generally a family. In this regard, the mechanism is similar to that posited by the theories of
commitment (Becker 1960) and efficiency wages (Shapiro and Stiglitz 1984). These bonds
are built of attachments to particular individuals, commitment to specific interests,
involvements in particular activities, and the internalization of beliefs about proper behavior.
As in the example of the diamond merchants, much more than economic transactions or
relationships are made to be at stake by social embedding and, as in the situations we explore
below, a live-in worker could lose home, friends, and a way of life in addition to a source of
income. Moreover, to the extent that the social relationship is part of a larger community, the
possibilities for alternative partners would be very limited.
The basic predictions of the influence argument are that exchange partners,
particularly those with sensitive functions, will be socially embedded. In contrast to the
previous argument where they partner choice was correlated to individual attributes, the
behavioral effects of social influence are orthogonal to individual human capital attributes.
Moreover, because both types of social embedding – based on current common activities or a
history of interaction – should accomplish the same end, the respective forms of social
embedding should substitute for each other.
Insurance against environmental contingencies
The first two arguments about social embedding focus on the operational issues
facing businesses in the (mutual) selection and management of partners in a few critical types
of exchanges. Relying on social embedding may reduce search costs and ameliorate
malfeasance problems. Yet immediate operations are not the only concerns of businesses
and, therefore, the preceding discussion may offer only a partial view of social embedding. It
is often advantageous to build a political coalition to protect one’s economic position