. Effects via the costs of conflict
incumbent has to pay the real market wage to future armiesmay be reluctant to raise incomes by investing in legal capacity
77
![Page 87: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/87.jpg)
Case 1 — cycles between peace and civil war
Effect of high and low resource rents
Z (RH ;π2, τ2) > Z > Z > Z (RL;π2, τ2) i.e., both opposition andincumbent arm if R2 = RH, while nobody arms if R2 = RL
investments in τ2 or π2 never large enough to change the regime
Payoff to period-1 incumbent
E[V Is(π2, τ2) | αs = αL] = w (π2) (1− τ2)
+E (λ2) [τ2w (π2) + E (z2)]
−ρh1− θ − (
³μ³AO −AI
´+ γO)
´(1− 2θ)
i2w (π2)A
I,
where E (λ2) =h1− θ +
³μρ³AO −AI
´+ γO
´(1− 2θ)
i2
and term on third line is expected cost of incumbent’s armyas in A., assume E (λ2) > 1 so investments are complements
78
![Page 88: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/88.jpg)
Case 1 — optimality conditions for state capacity choices
First order conditions
wp(π2)h{1 + τ2[E(λ2)− 1]}− ρ
h1− θ +
³μ³AI −AO
´− γO
´(1− 2θ
≤ λ1Lπ(π2 − π1)
w(π2)[E(λ2)− 1] ≤ λ1Fτ (τ2 − τ1) .
Assume that AI ≈ AO
conflict has no 1st-order effect on political turnover, soa higher ρ does not alter discount factor and hence not E (λ2)
Only one impact effect of higher conflict risk
marginal benefit of legal capacity falls, larger expected military costcf. second term on LHS of first condition
by complementarity, both types of investment fall
79
![Page 89: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/89.jpg)
Wasteful military spending and state capacity —Proposition 7
Suppose we have α2 = αL, conflict or peace depending on R2,and AI ≈ AO. Then, a higher ρ raises the probability of conflictand reduces the incentive to invest in both fiscal and legal capacity
if AI ≈ AO conflict is Pareto inferior: resources spent withoutmaterial change in who holds power
by the proposition, this static inefficiency is compounded by adynamic inefficiency: a lower incentive to invest in state capacity
80
![Page 90: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/90.jpg)
Case 2 — cycles between repression and civil war
Effect of high and low resource rents
Z (RH ;π2, τ2) > Z > Z (RL;π2, τ2) > Z i.e., both opposition andincumbent arm if R2 = RH, only incumbent arms if R2 = RL
Payoff to period-1 incumbent
E[V Is(π2, τ2) | αs = αL] = w (π2) (1− τ2)
+E (λ2)hτ2w (π2) +E (z2)− w (π2)A
Ii,.
where E (λ2) =h1− θ −
³μ³ρAO −AI
´+ γO
´(1− 2θ)
i2
again assume E (λ2) > 1
81
![Page 91: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/91.jpg)
Case 2 — optimality conditions for state capacity choices
First order conditions
wp(π2)h{1 + τ2[E(λ2)− 1]}−
h1− θ −
³μ³ρAO −AI
´+ γO
´(1− 2θ)
≤ λ1Lπ(π2 − π1)
w(π2)[E(λ2)− 1] ≤ λ1Fτ (τ2 − τ1) .
Conflict does have 1st-order effect on political turnover
even if AI ≈ AO, so higher ρ lowers discount factor and E (λ2)
Effects on marginal benefit of both types of state capacity
higher ρ decreases the LHS of both conditions, for legalcapacity because ∂LHS
∂ρ = −2 (1− 2θ)μAO(1−AI) < 0
by complementarity, both types of investment fall
82
![Page 92: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/92.jpg)
Conflict-induced instability and state capacity —Proposition 8
Suppose we have α2 = αL, and conflict or repression depending on R2.Then, a higher ρ raises the probability of conflict and reducesthe incentive to invest in both fiscal and legal capacity
the effect on the discount factor always outweighs effect onexpected cost of fighting
analogous to Prop 2 (d), but higher instability now equilibriumoutcome, where conflict makes incumbent survival less likely
83
![Page 93: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/93.jpg)
Implications
Caution for empirical studies of civil war
hazardous to consider a correlation between poverty and civilwar as causal effect from poverty to conflict.
two leading explanations: low opportunity cost of fightingdue to low wages, and low state capacity in poor countries
low state capacity, as well as low wages, may be simultaneouslydetermined with a high probability of civil war
Risk of different types of conflict
risk of internal and external conflict indeed appear to entail verydifferent incentives for state building
More generally
risk of conflict, low state capacity and low incomesmay be closely intertwined
84
![Page 94: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/94.jpg)
4. Summing up: A-C
Common themes
Political institutions and levels of development
may shape static policies at a point in time,but also dynamic state development
Public revenue and public goods preferences
realized and expected shocks shape current policies,as well as incentives to build the state
85
![Page 95: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/95.jpg)
What have we learned?
State capacity
major concept in other fieldshave shown a way to bring it into mainstream economics
low state capacity viewed as major problem in thedevelopment community
A low investment in legal capacity slows growthB low fiscal capacity may cause low income, via distortionsC low legal capacity may cause civil war, via low incomes
highlighted basic complementarity between the twoa way to think about the clustering of institutions
86
![Page 96: Lectures 7-8 090520perseus.iies.su.se/~tpers/courses/politec2009/Lecture7-8_090527.pdf · Lectures 7-8_090520.pdf Political Economics III Spring 2009 IV.StateCapacity,Conflict and](https://reader030.vdocuments.mx/reader030/viewer/2022041101/5eda8b35ca56d138585eb2d6/html5/thumbnails/96.jpg)
What next?
Conflict
have contrasted common-interest and redistributive statessuggests that risks of external and internal conflicts shapevery different incentives to build the state
Development
theory helps approach the data, but models very simple
try to better understand long-run development:add in private accumulation and full-fledged dynamics
try to endogenize political institutions: may revealadditional complementarity in the development process
87