Why do Autocrats Disclose?
James R. Hollyer B. Peter Rosendorff James Raymond Vreeland
Minnesota/Princeton
NYU
Georgetown
December 14, 2015
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
Question
When and why do autocratic governments disclose information(particularly economic information) to their publics?
Note: We will use the terms disclosure and transparency interchangeably
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project
1 Intro2 The HRV Index of Transparency
I apply an item response model to missing data from WDI, 125countries, 1980-2010
I Hollyer, Rosendorff & Vreeland (2014). Political Analysis.
3 Transparency and Investment
I lowers uncertainty for risk averse investors, signals pro-investmentpolicy
I increases in capital formation, fixed capital formation and FDI
4 Democracies Disclose More
I improves welfare, enables efficient economic decision-makingI democracy induces leaders to care more about citizen welfareI Hollyer, Rosendorff & Vreeland (2011). JOP
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project
1 Intro2 The HRV Index of Transparency
I apply an item response model to missing data from WDI, 125countries, 1980-2010
I Hollyer, Rosendorff & Vreeland (2014). Political Analysis.
3 Transparency and Investment
I lowers uncertainty for risk averse investors, signals pro-investmentpolicy
I increases in capital formation, fixed capital formation and FDI
4 Democracies Disclose More
I improves welfare, enables efficient economic decision-makingI democracy induces leaders to care more about citizen welfareI Hollyer, Rosendorff & Vreeland (2011). JOP
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project
1 Intro2 The HRV Index of Transparency
I apply an item response model to missing data from WDI, 125countries, 1980-2010
I Hollyer, Rosendorff & Vreeland (2014). Political Analysis.
3 Transparency and Investment
I lowers uncertainty for risk averse investors, signals pro-investmentpolicy
I increases in capital formation, fixed capital formation and FDI
4 Democracies Disclose More
I improves welfare, enables efficient economic decision-makingI democracy induces leaders to care more about citizen welfareI Hollyer, Rosendorff & Vreeland (2011). JOP
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project
1 Intro2 The HRV Index of Transparency
I apply an item response model to missing data from WDI, 125countries, 1980-2010
I Hollyer, Rosendorff & Vreeland (2014). Political Analysis.
3 Transparency and InvestmentI lowers uncertainty for risk averse investors, signals pro-investment
policyI increases in capital formation, fixed capital formation and FDI
4 Democracies Disclose More
I improves welfare, enables efficient economic decision-makingI democracy induces leaders to care more about citizen welfareI Hollyer, Rosendorff & Vreeland (2011). JOP
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project
1 Intro2 The HRV Index of Transparency
I apply an item response model to missing data from WDI, 125countries, 1980-2010
I Hollyer, Rosendorff & Vreeland (2014). Political Analysis.
3 Transparency and InvestmentI lowers uncertainty for risk averse investors, signals pro-investment
policyI increases in capital formation, fixed capital formation and FDI
4 Democracies Disclose More
I improves welfare, enables efficient economic decision-makingI democracy induces leaders to care more about citizen welfareI Hollyer, Rosendorff & Vreeland (2011). JOP
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project
1 Intro2 The HRV Index of Transparency
I apply an item response model to missing data from WDI, 125countries, 1980-2010
I Hollyer, Rosendorff & Vreeland (2014). Political Analysis.
3 Transparency and InvestmentI lowers uncertainty for risk averse investors, signals pro-investment
policyI increases in capital formation, fixed capital formation and FDI
4 Democracies Disclose MoreI improves welfare, enables efficient economic decision-makingI democracy induces leaders to care more about citizen welfareI Hollyer, Rosendorff & Vreeland (2011). JOP
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project, cnt’d
1 Intro
2 The HRV Index of Transparency
3 Transparency and Investment
4 Democracies Disclose More
5 Transparency and Regime Transition: Theory
I in aut.: facilitates common belief formation and coordination of protestI in dem: improves functioning of elections, reduces incentives to
overthrow regime
6 Transparency and Regime Transition: Empirics
I increases hazard of dem. transition, reduces hazard of dem. collapseI Hollyer, Rosendorff & Vreeland (2015) APSR; (2015) ‘Transparency,
Protest and Democratic Stability’
7 Why Do Autocrats Disclose?
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project, cnt’d
1 Intro
2 The HRV Index of Transparency
3 Transparency and Investment
4 Democracies Disclose More5 Transparency and Regime Transition: Theory
I in aut.: facilitates common belief formation and coordination of protestI in dem: improves functioning of elections, reduces incentives to
overthrow regime
6 Transparency and Regime Transition: Empirics
I increases hazard of dem. transition, reduces hazard of dem. collapseI Hollyer, Rosendorff & Vreeland (2015) APSR; (2015) ‘Transparency,
Protest and Democratic Stability’
7 Why Do Autocrats Disclose?
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project, cnt’d
1 Intro
2 The HRV Index of Transparency
3 Transparency and Investment
4 Democracies Disclose More5 Transparency and Regime Transition: Theory
I in aut.: facilitates common belief formation and coordination of protestI in dem: improves functioning of elections, reduces incentives to
overthrow regime
6 Transparency and Regime Transition: Empirics
I increases hazard of dem. transition, reduces hazard of dem. collapseI Hollyer, Rosendorff & Vreeland (2015) APSR; (2015) ‘Transparency,
Protest and Democratic Stability’
7 Why Do Autocrats Disclose?
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project, cnt’d
1 Intro
2 The HRV Index of Transparency
3 Transparency and Investment
4 Democracies Disclose More5 Transparency and Regime Transition: Theory
I in aut.: facilitates common belief formation and coordination of protestI in dem: improves functioning of elections, reduces incentives to
overthrow regime
6 Transparency and Regime Transition: EmpiricsI increases hazard of dem. transition, reduces hazard of dem. collapseI Hollyer, Rosendorff & Vreeland (2015) APSR; (2015) ‘Transparency,
Protest and Democratic Stability’
7 Why Do Autocrats Disclose?
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project, cnt’d
1 Intro
2 The HRV Index of Transparency
3 Transparency and Investment
4 Democracies Disclose More5 Transparency and Regime Transition: Theory
I in aut.: facilitates common belief formation and coordination of protestI in dem: improves functioning of elections, reduces incentives to
overthrow regime
6 Transparency and Regime Transition: EmpiricsI increases hazard of dem. transition, reduces hazard of dem. collapseI Hollyer, Rosendorff & Vreeland (2015) APSR; (2015) ‘Transparency,
Protest and Democratic Stability’
7 Why Do Autocrats Disclose?
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
The Larger (Book) Project, cnt’d
1 Intro
2 The HRV Index of Transparency
3 Transparency and Investment
4 Democracies Disclose More5 Transparency and Regime Transition: Theory
I in aut.: facilitates common belief formation and coordination of protestI in dem: improves functioning of elections, reduces incentives to
overthrow regime
6 Transparency and Regime Transition: EmpiricsI increases hazard of dem. transition, reduces hazard of dem. collapseI Hollyer, Rosendorff & Vreeland (2015) APSR; (2015) ‘Transparency,
Protest and Democratic Stability’
7 Why Do Autocrats Disclose?
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
FindingsStylized Facts:
transparent autocratic regimes more prone to collapse via mass unrestor dem’ization
but, are less prone to collapse due to coups
Argument: Autocratic leaders disclose because it insulates them fromthreats that emerge from within their regimes
in part, this is because transparency facilitates mass mobilization
in part, because transparency encourages FDI, increasing rents fromelite membership
Leaders disclose when threats from within regime are high relative to thosefrom populace
institutionalized regimes – competing power bases, designatedsuccession mechanisms
leaders new to office
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
FindingsStylized Facts:
transparent autocratic regimes more prone to collapse via mass unrestor dem’ization
but, are less prone to collapse due to coups
Argument: Autocratic leaders disclose because it insulates them fromthreats that emerge from within their regimes
in part, this is because transparency facilitates mass mobilization
in part, because transparency encourages FDI, increasing rents fromelite membership
Leaders disclose when threats from within regime are high relative to thosefrom populace
institutionalized regimes – competing power bases, designatedsuccession mechanisms
leaders new to office
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
FindingsStylized Facts:
transparent autocratic regimes more prone to collapse via mass unrestor dem’ization
but, are less prone to collapse due to coups
Argument: Autocratic leaders disclose because it insulates them fromthreats that emerge from within their regimes
in part, this is because transparency facilitates mass mobilization
in part, because transparency encourages FDI, increasing rents fromelite membership
Leaders disclose when threats from within regime are high relative to thosefrom populace
institutionalized regimes – competing power bases, designatedsuccession mechanisms
leaders new to office
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
FindingsStylized Facts:
transparent autocratic regimes more prone to collapse via mass unrestor dem’ization
but, are less prone to collapse due to coups
Argument: Autocratic leaders disclose because it insulates them fromthreats that emerge from within their regimes
in part, this is because transparency facilitates mass mobilization
in part, because transparency encourages FDI, increasing rents fromelite membership
Leaders disclose when threats from within regime are high relative to thosefrom populace
institutionalized regimes – competing power bases, designatedsuccession mechanisms
leaders new to office
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
FindingsStylized Facts:
transparent autocratic regimes more prone to collapse via mass unrestor dem’ization
but, are less prone to collapse due to coups
Argument: Autocratic leaders disclose because it insulates them fromthreats that emerge from within their regimes
in part, this is because transparency facilitates mass mobilization
in part, because transparency encourages FDI, increasing rents fromelite membership
Leaders disclose when threats from within regime are high relative to thosefrom populace
institutionalized regimes – competing power bases, designatedsuccession mechanisms
leaders new to office
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
FindingsStylized Facts:
transparent autocratic regimes more prone to collapse via mass unrestor dem’ization
but, are less prone to collapse due to coups
Argument: Autocratic leaders disclose because it insulates them fromthreats that emerge from within their regimes
in part, this is because transparency facilitates mass mobilization
in part, because transparency encourages FDI, increasing rents fromelite membership
Leaders disclose when threats from within regime are high relative to thosefrom populace
institutionalized regimes – competing power bases, designatedsuccession mechanisms
leaders new to office
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
FindingsStylized Facts:
transparent autocratic regimes more prone to collapse via mass unrestor dem’ization
but, are less prone to collapse due to coups
Argument: Autocratic leaders disclose because it insulates them fromthreats that emerge from within their regimes
in part, this is because transparency facilitates mass mobilization
in part, because transparency encourages FDI, increasing rents fromelite membership
Leaders disclose when threats from within regime are high relative to thosefrom populace
institutionalized regimes – competing power bases, designatedsuccession mechanisms
leaders new to office
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
FindingsStylized Facts:
transparent autocratic regimes more prone to collapse via mass unrestor dem’ization
but, are less prone to collapse due to coups
Argument: Autocratic leaders disclose because it insulates them fromthreats that emerge from within their regimes
in part, this is because transparency facilitates mass mobilization
in part, because transparency encourages FDI, increasing rents fromelite membership
Leaders disclose when threats from within regime are high relative to thosefrom populace
institutionalized regimes – competing power bases, designatedsuccession mechanisms
leaders new to officeHRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
Findings, cnt’d
Demonstrate that in autocracies:
1 disclosure more frequent in institutionalized regimes and less frequentin personalistic ones
2 leaders disclose more readily when new to office
3 transparency associated with increased net FDI inflows
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
Findings, cnt’d
Demonstrate that in autocracies:
1 disclosure more frequent in institutionalized regimes and less frequentin personalistic ones
2 leaders disclose more readily when new to office
3 transparency associated with increased net FDI inflows
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
Findings, cnt’d
Demonstrate that in autocracies:
1 disclosure more frequent in institutionalized regimes and less frequentin personalistic ones
2 leaders disclose more readily when new to office
3 transparency associated with increased net FDI inflows
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Qustion and Preview
Findings, cnt’d
Demonstrate that in autocracies:
1 disclosure more frequent in institutionalized regimes and less frequentin personalistic ones
2 leaders disclose more readily when new to office
3 transparency associated with increased net FDI inflows
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Motivation
Theoretical Priors
Autocratic leaders face two threats to rule:
1 displacement by regime members (e.g., coup)
2 displacement of regime – including the leader (e.g., by mass unrest)
Steps by elites to replace leader increase regime instability
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Motivation
Theoretical Priors
Autocratic leaders face two threats to rule:
1 displacement by regime members (e.g., coup)
2 displacement of regime – including the leader (e.g., by mass unrest)
Steps by elites to replace leader increase regime instability
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Motivation
Pacifying the Regime
Leaders may have an incentive to insulate themselves from accountabilityto elite:
Two Mechanisms:
1 increase rents flowing to regime members
2 increase external threat
Transparency achieves both ends
1 ease mobilization by members of the public
2 encourage investment (particularly FDI)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Motivation
Pacifying the Regime
Leaders may have an incentive to insulate themselves from accountabilityto elite:
Two Mechanisms:
1 increase rents flowing to regime members
2 increase external threat
Transparency achieves both ends
1 ease mobilization by members of the public
2 encourage investment (particularly FDI)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Motivation
Pacifying the Regime
Leaders may have an incentive to insulate themselves from accountabilityto elite:
Two Mechanisms:
1 increase rents flowing to regime members
2 increase external threat
Transparency achieves both ends
1 ease mobilization by members of the public
2 encourage investment (particularly FDI)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Motivation
Pacifying the Regime
Leaders may have an incentive to insulate themselves from accountabilityto elite:
Two Mechanisms:
1 increase rents flowing to regime members
2 increase external threat
Transparency achieves both ends
1 ease mobilization by members of the public
2 encourage investment (particularly FDI)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Motivation
Pacifying the Regime
Leaders may have an incentive to insulate themselves from accountabilityto elite:
Two Mechanisms:
1 increase rents flowing to regime members
2 increase external threat
Transparency achieves both ends
1 ease mobilization by members of the public
2 encourage investment (particularly FDI)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Question and Motivation Motivation
Pacifying the Regime
Leaders may have an incentive to insulate themselves from accountabilityto elite:
Two Mechanisms:
1 increase rents flowing to regime members
2 increase external threat
Transparency achieves both ends
1 ease mobilization by members of the public
2 encourage investment (particularly FDI)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Model Primitives
Actors: an autocratic leader LRegime Elites R and an Opposition O
Actions: L chooses d ∈ {0, 1}and a policy variable et ∈ {0, 1}R chooses v ∈ {0, 1}
Typespace: L is of type θ ∈ {0, 1}θ = 1 denotes a ‘convergent’ typeθ = 0 denotes a ‘divergent’ typePr(θ = 1) = π
State Space: st ∈ {0, 1}, Pr(st = 1) = 12
Timing: t ∈ {1, 2}
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Effects of DisclosurePrimitive assumption that disclose increases O’s mobilizational capacity.
Denote pr. O displaces R absent leader removal as p(d)
p(1) > p(0)
ρ ≡ p(1)− p(0)
ω ∈ (1, 1p(1)) represents effect of regime discord on stability
risk of regime collapse given by ωp(d) following removal of Lω declines with institutionalization, rises with leader’s time in office
Primitive assumption that disclosure increases investment:.
Denote potential pool of rents as y(d):
y(1) > y(0)
ψ ≡ y(1)− y(0)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Effects of DisclosurePrimitive assumption that disclose increases O’s mobilizational capacity.
Denote pr. O displaces R absent leader removal as p(d)
p(1) > p(0)
ρ ≡ p(1)− p(0)
ω ∈ (1, 1p(1)) represents effect of regime discord on stability
risk of regime collapse given by ωp(d) following removal of Lω declines with institutionalization, rises with leader’s time in office
Primitive assumption that disclosure increases investment:.
Denote potential pool of rents as y(d):
y(1) > y(0)
ψ ≡ y(1)− y(0)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Effects of DisclosurePrimitive assumption that disclose increases O’s mobilizational capacity.
Denote pr. O displaces R absent leader removal as p(d)
p(1) > p(0)
ρ ≡ p(1)− p(0)
ω ∈ (1, 1p(1)) represents effect of regime discord on stability
risk of regime collapse given by ωp(d) following removal of L
ω declines with institutionalization, rises with leader’s time in office
Primitive assumption that disclosure increases investment:.
Denote potential pool of rents as y(d):
y(1) > y(0)
ψ ≡ y(1)− y(0)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Effects of DisclosurePrimitive assumption that disclose increases O’s mobilizational capacity.
Denote pr. O displaces R absent leader removal as p(d)
p(1) > p(0)
ρ ≡ p(1)− p(0)
ω ∈ (1, 1p(1)) represents effect of regime discord on stability
risk of regime collapse given by ωp(d) following removal of Lω declines with institutionalization, rises with leader’s time in office
Primitive assumption that disclosure increases investment:.
Denote potential pool of rents as y(d):
y(1) > y(0)
ψ ≡ y(1)− y(0)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Effects of DisclosurePrimitive assumption that disclose increases O’s mobilizational capacity.
Denote pr. O displaces R absent leader removal as p(d)
p(1) > p(0)
ρ ≡ p(1)− p(0)
ω ∈ (1, 1p(1)) represents effect of regime discord on stability
risk of regime collapse given by ωp(d) following removal of Lω declines with institutionalization, rises with leader’s time in office
Primitive assumption that disclosure increases investment:.
Denote potential pool of rents as y(d):
y(1) > y(0)
ψ ≡ y(1)− y(0)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Effects of DisclosurePrimitive assumption that disclose increases O’s mobilizational capacity.
Denote pr. O displaces R absent leader removal as p(d)
p(1) > p(0)
ρ ≡ p(1)− p(0)
ω ∈ (1, 1p(1)) represents effect of regime discord on stability
risk of regime collapse given by ωp(d) following removal of Lω declines with institutionalization, rises with leader’s time in office
Primitive assumption that disclosure increases investment:.
Denote potential pool of rents as y(d):
y(1) > y(0)
ψ ≡ y(1)− y(0)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Utilities of Citizens
uR,t(et , st , d) =
{It [∆ + λy(d)] + (1− It)(1− λ)y(d) if et = stItλy(d) + (1− It)(1− λ)y(d) otherwise
uO,t(et , st , d) =
{It(1− λ)y(d) + (1− It)[∆ + λy(d)] if et = stIt(1− λ)y(d) + (1− It)λy(d) otherwise.
where
It =
{1 if R is in power0 otherwise.
and λ ∈ (12 , 1), ∆ > 0.
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Utilities of Leaders
uL,t(et , st , d ; θ) =
∆ + λy(d) if et = st and in powerλy(d) if et 6= st , θ = 1 and in powerrt + λy(d) if et 6= st , θ = 0 and in power0 if out of power.
Where rt is drawn from cdf G (.), and G (∆) = 0. E [rt ] = µ.
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Primitives
Game Form1 Nature draws the the leader’s type θ ∈ {0, 1}, the state variable s1
and the value of rents r1, which are revealed to the leader but not toany citizen.
2 The leader chooses d ∈ {0, 1} and the value of e13 R observes the choice of d and the realization of the policy outcome.
It chooses v ∈ {0, 1}.4 A contest for power between R and O takes place. O prevails with
probability p(d) if the leader was previously retained and withprobability ωp(d) if the leader was previously removed.
5 a If O prevails, it is in power in round 2 and a new leader is chosen byNature. This leader is of type θ = 1 with probability π.
b If R prevails after ousting the leader, a new leader is chosen by Nature.This leader is of type θ = 1 with probability π.
c Otherwise, L remains in office.6 Nature chooses values of s2 and r2.7 The sitting leader chooses e2. All payoffs are realized and the game
ends.HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Equilibrium
Equilibrium Concept
We characterize a perfect Bayesian equlibrium to this game
which satisfies the intuitive criterion (Cho and Kreps, 1987)
and in which R anticipates convergent types act on their primitiveinterests over disclosure
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Equilibrium
Equilibrium Concept
We characterize a perfect Bayesian equlibrium to this game
which satisfies the intuitive criterion (Cho and Kreps, 1987)
and in which R anticipates convergent types act on their primitiveinterests over disclosure
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Equilibrium
Equilibrium Concept
We characterize a perfect Bayesian equlibrium to this game
which satisfies the intuitive criterion (Cho and Kreps, 1987)
and in which R anticipates convergent types act on their primitiveinterests over disclosure
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Equilibrium
Stability Thresholds
R has a dominant strategy to set v = 0 if π∆ < p(d)y(d)(ω−1)(2λ−1)1−ωp(d)
Implicitly define ω̄ and ω s.t.:
π∆ =p(0)y(0)(ω̄ − 1)(2λ− 1)
1− ω̄p(0)
π∆ =p(1)y(1)(ω − 1)(2λ− 1)
1− ωp(1).
if ω > ω̄ no internal threat to leader
if ω < ω always an internal threat to leader
if ω ∈ [ω, ω̄] a threat absent disclose, but no threat given disclosure
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Equilibrium
Stability Thresholds
R has a dominant strategy to set v = 0 if π∆ < p(d)y(d)(ω−1)(2λ−1)1−ωp(d)
Implicitly define ω̄ and ω s.t.:
π∆ =p(0)y(0)(ω̄ − 1)(2λ− 1)
1− ω̄p(0)
π∆ =p(1)y(1)(ω − 1)(2λ− 1)
1− ωp(1).
if ω > ω̄ no internal threat to leader
if ω < ω always an internal threat to leader
if ω ∈ [ω, ω̄] a threat absent disclose, but no threat given disclosure
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Equilibrium
Stability Thresholds
R has a dominant strategy to set v = 0 if π∆ < p(d)y(d)(ω−1)(2λ−1)1−ωp(d)
Implicitly define ω̄ and ω s.t.:
π∆ =p(0)y(0)(ω̄ − 1)(2λ− 1)
1− ω̄p(0)
π∆ =p(1)y(1)(ω − 1)(2λ− 1)
1− ωp(1).
if ω > ω̄ no internal threat to leader
if ω < ω always an internal threat to leader
if ω ∈ [ω, ω̄] a threat absent disclose, but no threat given disclosure
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Equilibrium
Stability Thresholds
R has a dominant strategy to set v = 0 if π∆ < p(d)y(d)(ω−1)(2λ−1)1−ωp(d)
Implicitly define ω̄ and ω s.t.:
π∆ =p(0)y(0)(ω̄ − 1)(2λ− 1)
1− ω̄p(0)
π∆ =p(1)y(1)(ω − 1)(2λ− 1)
1− ωp(1).
if ω > ω̄ no internal threat to leader
if ω < ω always an internal threat to leader
if ω ∈ [ω, ω̄] a threat absent disclose, but no threat given disclosure
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Equilibrium
Equilibrium Disclosure
Proposition
The equilibrium strategy over disclosure can be characterized in thefollowing manner:
For ψ > ψ̄, d = 1 for all θ ∈ {0, 1}.For ψ ∈ [ψ, ψ̄], d = 0 iff θ = 0 and ω > ω̄.
For ψ < ψ d = 0 for all θ = 1. For θ = 0:I d = 0 for ω > ω̄.I d = 1 for ω ∈ [ω, ω̄] iff r1 > ∆ + ρ[µ+ λy(0)]− [2− p(1)]λψ.I d = 1 for ω < ω iff r1 > ∆− λψ + [1− p(0)][µ+ λy(0)]
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Comparative Statics
Investment and Disclosure
Proposition
Equilibrium disclosure is rising in the economic returns to transparency ψ.
straightforward intuition: higher economic benefits leads to higherdisclosure
important empirical implication: transparency should be associatedwith increased investment
I both due to a causal effect (it is assumed that ψ > 0)I and an endogenous equilibrium effect (↑ of ψ yield ↑ disclosure)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Comparative Statics
Investment and Disclosure
Proposition
Equilibrium disclosure is rising in the economic returns to transparency ψ.
straightforward intuition: higher economic benefits leads to higherdisclosure
important empirical implication: transparency should be associatedwith increased investment
I both due to a causal effect (it is assumed that ψ > 0)I and an endogenous equilibrium effect (↑ of ψ yield ↑ disclosure)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Comparative Statics
Investment and Disclosure
Proposition
Equilibrium disclosure is rising in the economic returns to transparency ψ.
straightforward intuition: higher economic benefits leads to higherdisclosure
important empirical implication: transparency should be associatedwith increased investment
I both due to a causal effect (it is assumed that ψ > 0)I and an endogenous equilibrium effect (↑ of ψ yield ↑ disclosure)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Model Comparative Statics
Institutions and Disclosure
Proposition
Leaders disclose for a wider range of values when the consequences ofleader removal for stability are low (ω ≤ ω̄) than when these consequencesare high (ω > ω̄).
greater disclosure in institutionalized than personalistic regimes
greater disclosure under new leaders than under entrenched leaders
Skip to Conclusion Skip to Results
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Evidence on Coups and Revolt
Mass Revolts and Democratization vs. Coups
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Data
Data Definitions
Test these predictions using:
HRV Transparency Index (HRV, 2014) as a measure of disclosure ofeconomic info
Two datasets on autocratic institutionsI GWF – party, personalistic, and military (and monarchies)I DD dataset – singleparty, multiparty, elected legislatures
PWT 7.1 economic data
UNCTAD data on FDI inflows (current USD)
Svolik (2012) for definitions of regimes and leaders’ time in office
Standardize all covariates that aren’t either indicators or time counts
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Who Discloses?
Empirical Model
Varying intercepts hierarchical model:
transparencyi ,t = ρtransparencyi ,t−1 + αi + Xi,t−1β + εi ,t
αi ∼ N(Ziγ, σ2α)
Zi denotes time invariant institutional characteristics
Xi ,t−1 denotes ec. data, leader time in office, cubic polynomial of time
i is an autocratic regime (some of which are quite short-lived)
Estimate via MCMC
Skip to Results
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Who Discloses?
Bias in LDV Varying Intercepts Models
To deal with bias, we estimate the following system of equations:
∆transparencyi ,t−1 = µ+ ζtransparencyi ,t−2 + ∆Xi,t−1ψ + νi ,t−1
∆transparencyi ,t = ρ̂∆ ˆtransparency i ,t−1 + ∆Xi,t−1β̂ + ηi ,t
transparencyi ,t = αi + ρ̂transparencyi ,t−1 + Xi,t−1β̂ + εi ,t
αi ∼ N(Ziγ, σα)
Similar results, somewhat inflated standard errors
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Who Discloses?
Who Discloses?: Results w. GWF Data
LDV ModelsModel 1 Model 2 Model 3
Party 0.002 0.002 0.002[-0.033, 0.038] [-0.039, 0.031] [-0.037, 0.036]
Personal -0.039 -0.038 -0.044[-0.083, -0.001] [-0.085, -0.007] [-0.087, -0.008]
Fuel Exporter -0.037 -0.036 -0.033[-0.082, 0.010] [-0.073, 0.006] [-0.070, 0.008]
Lag Transparency 0.960 0.961 0.964[0.943, 0.978] [0.943, 0.977] [0.947, 0.980]
New Leader 0.023 0.024 0.024[-4×10−4, 0.047] [0.001, 0.048] [0.002, 0.049]
# Obs 1530 1530 1530# Regimes 119 119 119
Ec. controls and cubic polynomial of time included in all specifications
Skip to FDI Skip to Conclusion
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Who Discloses?
New Leader Marginal Effect
Figure: Marginal Effect of a New Leader
2 4 6 8 10
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Marginal Effect of New Leadership
Time (in years)
Cha
nge
in H
RV
Sco
re (
Sta
ndar
d D
evia
tions
)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Disclosure and FDI
Empirical Model
Varying Intercepts Hierarchical Model:
FDIi ,t = αi + ρFDIi ,t−1 + γtransparencyi ,t−1 + Xi,t−1β + εi ,t
Estimated via MCMC
Nickell bias less of an issue here:
i denotes a country, rather than a regime, so long panels
GMM estimates from Stata return substantively similar results(slightly higher standard errors)
same with fixed-effects LSDV models (slightly larger point estimates)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Disclosure and FDI
Empirical Model
Varying Intercepts Hierarchical Model:
FDIi ,t = αi + ρFDIi ,t−1 + γtransparencyi ,t−1 + Xi,t−1β + εi ,t
Estimated via MCMC
Nickell bias less of an issue here:
i denotes a country, rather than a regime, so long panels
GMM estimates from Stata return substantively similar results(slightly higher standard errors)
same with fixed-effects LSDV models (slightly larger point estimates)
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Empirics Disclosure and FDI
Estimated Marginal Effects
2 4 6 8 10 12 14
050
010
0015
00
Marginal Effect of Transparency
Time (in years)
Cha
nge
in F
DI I
nflo
ws
(Mill
ions
US
D)
$198 million-1.2 billion long term ↑ in annual FDI flows (point est. $700million.).
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Conclusion
Conclusion
Construct a model of disclosure consistent with existing empirical findings
transparency increases the risk of mass mobilization
and reduces the risk of coup
Novel argument that autocratic leaders may gain from deliberatelydestabilizing the regime
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Conclusion
Conclusion
Construct a model of disclosure consistent with existing empirical findings
transparency increases the risk of mass mobilization
and reduces the risk of coup
Novel argument that autocratic leaders may gain from deliberatelydestabilizing the regime
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Conclusion
Conclusion
Demonstrate that:
1 newly installed leaders more likely to disclose
2 institutionalized autocracies (not personalistic, presence of electedlegislatures) more likely to disclose
3 transparency associated with increased net FDI inflows
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Conclusion
Conclusion
Demonstrate that:
1 newly installed leaders more likely to disclose
2 institutionalized autocracies (not personalistic, presence of electedlegislatures) more likely to disclose
3 transparency associated with increased net FDI inflows
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Conclusion
Conclusion
Demonstrate that:
1 newly installed leaders more likely to disclose
2 institutionalized autocracies (not personalistic, presence of electedlegislatures) more likely to disclose
3 transparency associated with increased net FDI inflows
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Conclusion
Conclusion
Demonstrate that:
1 newly installed leaders more likely to disclose
2 institutionalized autocracies (not personalistic, presence of electedlegislatures) more likely to disclose
3 transparency associated with increased net FDI inflows
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Examples
One Tool to Pacify Regime: Increased FDI
Transparency as part of broader attempts to increase size of pool for rentsfunneled to regime members
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Examples
Another Tool to Pacify the Regime: Facilitate MassMobilization
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Measurement Model
Transparency as Missing Data
World Development Indicators (Downloaded Dec. 2012)
Items: 240 variables from across WDIrecoded into indicator {0, 1} equal to 1 if non-missing
Panels: 125 countries
Time: Annual obs., 1980-2010
3875 observations
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Measurement Model
Measurement ModelItem Response Model
240 equations of the form:
Pr(yj ,c,t = 1|transparencyc,t) = logit(δj + βj transparencyc,t)
j ∈ {1, 2, ..., 240}c ∈ {1, 2, ..., 124}t ∈ {1, 2, ..., 31}
Priors: (δjβj
)∼ N(
(00
),
(100 0
0 100
))
transparencyc,1 ∼ N(0, 100) recentered at each iteration of the MCMC algorithm
transparencyc,t ∼ N(transparencyc,t−1,1
τc) ∀ t > 1
Cuba constrained to be negative, Sweden positiveHRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Transparency and State Capacity
Transparency v. GDP in Democracies and Autocracies
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Transparency and State Capacity
Transparency v. GDP in Democracies and Autocracies
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Transparency and Regime Transition
Hazard of Regime Transitions
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Relationship with Alternative Forms of Transparency
Freedom House v. HRV
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Relationship with Alternative Forms of Transparency
Newspaper Circulation per capita v. HRV
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Relationship with Alternative Forms of Transparency
Media v. HRV
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Relationship with Alternative Forms of Transparency
FOIs v. HRV
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Other Time Series
Other Country Time-Series: Declines
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Other Time Series
Other Country Time-Series:EU Accessors
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Other Time Series
Time-Series: One Regime Transition
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Other Time Series
Time-Series: One Regime Transition
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015
Extra Slides Other Time Series
Time-Series: Multiple Regime Transitions
HRV (Minnesota,NYU,Georgetown) Autocratic Transparency December 14, 2015