Results

Table 7.1 presents the main results of our explorative analysis. We find strong evidence supporting hypothesis 1. Both variables (columns 1 and 2) have statistically significant coefficients and a negative sign, indicating that higher levels of government effectiveness and regulatory quality are associated with less coercive government behavior in crises. We also find indication that debtor countries with a better rule of law and less corruption show a more cooperative (less coercive) stance on average. This gives some support to hypothesis 2. In contrast, we find no significant effects with regard to “Voice and Accountability” and, surprisingly, no significant coefficient for the indicator “Political Stability and Violence”

One obvious concern with the results in table 7.1 is that we do not control for the level of economic development and the degree of debt distress of the sovereign. A country’s per capita income and its level of indebtedness are likely to be crucial for crisis resolution. We therefore gather data for these two variables from the World Bank’s World Development Indicators. More specifically, we use per capita GNI (gross national income) in PPP terms (purchasing power parity) and the share of total external debt to GNI and construct their average values for the years 1980 to 2005. As stated, we also include time dummies on an annual level.

Table 7.2 shows that our results hold with time-fixed effects and even when controlling for the level of indebtedness and economic development of countries. This gives some support to our findings mentioned earlier. Only the indicator on “Control of Corruption” turns borderline significant (at the 10% level only). Overall, we therefore find governance to matter, especially with regard to regulatory quality, the rule of law, and the state’s capacity to implement effective policies.

A caveat is in order at this point: while we do unveil interesting correlations between our index and the set of governance indicators, the results should not be misinterpreted. We have not shown any causal effect nor have we identified specific mechanisms for our findings. In particular, it should be highlighted that the governance indicators are outcome variables and partly based on expert opinions. This raises the concern of reverse causality, as debtor governments that adopt a particularly aggressive stance toward Western creditors may also be perceived and evaluated as having lower levels of governance. Additionally, there might be measurement error. One explanation for the insignificant effect of “political instability and violence” could, for example, be that these factors have strong short-term effects, which are clouded by our use of time-variant explanatory variables. Further analysis, possibly on a year-by-year level, is needed to address some of these concerns.

Government

Effectiveness

(eoff/se)

Regulatory

Quality

(eoef/se)

Rule of Law (eoff/se)

Control of Corruption (eoef/se)

Voice and Accountability (eoff/se)

Political Stability and Violence (eoff/se)

Government

Effectiveness

  • -0.492**
  • (0.210)

Regulatory

Quality

  • -0.380*
  • (0.211)

Rule of Law

  • -0.446**
  • (0.193)

Control of Corruption

  • -0.449**
  • (0.199)

Voice and Accountability

  • -0.301
  • (0.205)

Political Stability and Absence of Violence

  • -0.196
  • (0.193)

/cutl

  • -1.082***
  • (0.171)
  • -1.133***
  • (0.170)
  • -1.014***
  • (0.173)
  • -1.037***
  • (0.168)
  • -1.143***
  • (0.170)
  • -1.038***
  • (0.176)

/cut2

  • -0.200
  • (0.152)
  • -0.261*
  • (0.156)
  • -0.131
  • (0.158)
  • -0.150
  • (0.152)
  • -0.279*
  • (0.155)
  • -0.180
  • (0.167)

Government

Effectiveness

(coff/se)

Regulatory

Quality

(coef/se)

Rule of Law (coff/se)

Control of Corruption (coef/se)

Voice an d Accountability (coff/se)

Political Stability and Violence (coff/se)

/cut3

  • 0.116
  • (0.149)
  • 0.051
  • (0.149)
  • 0.187
  • (0.157)
  • 0.170
  • (0.151)
  • 0.031
  • (0.154)
  • 0.128
  • (0.167)

/cut4

  • 0.588***
  • (0.164)
  • 0.517***
  • (0.158)
  • 0.658***
  • (0.173)
  • 0.644***
  • (0.174)
  • 0.491***
  • (0.163)
  • 0.582***
  • (0.180)

/cut5

  • 0.918***
  • (0.169)
  • 0.842***
  • (0.161)
  • 0.987***
  • (0.183)
  • 0.975***
  • (0.190)
  • 0.815***
  • (0.161)
  • 0.901***
  • (0.184)

/cut6

  • 1.372***
  • (0.201)
  • 1.290***
  • (0.196)
  • 1.439***
  • (0.205)
  • 1.425***
  • (0.219)
  • 1.266***
  • (0.194)
  • 1.346***
  • (0.205)

/cut 7

  • 1.748***
  • (0.189)
  • 1.662***
  • (0.185)
  • 1.820***
  • (0.181)
  • 1.800***
  • (0.195)
  • 1.639***
  • (0.187)
  • 1.718***
  • (0.186)

/cut8

  • 1.987***
  • (0.264)
  • 1.900***
  • (0.257)
  • 2.066***
  • (0.270)
  • 2.040***
  • (0.279)
  • 1.878***
  • (0.253)
  • 1.959***
  • (0.264)

Number of observations

100

100

100

100

100

100

Results of ordered probit estimation. The dependent variable indicates the degree of government coerciveness related to each sovereign debt restructuring on a scale of 1 (very low coerciveness) to 10 (very high coerciveness). ***/**/* denote significance at a 1/5/10 percent level respectively. Country-clustered standard errors in parentheses.

table 7.2 Robustness Checks

Government

Effectiveness

(coff/se)

Regulatory

Quality

(eoef/se)

Rule of Law (eoef/se)

Control of Corruption

Voice and Accountability (coff/se)

Political Stability and Violence (eoef/se)

GDP per capita (in PPP)

  • 10.801*
  • (6.004)
  • 0.717
  • 6.652
  • (5.960)
  • 0.512
  • 9.265
  • (6.204)
  • 0.954
  • 6.738
  • (6.432)
  • 0.494
  • 2.786
  • (8.742)
  • 0.256
  • 5.360
  • (7.727)
  • 0.483

Ext. Debt / GDP

(1.070)

(1.076)

(1.111)

(1.102)

(1.124)

(1.161)

Year Fixed Effects

Yes

Yes

Yes

Yes

Yes

Yes

Government

Effectiveness

  • -0.830***
  • (0.320)

Regulatory Quality

  • -0.616**
  • (0.282)

Rule of Law

  • -0.690**
  • (0.317)

Control of Corruption

  • -0.570*
  • 0.320)

Voice and Accountability

  • -0.300
  • (0.429)

Political Stability and Absence of Violence

  • -0.422
  • (0.366)

Number of observations

80

80

80

80

80

80

Results of ordered probit estimation. The dependent variable indicates the degree of government coerciveness related to each sovereign debt restructuring on a scale of 1 (very low coerciveness) to 10 (very high coerciveness). ***/**/* denote significance at a 1/5/10 percent level respectively. Debt/GNI and GDP p.c. are averaged for the period 1980-2005. Country-clustered standard errors in parentheses. Coefficients of cutpoints not reported.

 
Source
< Prev   CONTENTS   Source   Next >