In recent years abundant evidence has been put forth to show that some
thing about a country's trade policy stance improves economic performa
nce. However, less examined is the question of what exactly that trade
policy something that matters for performance is. Examination of the
link between various empirical indicators used in the literature to me
asure trade policy stance reveals that, with minor exceptions, they ar
e pairwise uncorrelated. This finding raises obvious questions about t
he their reliability in capturing some common aspect of trade policy a
nd the interpretation of the empirical evidence on economic performanc
e.