where TFP_max is the maximum TFP level indicating the technology frontier in the industry of
firm i. A high level of relative efficiency is assumed to be close to that of the industry leaders.
Second, we divide our subsample of domestic firms into three subsamples: the top 20 percentage
of the levels of technology gap (5th quintile), between the 20 and 80 percentage (2-4th quintile),
and below the 20 percentage (1st quintile). We expect the spillovers from the foreign presence to
have more influence on domestic firms in the second subsample than in the first and third
subsamples. The estimation with the first and third subsample enables us to investigate the cases