development of social capital entails costs, alternative mechanisms may be attractive options.
Vietnam is a particularly appropriate research site because the combination of a state
apparatus still in need of reform and rapid economic growth suggest that the use of informal
mechanisms of economic regulation, such as social embeddedness, would be widespread.
Our empirical explorations suggest that the problems of partner choice, performance
enforcement, and environmental contingency are real and that social embeddedness is used,
to varying degrees, to address all three concerns. We uncovered a more complex situation
than is sometimes asserted. That outcome should come as no surprise given that the three
arguments about social embeddedness that we investigate – and others are possible – make
somewhat contradictory predictions. Briefly stated, we find the following:
1) The level of embeddedness is substantially higher than that implied by the “arm’s
length” relationships of textbooks but less than some descriptions would suggest. We find a
moderate level of embedding among the external business relationships, most owners had
contact with their partners several times a month or less and that the level of contact was
driven by operational need (suppliers are seen most often). Employees are fairly well
socially embedded into businesses but still over 40 percent do not eat meals at the business
and approximately 15 percent do not socialize (the most common social activities).
Unfortunately, since ours is the only dataset that we know of that has attempted to
systematically measure the level and pattern of social embeddedness in a random sample of
small businesses, we are unable to compare our findings to other regions.
2) Families (socially embedded actors) are not especially reliable sources of help.
Approximately two-thirds of our respondents received no internal or external help from their
relatives. Even when family members were nearby, help was forthcoming in only about half
the cases. Family members serve as a source of help more frequently than a baseline