The studies which have yielded rich insights have their own limitations,one of which arises from the narrow range of host-country variables being used.Most often,these studies incorporate a measure of host-government policies on foreign ownership into their regressions. They mainly approach institutions as determinants of FDI flows rather than as determinants of equity structures.In this type of studies,the common control variables are market size,the export orientation of the economy,infrastructural quality,and political and economic stability.3 While it is natural and commonsensical to include foreign ownership policies in the studies of foreign ownership across countries,there is no need to do so in a study of FDI in China because in the 1990s the FDI policies of China were homogenously liberal. Another limitation related to the use of host-county variables is resulted from the negligence of the fact that FDI policies is just one of the many host-country variables that affect the ownership structures of FIEs.This can be judged by the inconsistent findings of the effect of FDI policies on FDI inflows in empirical studies.Some studies find the effect to be significant while others do not and even have results contrary to anticipation.There are countries which have liberalized the FDI regimes but fail to garner much FDI which others inundated with FDI despite their highly restrictive policies.Taiwan,for example,considerably liberalized its FDI regime in the late 1980s but its FDI/capital formation ratio remained virtually unchanged in the 1990s.India undertook substantial FDI liberalization in the 1990s but its FDI inflows were a fraction of the FDI inflows that went to China,which in fact is quite comparable to India by various FDI liberalization measures.4 3 See the survey by Lim(2001)for more about the findings. 4 Huang(2003)provides a number of such examples. 44 The studies which have yielded rich insights have their own limitations, one of which arises from the narrow range of host-country variables being used. Most often, these studies incorporate a measure of host-government policies on foreign ownership into their regressions. They mainly approach institutions as determinants of FDI flows rather than as determinants of equity structures. In this type of studies, the common control variables are market size, the export orientation of the economy, infrastructural quality, and political and economic stability.3 While it is natural and commonsensical to include foreign ownership policies in the studies of foreign ownership across countries, there is no need to do so in a study of FDI in China because in the 1990s the FDI policies of China were homogenously liberal. Another limitation related to the use of host-county variables is resulted from the negligence of the fact that FDI policies is just one of the many host-country variables that affect the ownership structures of FIEs. This can be judged by the inconsistent findings of the effect of FDI policies on FDI inflows in empirical studies. Some studies find the effect to be significant while others do not and even have results contrary to anticipation. There are countries which have liberalized the FDI regimes but fail to garner much FDI which others inundated with FDI despite their highly restrictive policies. Taiwan, for example, considerably liberalized its FDI regime in the late 1980s but its FDI/capital formation ratio remained virtually unchanged in the 1990s. India undertook substantial FDI liberalization in the 1990s but its FDI inflows were a fraction of the FDI inflows that went to China, which in fact is quite comparable to India by various FDI liberalization measures.4 3 See the survey by Lim (2001) for more about the findings. 4 Huang (2003) provides a number of such examples