DHawkins present

Slide1: 

Dependence Networks and the Diffusion of Democracy Jay Goodliffe Brigham Young University Darren Hawkins Brigham Young University

How and to what extent do states influence the level of democracy in other states?: 

How and to what extent do states influence the level of democracy in other states?

What We Know: 

What We Know “Democratization cannot be seen exclusively as a result of functionally similar processes unfolding independently within each country.” (Gleditsch and Ward 2006) Many studies show that key predictors of democracy are: Percentage of nearby democracies Percentage of global democracies

Two Problems: 

Two Problems 1-Studies have focused on geographic proximity without investigating other possibilities, despite recent theories highlighting other diffusion mechanisms 2-Studies focused on international influences on democracy have slighted domestic factors and vice versa

Why should foreign states matter?: 

Why should foreign states matter? Governments depend not only on their own society for resources but also on other states for aid, trade, recognition, foreign policy goals, and security Domestic nongovernmental groups also at times depend on foreign states Governments face difficult domestic problems and look to others for possible answers

Dependence: 

Dependence Value that actors place on the goods they can obtain through an exchange relation, taking into account their alternative sources of those goods “Need fulfillment that would be costly to forgo” (Baldwin)

Dependence Network: 

Dependence Network For any given state, a dependence network is a set of partner states with whom it regularly engages in exchanges of valued goods. Measurement: -- Trade partners -- Alliance partners -- International Organization (IO) partners All partners are weighted by their material capabilities to capture dependence rather than just connections

Why Dependence Networks Matter: 

Why Dependence Networks Matter Many states offer rewards or punishments (explicit, implicit or anticipated) for getting the “right” domestic institutions Frequent contact allows powerful states to create or influence domestic actors in other states Governments facing domestic difficulties are likely to learn from states with whom they have frequent contact (“availability heuristic)

Key Independent Variables: 

Security Network: Average Polity score of all Alliance Partners, weighted by partner’s security capabilities Trade Network: Average Polity score of all Trade Partners, weighted by volume of trade IO Network: Average Polity score of all partners in significant IOs, weighted by number of shared memberships and by partner’s GDP (all one-year lags) Key Independent Variables

Dependent Variable: 

Dependent Variable Polity Score (-10 to 10) Fixed effects model estimates changes of this variable within a country Cold War effect: Break models into two time periods: 1972-1990 and 1991-2001

International Control Variables: 

International Control Variables Regional Diffusion: Average Polity score in region using “Gravity Model” -- inverse weighting by distance (one-year lag) Global Diffusion: Average Polity score of all states (one-year lag)

Domestic Control Variables: 

Domestic Control Variables GDP/capita (logged) Short-term Inflation (logged, 1 year lag) Long-term Inflation (logged, 10 year average) Short-term Growth GDP/capita (1 year lag) Long-term Growth GDP/capita (10 year avg.) Polity Score (1 year lag)

Slide13: 

Table 1: Influences on Democracy, 1972-1990

Slide14: 

Table 1: Influences on Democracy, 1972-1990

Slide15: 

Table 2: Influences on Democracy, 1991-2001

Slide16: 

Table 2: Influences on Democracy, 1991-2001

Slide17: 

Table 3: Predicted Changes in Polity Score, 1972-1990

Slide18: 

Table 3: Predicted Changes in Polity Score, 1991-2001

ConclusionTruth and Justice (and our model) Will Prevail: 

Conclusion Truth and Justice (and our model) Will Prevail

Conclusion: 

Conclusion States are influenced by their network partners, even in issues like domestic political institutions (and even when controlling for fixed effects and a lagged DV) This influence is more consistent across time than geographic diffusion and is of about the same substantive importance IO partners mattered more during the Cold War; trade partners matter after the Cold War Economic factors don’t matter