overall volume of cross-holdings for a given country in a given year. The variable p2
measures instead the volume of cross-holdings in equity.
4. Econometric results
4.1 Financial openness and per-capita income-gaps.
The results for the income gap equation are reported in Table 1. The estimates for the
group of emerging economies are shown in columns I to IV; the estimates for the full
sample of 44 countries are displayed in columns V to VIII.
Starting with emerging economies, column I reports the baseline specification of
equation (3) without dk. The estimated coefficient on dz is positive and statistically
significant. This means that the more a country approaches the degree of financial
openness of the EU-15, the smaller its income gap relative to the EU-15 average will
be. The effect holds over and above the impact of differences in domestic financial
depth and in the rates of factors accumulation. Columns II and III show the same
baseline equation re-estimated with measures of gaps in the degree of international
financial integration (dp1 and dp2). The evidence is complementary to that in column
I: countries that fall behind the EU-15 average in terms of their degree of international
financial integration tend to experience greater income gaps. Finally, in column IV the
gap in physical capital accumulation, dk, is included. A couple of interesting findings
emerge. The coefficient on dz remains positive and different from zero; actually, it
substantially increases, whilst the coefficient on dq decreases (even though it remains
statistically significant at high confidence levels). Moreover, the effect of gaps in
investment rates appear to be marginally very small. Taken all together, these findings
suggest that in emerging economies financial openness impacts catching-up not only
through its effect on the level of the investment rate in the economy and/or its
contribution to domestic financial development. The discussion in Section 2 has in
fact emphasised other possible channels, including investment-composition effects
and productivity/technological spillovers.
Turning to sample of all countries, it is evident that results are qualitatively very
similar to that reported for the group of emerging economies. If anything, the
marginal effects of both financial openness and domestic financial development
appear to be larger, as denoted by the point estimates in column V. Interestingly, the
inclusion of dk also leads to an increase in the coefficient on dz and a
contemporaneous decrease in the coefficient on dq. However, differently, from the
12