In their article in this issue, Donald P. Green, Soo Yeon Kim, and David H.
Yoon claim, contrary to liberal theory and extensive evidence, that neithe
r joint democracy nor economic interdependence significantly reduces the fr
equency of militarized interstate disputes in pooled time-series analyses w
hen dyadic fixed effects are taken into account. Similarly, their fixed-eff
ects analyses contradict theory and previous evidence that democracies have
higher levels of trade with one another than do other types of stares. Our
reexamination, however, refutes both claims and reinforces previous findin
gs. Their fixed-effects analyses of disputes produces distorted results bec
ause they consider a relatively short time period, 1951-92. in which variat
ion in the binary dependent variable and the key independent variables, dem
ocracy and trade, is limited. When we analyze a longer period (1886-1992),
the results confirm liberal theory. The differences between our analyses of
bilateral trade and those of Green, Kim. and Yoon primarily arise from a s
eemingly minor methodological decision. A more reasonable method confirms t
hat democracies do have higher levels of trade than expected on purely econ
omic grounds. Though we do not advocate a fixed-effects model for analyzing
these data and have serious reservations about its general usefulness, our
findings provide additional confirmation of liberal theories of internatio
nal relations.