and in 14 out of 19 sectors the difference is statistically significant. A similar result
emerges from comparing efficiency differences between incumbents and exits in
column (2). Our results therefore show that exit is concentrated among the least efficient
firms, consistent with the findings of for example Aw et al. (2001) for Taiwanese firms
in the 1980s. However, focusing on efficiency differences between “real” entrants and
“real” exits in column (3) reveals that exits have on average higher efficiency levels
than new entrants. This is so in 15 out of 19 cases - and in five sectors the difference is
statistically significant. This suggests that the standard finding that firm entry and exit
contribute to a more efficient reallocation of resources (i.e. transferring resources from
less efficient to more efficient producers) may not hold in the case of the Vietnamese
manufacturing sector. 27 Switchers certainly merit separate investigation.
Columns (4) and (5) in Table 8 compare sector switchers with incumbents. Our findings
confirm that incumbents produce more efficiently than sector switchers (both “IN” and
“OUT”), in line with the predictions of traditional life-cycle theories. The two
remaining columns (columns (6) and 7)) compare mean efficiency differences between
sector switchers and entrants and exits, respectively. In nine out of 19 sectors we find
that sector switchers are (on average) significantly more efficient producers than “real”
new entrants. This is also the case for three sectors when sector switchers are compared
to exits, even if less by way of statistical significance emerges from this comparison.
Overall, our results indicate that significant efficiency differences exist between
different types of firms, and the positive contribution to growth from firm dynamics is
more associated with sector switchers than with standard turnover. Pooling switchers
with new and closed down firms (i) is potentially misleading in coming to grips with
relative levels of efficiency, and (ii) would clearly constrain our understanding of the
underlying process of development and the potential contribution of firm dynamics to
productivity growth. On this background we turn to analysing the determinants of firm
exit and switching.
27 For example, results for Chile and Columbia find that inefficient plants are replaced with slightly more
efficient plants, increasing overall productivity in the sector albeit by a small amount (Tybout, 2000).
18