is included. Finally, dummy IMF stands for the role of the IMF in the debt
rescheduling process and it is expected to be positive and signi...cant.
In sum, regarding the external factors, we should .nd a positive correlation
between the dependent variable and both EDT and TDS and a negative relation
between DRES and the variables BOP, CA, EXP and RES. The sign is instead
expected to be positive for the coe^ients of both IAR and PAR. As the domestic
factors are concerned, per capita GDPPC, INFL and GDI should have negative
coe^ients (as a consequence of the debt overhang e∏ect and also because we are
dealing, generally, with countries in troubles). We expect to .nd a signi.cant and
positive coe^ient for the dummy BB, meaning that those countries which have
adhered to these plans are more likely than others to obtain a rescheduling of
their debts.
This equation did not include any of the so called “balance sheet” variables
(see Section 2.2). This is the case because we are more interested in more fun-
damental, longer-term determinants of a country’s solvency and macro-variables
are better proxies for this type of information, while .nancial variables tell more
about a country’s current liquidity. The only two variables which give an indica-
tion on a country’s.nancial situation we considered are the interest and principal
arrears on long-term debt.
As dummies variables are concerned, we have no regional dummies (in earlier
regressions we have actually tried to insert them, in both equations, but they were
not signi.cant). We included a dummy for Backer and Brady plans countries
(BB), that we did not .nd in the related literature.19
19In a b road er sen se, BB could also be interpreted as a dummy for middle-income countries.
25