4.2.1. Sample composition
In Panel A of Table 3 we examine whether our main results above are driven by the uneven distribution of observations across
the U.S. and Japan, which contribute a large fraction of observations to our sample firm-years. First, we repeat our baseline
regression after excluding firms from Japan (Model 1), the U.S. (Model 2), and both Japan and the U.S. (Model 3). Second, we use
weighted least squares, where a country's weight is equal to the reciprocal of the number of observations for that country (Model
4). Third, we estimate a country-level regression in which the dependent and independent variables are averaged at the country
and year levels (Model 5). The results consistently show a negative and statistically significant (at the 1% level) coefficient on
Creditor Rights, indicating that the significantly negative effect of creditor rights on long-term debt that we document is not driven
by the uneven distribution of observations across countries.