17 Apr 2015

Argument

  • growing use of a "language of democracy" by international organizations
  • democratic rhetoric reflects IO legitimation efforts (procedures & outcome)
  • democratic rhetoric varies over time and across IOs
  • explanation 1: world cultural dynamics drive democratic rhetoric
  • explanation 2: democratic rhetoric is a response to demands from IO environments

Structure

  1. Patterns of Democratic Rhetoric
  2. Model 1: World Culture
  3. Model 2: Democratic Visibility
  4. Summary & Next Steps

Patterns of Democratic Rhetoric

  • qualitative content analysis of 20 IOs' Annual Reports from 1980-2011 (29.611 observation units, 36-60 paragraphs per report, 516 complete IO-Years)
  • focus on identity and/or purpose statements and their possible democratic content
  • democratic rhetoric as a kind of normatively laden language, targeted at audiences for legitimation and legitimacy management

Share of Id/Purpose statements and Democracy Statements

  • strong increase in democratic rhetoric

Variation across IOs

  • considerable variation across IOs.

Variation across IO and time

Which democratic values do IOs refer to?

preliminary data:

How do we explain the variation over time and across IOs?

Model 1 : World Culture

  • IOs follow universal scripts of good and rational governance
  • democracy has become such an important script
    • post 1990: age of democracy, we expect strong growth in democratic rhetoric
    • global norm change towards democratic governance in IOs
    • exposure to world culture: UN family IOs and IO HQ hot-spots

Results of our work-in-progress multivariate analysis

Dependent variable:
dummy_dem raw_roll3_dem_rhet raw_roll3_dem_no_promo
logistic Tobit Tobit
(1) (2) (3)
Constant -3.113***
(0.394)
dummy_dem_lag 2.013***
(0.237)
raw_roll3_dem_rhet_lag 1.064***
(0.034)
raw_roll3_dem_no_promo_lag 1.034***
(0.038)
TRAC.part_disc 0.0002** 0.00000 0.00000
(0.0001) (0.00000) (0.00000)
HQD 0.092*** 0.002** 0.002**
(0.033) (0.001) (0.001)
prepost90=1 0.999***
(0.357)
F.un=1 0.451*
(0.247)
prepost901 0.020** 0.019**
(0.008) (0.008)
F.un1 0.014** 0.015***
(0.006) (0.006)
Constant -3.113***
(0.394)
Observations 457 457 457
R2 0.388
Log Likelihood 284.417 277.441
chi2 153.118*** (df = 5)
Wald Test (df = 5) 1,183.979*** 929.464***
Note: p<0.1; p<0.05; p<0.01

Model 2: Democratic visibility

  • IOs respond to demands from their environments (e.g. mediated by media salience), especially when many member states are democracies (and thus more likely to frame governance beyond the state as democratic).
    • high share of democratic members (POLITY IV > 7)
    • high visibility in NYT (highly correlates with global visibility)
    • IOs with a contentious ideology (e.g. highly politicized economic IOs) also have witnesses public protest

Dependent variable:
dummy_dem raw_roll3_dem_rhet raw_roll3_dem_no_promo raw_roll3_dem_rhet raw_roll3_dem_no_promo
logistic Tobit Tobit Tobit Tobit
(1) (2) (3) (4) (5) (6)
Constant 0.071 -10.146***
(0.476) (2.727)
dummy_dem_lag 1.910*** 0.052
(0.242) (0.405)
raw_roll3_dem_rhet_lag 1.032*** 0.822***
(0.037) (0.051)
raw_roll3_dem_no_promo_lag 0.977*** 0.861***
(0.041) (0.054)
DM.share -0.078*** -0.051 -0.001*** -0.001*** -0.001 -0.0005
(0.018) (0.038) (0.0004) (0.0004) (0.001) (0.001)
DM.m.pol 0.494*** -0.614 0.008*** 0.009*** -0.019 -0.014
(0.118) (0.566) (0.003) (0.002) (0.013) (0.012)
VIS.nyt 0.596 -0.482 0.051 0.046 -0.137 -0.041
(3.659) (4.600) (0.076) (0.071) (0.091) (0.088)
Ideo.cont 0.941*** 0.023*** 0.028***
(0.274) (0.006) (0.006)
Ideo.cont=1 -0.144
(0.689)
Ideo.cont1 0.005 0.022*
(0.014) (0.013)
vis_dem_scaled 5.269 6.196 -0.069 -0.034 0.742* 0.382
(16.743) (20.170) (0.337) (0.320) (0.389) (0.375)
ln_protest_l -0.346* -0.005 -0.005
(0.185) (0.004) (0.004)
ln_protest_isre 1.284*** 0.021*** 0.021***
(0.290) (0.005) (0.005)
mediacov_IO 0.367 0.002 0.001
(0.257) (0.005) (0.005)
MS_affin -6.639* -0.189** -0.063
(3.507) (0.075) (0.070)
democracy_ms 0.661 0.017 0.010
(0.544) (0.012) (0.012)
Constant 0.071 -10.146***
(0.476) (2.727)
Observations 457 219 457 457 219 219
R2 0.407 0.493
Log Likelihood 288.486 285.451 223.709 222.582
chi2 162.416*** (df = 6) 100.995*** (df = 11)
Wald Test 1,223.917*** (df = 6) 977.680*** (df = 6) 812.946*** (df = 11) 667.582*** (df = 11)
Note: p<0.1; p<0.05; p<0.01

Summary

  • large variance of democratic rhetoric that needs to be explained
  • explanatory models capture some aspects
    • effect of world culture on occurrence of democratic rhetoric post 1990
    • spread of democratic rhetoric in UN family and in IO HQ hot-spots
    • limited effect of visibility, democratic membership
    • contentious ideology raises democratic rhetoric

Next steps

  • engaging with crises of legitimacy
  • gather an IO protest variable
  • "unpacking" democratic rhetoric
  • thinking about an institutionalist model
  • finalizing multivariate analysis

Dependent variable:
dummy_dem raw_roll3_dem_rhet raw_roll3_dem_no_promo
logistic Tobit Tobit
(1) (2) (3) (4)
Constant -1.550*** -1.653***
(0.172) (0.316)
dummy_dem_lag 2.168*** 1.130***
(0.232) (0.328)
F.purp=general 1.461*** 0.217
(0.441) (0.529)
F.purp=political/military 0.163 2.589**
(0.398) (1.101)
F.purp=social 0.132
(0.280)
lav_rdcc 2.959***
(0.879)
raw_roll3_dem_rhet_lag 1.090***
(0.033)
raw_roll3_dem_no_promo_lag 1.081***
(0.037)
F.purpgeneral 0.023** 0.019**
(0.009) (0.009)
F.purppolitical/military -0.002 0.0002
(0.010) (0.010)
F.purpsocial -0.004 -0.005
(0.007) (0.007)
Constant -1.550*** -1.653***
(0.172) (0.316)
Observations 457 219 457 457
R2 0.347 0.343
Log Likelihood 279.457 269.838
chi2 (df = 4) 134.278*** 65.092***
Wald Test (df = 4) 1,191.357*** 923.655***
Note: p<0.1; p<0.05; p<0.01

Thank you