Transcript 7th Lecture, STV4346B: “The economic effects of institutions”

7th Lecture, STV4346B:
“The economic effects of institutions”
Carl Henrik Knutsen, Department of Political Science, UiO
24/11-2008
Some introductory notes




Last lecture, we looked at studies that claimed a positive
effect on the economy from a complex (of little specified)
institutional structures.
Today we are going to be more specific when it comes to
the political/institutional (independent) variables.
What is the effect of democracy on economic growth and
property rights?
What are the economic effects of particular electoral
systems and forms of government, within the set of
democracies?
Democracy and growth




Conceptual and measurement issues with regard to
democracy, see lecture 5
Which aspects of democracy affect the economy, the
benefits of an analytical approach: the quest for more
concrete mechanisms
The same with regard to ecoonmic growth: Which
immediate factors behind growth is affected?
Technological change, A (incl efficiency in markets,
organizational change that improves efficiency/reduces
transaction costs), physical capital, K, Human capital, H,
Labor input, L
Y= F(A, K, H, L), with all partial derivatives>0.
The Lee-thesis


Amartya Sen (1999:15):
“[A] great many people in different countries of the world are
systematically denied political liberty and basic civil rights. It is
sometimes claimed that the denial of these rights helps to
stimulate economic growth and is “good” for economic
development. Some have even championed harsher political
systems – with denial of basic civil and political rights – for
their alleged advantage in promoting economic development.
This thesis (often called “the Lee thesis”, attributed in some
form to the former prime minister of Singapore, Lee Kuan Yew)
is sometimes backed by some fairly rudimentary empirical
evidence”
Theoretical arguments




A vast catalogue of potential mechanisms. Knutsen (2006,
MA-thesis in political science): at least 20!
These mechanisms point in different directions with
regard to whether democracy affects growth positively or
negatively, relative to dictatorship.
Interaction effects
Dictatorship as a very eclectic category. Can be classified
along different dimensions. See for example Linz and
Stepan’s book or Hadenius and Teorell (2007), or Knutsen
(2007, MA-thesis in economics)
General theoretical arguments, an overview of the
theoretical presentations in P&L (1993) and Knutsen
(2008)
Argument
Przeworski and Limongi's conclusions
My (Knutsen’s) conclusions
1) Democracy and property rights
Either way
For democracy
Either way (for democracy when adding
2) Democracy and investment
Against democracy
human capital)
particularistic pressure
Against democracy
Against democracy, but with qualifications
4) Autonomous rulers are predatory
For democracy
For democracy
5) Democracy and technological change
-
For democracy
3) Dictatorship and insulation against
The arguments: read through the detailed
arguments in P&L and K!





1) The poor median voter and redistribution of property in
democracies (Marx, Mill etc). But, rulers with private gains or whose
constituencies have private gains, of grabbing property (North)
2) Dictatorship increases investment. Reduces public and private
consumption through different policies, since generally less
responsive to citizens’ demand for consumption. But: democracy and
human capital
3) Autonomous dictatorial regimes, and the ability to ignore special
interestshardhanded reforms and policies for economic growth
4) Dictators might be predatoryDisincentives for growth in these
regimes. Concentration and abuse of power
5) Democracies and freedom of speech, media and movement
more innovation and better diffusion of foreign and domestic
technologies in the economy
P&L (1993): the abstract
Methodology


Earlier studies using OLS and problems
Endogeneity and selection of regimesselection models
and sugestive simulations



We might find effects through OLS that are actually not there if
there is systematic selection of regimes after how countries
perform economically
Panel data vs cross section analysis
Meta-study of earlier studies: even split between those
that find a positive and negative general effect on growth
from democracy.No actual effect?
The earlier studies investigated by P&L
Later studies and nuances



My results (Knutsen, 2008): longer time series, various econometric
methods, different measures There seems to be a positive general effect
from democracy on growth. About 1% extra annual growth. This result
stands in contrast with the ”no significant effect”-result in P&L 1993.
Barro’s study and the inverse U-curve: semi-democracies grow faster than
both ”extremes”. But in Knutsen(2008), this result does not find support.
Rodrik: Democracies have higher wages, show less fluctuation in growth
rates (within country), have less variation between countries



The last point is important: Dictatorships vary much more in their economic
performances: both the best and the worst performances ..
Type of dictatorship and growth?
Different effects from democracy in different regions of the world? Asia vs
Africa.
Democracy and property rights



Due to the centrality of (stable) property rights for economic
growth, it is particularly important to investigate the effect of
democracy on property rights.
Boix empirical study: less expropriation in democracies
Olsons’s theoretical study: Only in particular instances do
dictators have the incentive to protect property rights to the
same degree that democratic governments do (representation
of broad interests in democratic government). Dictators with
long time horizons or with encompassing interest might have
strong incentives to protect PR.

But, see James Robinson’s working paper from 2001: If dictators
acknowledge that development might affect their political position,
the positive effects from encompassing interest and patience on
policy might disappear.
Olson (1993)
•
•
Why is dictatorship better than anarchy and democracy better than dict. in
terms of generating development
Key assumptions
–
–
–
–
–
–
Externalities from roving bandits on other bandits and on society, everybody
maximizing consumptionlittle production (kind of Pris. D. game)
Rational dictator: stationary bandit that monopolizes power. Maximizes utility,
which is a function of consumption. Brings security
Taxes have negative effects on productivity, but revenue raised can be used on
public goods with positive effects.
Dictator maximizes tax revenue –expenditure on public good: T(Y)-G, let T(Y) be
tY, where t is the tax rate.
However;Y is also a function Y(t), where Y’(t)<0: Higher tax rate gives less
incentive for people to produce
Interpret tax broadly: Any government activity that increases government’s
revenue, while affects total production (negatively). More general protection of
individual property rights than tax in the everyday use of the word: private
entitlement to own production
Olson cont’d
•
•
•
•
•
Hunting, gathering societies, small groups and ability to solve
collective action problems (Olson, 1965)
Agriculture and the problem of roving bandits. If no
state/protectionincentives for people to become bandits and
raid farms rather than produce.
In such a world, low production, since no incentives to
produce even for those who keep farming
A roving bandit settles down, picks up a crown and calls
himself King. The stationary bandit has emerged.
Why does production increase under the stationary bandit?
Shouldn’t he be “predatory”? Olson: the stationary bandit as a
cattle rancher rather than a predator. Understands that
production will increase if tax set at lower level than max and
provide some public goods more income to tax from!!
Olson cont’d
•
Dictator does not care about subjects income but will provide just
enough incentives to maximize his own revenue
–
Tax revenue set at point where direct positive effect from extra tax
revenue gathered equals the negative indirect effect on tax revenue via
reduced total production.
•
•
Ex: If assume that revenue maximizing tax rate is 1/3  last tax dollar reduces
national income by 3 dollars
Dictators will thus tax at a too high rate relative to the social
optimum. Which dictators will produce the lowest tax rates, (i.e. be
closer to optimum)?
–
–
Dictators with some stake in the market (dictators with income from
market activity not only from taxes/expropriation etc. The more
encompassing the interests are, the more the dictator will internalize the
negative externalities on production from taxation
Dictators with a long time horizon (political stability, patience, dynasties):
Low taxes More investment  Bigger cake in the future to tax from.
When dictator has very short time horizon acts like the roving bandit
and grabs all he can..
Will democracy do better?
•
Yes! Even if we assume that all actors are similar (no nicer guys under democracy),
democracy will generally have higher production. Why?
–
–
•
•
•
Expropriation and negative effects on today’s production and investment, but also
on incentives to innovate..
The importance of rule of law, independent judiciary and individual rights both for
property rights and thus production, but also for stability of democracy
Many ways to expropriate/violate property rights:
–
•
Majority earns a large share of the market income, will thus take into account that taxes/
expropriation reduce their income from the market. Still tax higher than optimal rate, but
lower than dictator, which again is lower than when you have scores of roving bandits
(grab all you can!)
But, the negative effects of powerful special interests under democracy (Olson, 1982)
F.ex. governments with debt can run hyperinflation government wins, but private capital
eroded.
When will we have transition to democracy? Balance of power (different groups)
and the difficulty of being the only stationary bandit
Boix ch. 6 (important, but short presentation: a lot of
the points have been covered in other texts)
•
•
•
Dictatorship violation of property rights, corruption and patronage
negative effects on development
Empirical evidence for the first link provided. In addition to democratic
institutions, newspaper circulation has a positive effect. Look at these
results yourselves!
The role of political accountability for disciplining democratic governments.
Horizontal (group conflict) and vertical (politician-citizen) dimensions of
politics. Accountability and the vertical dimension.
–
–
•
•
Politicians with better info
Citizens’ incomplete control mechanisms
Can have some degree of accountability in dictatorships as well (revolution
threat, coups etc), but generally weaker than in dem.
Differences between dictatorships: those with legislatures do better: The
negative effects of power concentration and personalization of power and
politics. Absolutism’s negative effects
Boix cont’d
•
Democracy as a political equilibrium:
–
–
•
•
•
•
Non-specific assets and equality democracy increased equality and further
diversification of economy (through less corruption and expropriation)stabilizes
democracy..
From economic structure to politics (Ch.1) and from politics to economic structure
(Ch.6)
Agrarian economies and specific assets Incentives to grab and difficulty of
democratization
Industrialization  reduced asset specificity democratization (link between Boix’
model and modernization theory, in addition to a possible effect from increased
GDP/industrialization on economic inequality)
Cycles of revolutions and left-wing authoritarian govt’s. When left-wing govt take
power gather resources for themselves new economic elitenew revolutions
Draws on North And Olson’s frameworks, but additional points
–
–
Threat of war and impetus for industrialization
The scattered map of Europe: Increased mobility of physical and human capital: I’ll just
move to neighbouring country if King rules like a TyrantIncentives to not expropriate to
heavilyequality (plus the capital mobility)Democracy
Persson and Tabellini (2004)


For those interested, I can really recommend the 2003
book by the same authors
Among democracies, how do constitutions affect different
economic aspects?




Electoral rules (Proportional representation vs plurality rule;
but also other details of the electoral system)
Form of government (Parliamentary vs Presidential system)
JEP- article is a general overview of the field, results from
2003 book a natural focus
Theoretical arguments and empirical results
Persson and Tabellini
•
Some theoretical points, electoral rules
–
–
–
–
Plurality rule and political accountability. Voters know who to punish
and reward Improve economic performance, particularly through
reduced rents and corruption.
But, plurality rule and small district magnitude: incentive to spend on
programs in “swing states” (size of district crucial for how narrow
program is), and to spend on narrow groups that are pivotal. PR and
broader, more universal programs (“representation issues”)
Open lists no barriers to entry and electoral accountability less
corruption (can throw out the corrupt). Intraparty competition and
patronage..Better to open up the competition to the electorate and
have competition between parties
PR coallition governments increased spending and maybe
increased deficits (narrow, small parties’ demands and collective
action problems)
P&T cont’d
•
Some theoretical points; forms of government
–
–
–
–
More accountability under Presidentialism? Simpler chain of
delegation and checks and balances less corruption
But, fixed term for Presidents Do not have the disciplining
device of a no-confidence vote do not need a stable
majority specific interests benefit
Gridlock between presidents and constituencies less
spending and less ability to implement reform and react to
crisis (but increased time horizon under Presidentialism)
Winner-take-all nature of presidential election and possibility
of strong political conflict
Methodological issues, P&T
•
•
•
Selection of electoral rules and form of government:
Specific countries tend to have specific political
structures. The danger of omitted variable bias
Possible to use intra-national variation is small due to
inertia in the political structures
Solutions:
–
–
IV-approach, find exogenous variation in electoral rules and
forms of government due to “international fashions” changing
over time
Uses matching-approach to allow for non-linear effects. Find
most similar countries with differences on the constitutional
variables
Empirical findings summarized in P&T
•
•
•
•
•
•
Plurality rule, and voting on individuals rather than party lists
reduce corruption
Less corruption in districts with tough party competition
(evidence from Italy)
Less corruption in districts with small magnitude (less barriers
to entry)
Under Plurality rule, politicians cut taxes and spending in
election year, but not to same degree under PR
PR larger public spending (5-10% than plurality) Much of the
effect due to higher frequency of coallition governments under
PR
PR1-2% larger budget deficit/GDP
Empirical findings summarized in P&T
•
•
•
•
•
Presidentialism only reduces corruption in “good democracies”
(High Polity, low FHI-score). A higher share of presidential countries
are “bad democracies” (mere correlation or effect from
presidentialism? Winner take all nature of presidential elections and
incentives to violate democratic rules of the game). overall more
corruption in presidential countries
Presidentialism reduces public spending (5% of GDP)
Pro-cyclical fiscal policy under presidentialism. Spend more in good
times, contrary to Keynes’ recommendation
Positive effect on growth and “good” economic policies from
parliamentarism compared to presidentialism (but again, low quality
of democracy in presidential countries..is the effect on growth a
causal effect?)
In large sample with authoritarian regimes included: a positive effect
on growth from parliamentary democracy