10
Atypically, the determinants of right-to-work laws have been fairly extensively studied in
the labor economics literature. But the purpose is again indirect: to obtain unbiased
estimates of the effects of the law on several outcome indicators. Chief among these has
been union membership, hypothesized to be reduced in the presence of right-to-work
laws either by reason of increased union organizing and maintenance costs (as union
shops cannot be used to curb free riding), or because of reduced bargaining power
(stemming from the lack of universal membership within the bargaining unit) leading to
reduced benefits from unionism and a long-run decline in membership. To the extent that
right-to-work laws mirror existing tastes for unionism and the extent of unionism there is
both an omitted variables problem and a simultaneous equations bias. The effects
literature has therefore sought in often ingenious ways to control for the nonrandom
presence of right-to-work laws (for surveys, see Moore and Newman, 1985; Moore,
1998).
Variables in the right-to-work equation have tended to reflect the simple view that
employers favor and unions oppose such legislation. Union density is found to have a
strongly negative impact on the likelihood that a state has a right-to-work law, although
distinct employer arguments do not seem to have been deployed. Other variables have
included economic development (poor states tend to adopt RTW laws to promote
growth), degree of urbanization/population density (higher values for both of which are
presumed to indicate ‘collectivist views’ facilitating the passage of legislation), and the
proportion of the workforce that is female (higher shares are supposed for various reasons
- tastes, labor force attachment, and job composition grounds - to favor right-to-work
laws).
After taking the taste effects of right-to-work laws into account or treating the right-to-
work status of states as endogenous, the point effects of the laws are often poorly
determined. The same appears to be true of fixed-effect and disequilibrium models as
well. The main exception is the fixed effect stock-adjustment model of Ellwood and Fine
(1987) who find that right-to-work laws have a sizeable initial effect on organizing that