Transcript measure of property rights institutions

Political Economy of
Economic Policy
Lecture 2
Property Rights Theory
and Property Rights Institutions
Chenggang Xu
copyright@Chenggang Xu
Institutions and ownership
• Ownership of means of production as the cornerstone dividing two
types of institutions:
– Socialist institutions: public ownership (Marx, 1870; von Mises, 1920; Lange,
1930; Hayek, 1940; Kornai, 1980)
• e.g. former Soviet Union (before 1990), China before mid 1990s, etc.
– Capitalist institutions: private ownership
• e.g. private ownership dominates in US, UK, Western Europe, Japan, South
Korea, etc.
– Transition: the most fundamental change is to privatize public assets
• e.g. Russia, China, Central Eastern European countries, etc.
• A classical socialist economy is a one firm economy
– All means of production are owned by this ‘firm’
– no market, no market prices
• Can state ownership be as efficient as market economy?
– Market socialism (Lange, 1930s): keep public ownership; use markets
• Can incentive problems be solved in a market socialist economy?
– Ownership involves incentives (Hayek vs Lange, 1940s)
Ownership (Why should ownership matter?)
Incentive problems in a team (Alchian-Demsetz):
- to produce a product needs efforts from all team members
- exerting efforts is costly to team members that every one wants
to shirk if possible
- if every one shirks there is no production
- how to solve the problme?
- can hire a manager to monitor team members
- but the manager’s efforts are also costly, thus does not want to
work hard
-Manager does not work: team members do not work
Incentive problems in a team
• How to solve manager’s incentive problem?
– who monitors the manager?
• Allocating assets ownership to some one can solve the problem:
– The owner cares residual income of his assets
– Thus has ultimate incentives to monitor the manager
– The chain of incentives of team members and managers is solved.
• To solve incentive problems in a team one may not necessarily rely
on assigning ownership:
– if contracts are complete, an optimal incentive schemes can solve the above
problem regardless of ownership structure.
– apply to large firms?
– if all assets are owned by a dictator, will this be efficient?
– separation of ownership and control -- the residual control and residual
income are separated;
Why should ownership matter?
Tragedy in the commons (Demsetz)
- A lake is ‘owned’ commonly by a large number of
citizens
- Protecting the lake
- costly to private; benefits to the public
- no individual has incentive to protect
- over fishing in a commonly owned lake
- Negative externalities associated with the commons
- allocate private ownership as a solution
Tragedy in the commons
• To solve tragedy in the commons problems one may not
necessarily rely on assigning ownership:
- If bargaining is efficient, then who owns what doesn't
matter (Coase theorem).
- For Coase theorem to work requires zero negotiation cost and
zero enforcement cost for all citizens who own the lake
- A complete contract achieves efficient solution.
- The contract specifies rewards, punishments, charges,
maintenance, etc. for all people involved under all
contingencies
- Requires low costs of negotiating/enforcing the contract
Complete Contract model setup
• Quality of an output of the agent: q = g (e, ε)
– q: observable and verifiable
– e: effort of the manager
– ε: exogenous shocks realised after choosing e
• Cost of effort H(e)
•
•
•
•
•
revenue from the quality: r(g)
Principal's utility: Up(r(q)-t) = r(q)- t
RN
Agent's utility:
UA(t,e) = V(t) - H(e)
V(t) is concave; Agent is RA
there are i = 1, 2, …, N contingencies, each with a
probability  i
Complete Contract model solution
• a complete contract {p i } covers all the N contingencies :
Max
i  i E{ri (qi(ei, εi)) – pi(qi(ei, εi))}
{e i, p i }
s.t.
ei ε argmax i  i {E[V(pi (g(ei', εi)))]-H(ei)}
ei'
(IC)
i  i E{V(pi (g(ei', εi))) - H(ei) }  U
(IR)
• SOLUTION: {pi},
i = 1, 2, …, N
Complete Contract Solution
• a complete contingent contract that corresponding each
contingency i there is pi (.)
• Pr(contingency i) contract
– 1
– 2
– 3
–
p1 (.)
p2 (.)
p3 (.)
……….
– N
pN (.)
Incomplete Contract and Property Rights
• Contract complete ==>
Ownership does not matter
• Coase Theorem: As long as bargaining is efficient
(e.g. parties can achieve complete contract), the efficient
result should always be achieved through bargaining
regardless of the initial allocation of property rights
(ownership structure doesn't matter).
• if for some i, pi (.) is missing, or is not unambiguously
defined, the contract is not complete.
Why is allocating property rights important?
• Incomplete contracts if writing complete contracts is
 too costly
 too difficult
 not possible
• some items are non-verifiable or un-observable
• when contracts are incomplete or cannot be enforced
– What to do if unspecified contingencies occur?
– Who is in charge if unspecified contingencies occur?
– Who has the right to make decisions if unspecified contingencies occur?
• Residual rights of control: the right to make decisions when
unspecified contingencies occur
• when contracts are incomplete there are residual rights left to be
allocated – the owner of assets has the right
Hold-up & incomplete contracts
Hart (1995), Ch 1-2
• Complete contract: with a full coverage of all contingencies
– Contract can be enforced effectively
• Incomplete contract: contingencies not fully covered
– Contract cannot be enforced effectively
– Knowing this, parties may take advantages in implementation
process
• Relation specific (rs) investments
– Two parties both gain from the relationship if both invest
• With rs investment and incomplete contract
– One party may hold up the other when the other party has invested
in the relationship
– Anticipating this, both parties may reduce rs investments
Implications and limitations of
Property Rights Theory
• Implications
– How small is a firm too small (e.g. family business)
• Never be too small if no complementary assets are involved
– How large is a firm too large (e.g. a centralized economy)
• Existence of independent assets in the economy
– What should be privatized together or separately
• Limitations (fail to explain some important phenomena)
– Outsourcing in Japanese manufacturing, e.g. car industry
• Auto maker provide design to subcontractors who make relation specific
investments)
• Contracts are highly incomplete: no details on quality, prices
• Instead of ownership, long term relationship is a mechanism to avoid holdup
Functions of the state:
Contracting vs. Property Rights Institutions
• Two theories of the state (North, 1981)
– ‘contract theory’ of the state: state as the contracting institution
– ‘predatory theory’ of the state: state as the property rights institution
• Contracting institutions: the rules and regulations governing
contracting between ordinary citizens
– the functioning of the legal system
– regulate transactions between private parties, e.g. a debtor and a creditor
– the state and associated institutions provide the legal framework that enables
private contracts to facilitate economic transactions (i.e., reduce transaction
costs.)
• Property rights institutions: the rules and regulations protecting
citizens against the power of the government and political elites
– the functioning of the government
– Institutions protecting investors against government expropriation
– the state is an instrument for transferring resources from one group to another.
Unbundling Institutions
(Acemoglu and Johnson, JPE, 2005 )
• Good institutions will support private contracts and protect
private property rights (North, 1981)
• Contracting institutions affect what kind of contracts can be
written and enforced, thus affect efficiency of organizations
and societies (Coase, Williamson, Hart)
• Property rights institutions affect protection of private
properties, thus investment and efficiency of economies
• Which institution has larger impacts to societies/economies?
• Can we distinguish these two institutions in reality?
• Can we separate the roles of these two institutions?
Measuring Contracting institutions
• Index of legal formalism (Djankov et al., 2003):
– The number of formal legal procedures necessary to resolve a case of collecting
on an unpaid check
– The larger the higher cost the legal system
• index of procedural complexity (The World Bank, 2004):
– The difficulties in resolving the case of an unpaid commercial debt
– The larger the less effective the legal system
• The number of procedures necessary to resolve a court case involving
an unpaid commercial debt (The World Bank, 2004):
– The more the worse
• Example: enforcing a simple commercial debt contract
– Dominican Republic: costs 440% of income per capita and requires a process of
495 days
– New Zealand: costs 12% of income per capita, and requires 50 days.
Measuring property rights institutions
• Constraint on the executive (the Polity IV dataset)
– This measure ranges from 1 to 7: the higher the greater
constraints on politicians and politically powerful elites
• Protection against expropriation by government (Political Risk
Services)
– A value between 0 and 10 for each country and year: the higher
the better protection
• Private property index (the Heritage Foundation)
– The higher the better private property is protected against
government and other sources of expropriation.
Data
• 71 former European colonies
– 25 common law origin countries
– 46 civil law origin countries
• Outcomes:
– Growth: GDP/capita
– Investment: Investment to GDP ratio
– Financial development: Credit to private sector; stock market capitalization
• Contracting institutions
– Index of legal formalism
– index of procedural complexity
– The number of procedures necessary to resolve a court case involving an
unpaid commercial debt
• Property rights institutions
– Constraint on the executive
– Protection against expropriation by government
– Private property index
World average of institution measures
Basic empirical model
• Yc = a Fc + b Ic + Zc’go + ec
• Yc: outcome for country c
–
–
–
–
•
•
•
•
Per capita GDP;
investment/GDP;
private credit/GDP;
stock market capitalization/GDP
Fc: measure of contracting institutions
Ic: measure of property rights institutions
Zc: vector of other controls.
go: vector capturing effects of the control variables
in Zc
Contract institutions vs. GDP/capita
Property right institutions ‘help’
GDP/capita
Establish causalities
• Yc = a Fc + b Ic + Zc’go + ec
(1)
• Both institutions are important to outcomes: a and b are
both positive and significant
• From there do we know institutions determine better
outcomes?
• Or outcomes determine institutions?
• Both contracting and property rights institutions are
endogenous (better outcomes may lead to better
institutions):
– Reversed causality can have the same results
– impacts of institutions may reflect omitted characteristics (e.g.,
geography, religion, etc.)
• How to determine which causes which?
2SLS with instruments for institutions
• Estimate equation (1) using Two-Stage Least Squares
(2SLS) with instruments for institutions
– Instruments should be correlated with the endogenous
regressors: Cov (Ic; Lc) > 0, Cov (Fc; Mc) > 0
– uncorrelated with outcomes: Cov (ec; Lc)=Cov (ec; Mc)=0
• Fc = d1 Lc + h1 Mc + Zc’g1 + u1c
• Ic = d2 Lc + h2 Mc + Zc’g2 + u2c
– Mc : instrument for property rights institutions:
• log mortality rate of European settlers; or
• log of the indigenous population density in 1500
– Lc : instrument for contracting institutions:
• a dummy for English legal origin
– whether or not the country was a British colony
Instrument variables
for property rights institutions
• Settler mortality in countries that were colonized by
European nations between 1500 and 1900
– Where the mortality rates were low for European settlements,
they migrated in large numbers and developed institutions that
protect property rights better (e.g., North America, Australia,
etc.)
– Where the mortality rates were high for European settlements
(e.g. >50%), the colonizers exploited the native population and
did not protect property rights (e.g., sub-Saharan Africa, South
Asia, and Central America, etc.)
• initial indigenous population density
– Where this was high, Europeans were more likely to develop a
forced labor system
– Where this was low, Europeans were more likely settle
themselves, and develop institutions that protect their properties
Instrument variables
for contracting institution
• Legal origin
– Common law vs. civil law
• Legal origin is an important determinant of all three measures of
contracting institutions
– legal origin explains about 40 percent of the variation in legal formalism
– for former colonies legal origin is exogenous
• The British imposed common law on their colonies
• The French imposed civil law on their colonies
– Problem: legal origin affect economic outcomes through
channels other than contracting institutions
• given Cov(ec , Lc) ≥ 0 , the estimate of the impact of contracting
institutions on economic outcomes is biased upwards
1st stage regressions with instruments
• Fc = d1 Lc + h1 Mc + Zc’g1 + u1c
• Ic = d2 Lc + h2 Mc + Zc’g2 + u2c
– Mc : instrument for property rights institutions:
• log mortality rate of European settlers; or
• log of the indigenous population density in 1500
– Lc : instrument for contracting institutions:
• a dummy for English legal origin
– whether or not the country was a British colony
First Stage OLS Regressions for
Contracting Institutions
Dependent variable is measure of contracting institutions
(standard errors are in parentheses)
•Discoveries:
•English legal origin has a large and significant effect on legal formalism
•Settler mortality and population density have no significant effect on legal
formalism
First Stage OLS Regressions for
Property Rights Institutions
Dependent variable: measure of property rights institutions
(standard errors are in parentheses)
•Discoveries:
•English legal origin has no significant effect on constraint on executive
•Settler mortality and population density have a large effect on the
constraint on executive
Discoveries
from 1st stage regressions
• English legal origin mainly affects contracting
institutions
– Plausible explanation: Who colonized (British vs.
French/Spanish), but not the details of colonization
strategy, shapes contracting institutions.
• Settler mortality and population density before
colonial times mainly affect property rights
institutions
– Plausible explanation: The way in which countries were
colonized (settling vs. exploiting), but not who colonized
them, is a robust determinant of property rights institutions
2nd stage regressions with instruments
• Yc = a Fc + b Ic + Zc’go + ec
(1)
– Where, Fc and Ic are estimated by the following 1st
stage regressions
– Fc = d1 Lc + h1 Mc + Zc’g1 + u1c
– Ic = d2 Lc + h2 Mc + Zc’g2 + u2c
• Mc : instrument for property rights institutions:
– log mortality rate of European settlers; or
– log of the indigenous population density in 1500
• Lc : instrument for contracting institutions:
– a dummy for English legal origin
» whether or not the country was a British colony
Contracting vs. Property Rights Institutions:
Dependent variable is log GDP per capita
(standard errors are in parentheses)
• 2nd stage of 2SLS:
with log settler mort.
as instrument
– Contracting institution:
• Insignificant effects
– Property rights institution:
• Significant effects
• Equivalent OLS:
Contracting vs. Property Rights Institutions:
Dependent variable is stock market capitalization
(standard errors are in parentheses)
• 2nd stage of 2SLS:
with log settler mort.
as instrument
– Contracting institution:
• Significant effects
– Property rights institution:
• Significant effects
• Equivalent OLS:
Major discoveries
• Property rights institutions have a first order effect
on income per capita, the ratio of investment to
GDP, the level of credit, and stock market
development
• Contracting institutions only have an effect on
stock market development
• For all variables, the effect of contracting
institutions is quantitatively much smaller than the
effect of property rights institutions
Semi-Reduced Form
regressions
• It is possible that legal origin may affect
outcomes through channels other than
contracting institutions
– Using legal origin as instrument for contracting
institutions may miss these effects!
• To address this issue, let legal origin enter the second
stage regression directly; whereas property rights
institutions are instrumented. Results:
– English legal origin has no significant positive effect, except on
stock market capitalization
– All three measures of property rights have a significant positive
effect on the outcomes independent from instrument used
• No evidence of a direct positive effect of English
legal origin on the outcomes
Alternative Samples
• Examine whether there is an effect of property
rights institutions within each legal family
– Drop contract institutions in regressions
– Restrict the sample to common law countries (i.e.,
former British colonies)
– Restrict the sample to French legal origin countries
• Essentially the same results are retained for both groups
of legal family countries
– much of the variation in institutions within each legal family
countries can be explained by the colonization strategy of
European powers
General discussion
• Contracting institutions and legal rules have some effect on the form
of finance
– Stronger contracting institutions use more equity than debt contracts
• Contracting institutions have limited or no effects on major
economic outcomes
– long-run growth (current income levels)
– Investment to GDP ratio
– The overall amount of financial intermediation in the economy
• Property rights institutions matter significantly for all these outcomes
– determine how the government, politicians, and elites are constrained in their
relationships with the rest of the society
• Question for further thoughts:
– What do the proxies (indices) used by the paper capture in reality?
– Do they really correspond to property rights institutions and contracting
institutions?
– What are mechanisms that property rights institutions and contracting
institutions affect outcomes?