Transcript The economic effects of democracy in developing countries

Africa's Growth Tragedy Revisited: Weak
States, Strong Rulers
Tirsdagsseminaret ved ISV
16/11 2010
Carl Henrik Knutsen
ISV, UiO
Core questions and argument
 How do democracy and state capacity affect economic growth?
 Direct, linear effects?
 Democracy (e.g. Przeworski and Limongi 1993; Przeworski et al. 2000; Baum and Lake 2003; Bueno de Mesquita
et al. 2003; Doucouliagos and Ulubasoglu 2008)
 State capacity (e.g. Wade 1990; Evans 1995; Kohli 2004; Fukuyama 2005; Evans and Rauch 1999)
 Interaction between state capacity and regime type?
 The relationship between regime type and growth in Sub-Saharan Africa?
 Low state capacity in post-colonial Africa
 Argument: Dictatorship is particularly bad when there are no other institutional restraints on
dictator. In weakly institutionalized states, the vertical checks from democratic institutions play
vital role in keeping dictators from pursuing “bad policies”.
Main results
 Democracy increases economic growth in Africa
 The positive effect of democracy on growth is stronger in Africa than globally
 Democracy’s effect on economic growth depends on state capacity
 Significant interaction in both African and global samples
 Democracy has positive growth-effect in weak capacity states
 The democracy-state capacity interaction contributes to explaining Africa’s growth
tragedy
 Promote democracy in poor, weakly institutionalized countries!
Africa’s growth tragedy
 Economic stagnation in African countries after decolonization, why?
 The international system, commodity dependence, etc.
 Ethnic fractionalization (Easterly and Levine, 1997)
 Climatic and geographical factors (e.g. Sachs)
 Domestic politics
 The colonial legacy and structure of state institutions (e.g. Englebert, 2000; Acemoglu et al., 2001)
 The dominance of neo-patrimonialism/clientilism (e.g. Medard, Chabal and Daloz)
 Does political regime type matter? Dictatorship and power concentration
“Democracy doesn’t fit/work in
Africa”
 Similar arguments for Germany, Catholic countries, Latin America, Asia
 The relevant comparison: the ideal vs the counterfactual (dictatorship)
 Optimism early 1990s, pessimism thereafter. Lindberg’s results (2006)
 “The same leaders are elected..”
 Candidate selection vs disciplining effects
 “Neo-patrimonialism survives in democracies” (Chabal and Daloz 1999)
 True, but inertia does not imply everything is constant
 Relative power in patron-client relations
 The size of client-pools likely to increase, with positive economic effects (Bueno de
Mesquita et al., 2003)
Interaction between weak states and strong rulers
 Political economy and the incentives of rulers: Why would leaders want to promote
growth enhancing policies?
 Democracies: need to enact popular policies to stay in office
 In dictatorships:
 Depends on autonomy and type of winning coallition
 Alternative institutional checks on dictator (independence, capacity)
 Other factors: security threats, internal vs external
 In weakly institutionalized states, dictators are able to enhance political survival
and private consumption through “bad policies”
 Property rights, corruption, and other economic institutions
 Fiscal policy
 Markets and industrial policy
 Monetary policy
Robustness checks
 Random Effects
 PPP-adjusted GDP
 Exclude Botswana and Mauritius
 Control for conflict and post-conflict situations
 Control other variables (urbanization, trade, etc.)
 Lags
 Granger-tests and 2SLS
 Arrellano-Bond with one and two lags on DV
A comparison: “Investigating the Lee
thesis…”, EPSR 2(3):451-73.
 This paper presents the hypothesis that democracy hurts economic growth and
development, also known as the Lee-thesis. The paper discusses why one
could expect dictatorship to be particularly beneficial for growth in the
Asian context, and then presents three general theoretical arguments
that support the Lee-thesis. However, the empirical results, based on extensive
time series for more than 20 Asian countries, disconfirm the hypothesis
that dictatorship increases economic growth in Asia: There is no significant,
average effect from democracy on growth. Asian dictatorships do
however invest a larger fraction of their GDP than democracies, but
they are worse at generating high enrollment ratios in education
after primary school.
Results from Asia
Summing up the empirical results
 Significant interaction state capacity and democracy
 Positive effect of democracy on growth in Africa
 Significant and substantially large effect
 Democracy matters more for growth in Africa than elsewhere
 Democracy is particularly beneficial for growth when state capacity is low
 No significant effect on growth in Asia
 Positive effect of democracy on some school enrollment measures
 Negative effect of democracy on capital investment
 Large variation in economic outcomes among dictatorships