The seminal work of Jovanovic (1982) considers a passive learning model in which
information is gathered at no cost. Firms entering a new product market do not know
their exact cost structure and assuming that firms differ with regard to efficiency, they
incur different costs when producing the same levels of output. Entrants do not know
their exact abilities (productivity) so their performance is unknown, and each participant
has to go through a learning process, accumulating information from actual market
experience. Gradually firms may discover whether their abilities meet prior
expectations, and if not they exit. Consequently, efficient firms survive and experience
growth, whereas over-optimistic firms eventually switch sector or close down. The
longer a firm has been in a sector the more knowledge it has about its own abilities,
suggesting that the probability that a firm switches is negatively related to firm age.
Accordingly, we hypothesize that the probability of a firm switching sector is
decreasing in firm level efficiency and age as well as size, noting that the motivation for
including size is well established in the firm dynamics literature
Ownership structure may also influence sector switching, even when firm specific
efficiency, age and size are controlled for.4 Wen et al. (2002) study the reforms of state
owned enterprises (SOEs) in China and note that responsibility for many production
decisions have gradually been decentralized to individual firms. However, although the
importance of central planning departments and committees has been shrinking,
decisions about the industries in which SOEs should engage continue to be made at
central level. This political hierarchy in SOE management structure is likely to limit
inter-sector dynamics, so we expect SOEs to switch sector less frequently. At the same
time, the ongoing reform/privatization process in Vietnam appears to close down
relatively many SOEs. We therefore hypothesize that there is a positive association
between state ownership and firm closure in the exit specification at firm level.
Foreign owned enterprises, or enterprises with some foreign participation, are also
expected to be more “locked into” specific sectors due to the legal constraints. Until
4 Choices are involved in classifying firms by ownership type as discussed in the data appendix, and there
are “grey” areas involved in distinguishing between state and foreign owned firms. This is however
unavoidable in the type of analysis put forward here. As regards mergers these are very limited in number
as this process is just starting in Vietnam. As such merging is of no quantitative significance to our
results. In most mergers and acquisitions firms kept at least one business registration licence and one tax
code registration. This means that one firm stays in our data as an incumbent and the rest of the merger
firms exit. Note also that mergers and acquisitions are often followed by sector switch.