Table 12. Fixed-effects regression results (dependent variable: Gini
coefficient of distribution of disposable income/c)
Source: author’s calculations. Notes: , , , ^ significant at 1%, 5%; 10%; between 10 and
15%. a/ In the fixed effect estimation procedure, this time-invariant variable is omitted and is
subsumed in each country’s constant.
(vi) Finally, the dummy variable ‘LOC’ is significant and indicates
that left of centre governments tend to have, on average, a Gini
coefficients lower by around one point than NO-LOC countries, in
addition to the effects mediated by the adoption of more
progressive economic social policies. The statistical fit of the
regression is broadly satisfactory. Yet, these results need to be
probed further, and have to be considered as an initial step in
disentangling the sources of the inequality decline observed from
2003 to2007. For instance, the validity of the conclusions drawn on
the basis of these results needs to be probed in a few ways, starting