notches. As we mentioned, a reduction of the unemployment only affects the rating if
this reduction is sustainable. If that is the case, a reduction of 4 percentage points
increases the rating by 0.2 to 0.35 notches. The effect of inflation is quite small, a
reduction of 20 percentage points on inflation increases the rating by 0.05 to 0.1
notches. These values are too small to be noteworthy for industrial countries, but if one
does the same calculation with the value estimated for high rated countries a reduction
of 4 percentage points in inflation would increase the rating by 0.15 notches.
Now that we have an idea of the estimated impact of the variables we can do some
specific country analysis. As an example, in Table 17 we show the rating for some
European countries and some emerging markets both in 1998 and 2005. Then, we use
the estimated short-run coefficients of the random effects ordered probit together with
the values for the relevant variables. Afterwards, we divide the overall prediction
change in the rating for each agency into the contributions of the different blocks of
explanatory variables: macroeconomic performance, government and fiscal
performance, external elements and European Union6. The upper and lower bound
presented are computed by adding and subtracting one standard deviation to the point
estimate of the coefficients.
[Insert Table 17 here]
Let’s compare, for instance, Portugal and Spain. In 1998 they both had an AA (Aa2)
rating but in 2005 while Spain had been upgraded to AAA (Aaa) by all agencies,
Portugal had been downgraded by S&P. If we analyse the contributions of the main key
variables we see that, for Portugal the positive contribution of the macroeconomic
performance was overshadowed by the negative government developments. For such
government performance contributed the worsening of the budget deficit since 2000, the
upward trend in government debt and the worsening in the World Bank governance
effectiveness indicator. As for Spain, the good macroeconomic performance was the
main cause of the upgrade, specially the reduction of structural unemployment since the
mid nineties and the increase of GDP per capita due to the persistent high growth.
6 As an exception, we used the long-run coefficient of unemployment instead of the short-run coefficient.
We are implicitly assuming that all the changes in unemployment between these two years were
structural.
27
More intriguing information
1. The name is absent2. Expectation Formation and Endogenous Fluctuations in Aggregate Demand
3. The name is absent
4. Expectations, money, and the forecasting of inflation
5. Who runs the IFIs?
6. Midwest prospects and the new economy
7. On the Relation between Robust and Bayesian Decision Making
8. Consumer Networks and Firm Reputation: A First Experimental Investigation
9. Evolution of cognitive function via redeployment of brain areas
10. The name is absent