Conclusion.
The estimates reveal evidence that the commonly applied mean regression model
conceals a more complex pattern of returns to labor market institutions and worker
characteristics. Quantile regression makes it possible to explore how public sector
employment, unions, firm size, worker education and experience affect wages at different
points of the conditional wage distribution in Panama.
For men as well as for women, public sector employment and employment by a large
firm increase wages most at the lower quantiles of the wage distribution. Men at the
highest quantiles who work for the public sector suffer a substantial wage penalty.
Presence of a union at the workplace primarily increases wages of men at the lower
quantiles. Union membership and collective bargaining mainly affect wages of men in
the middle quantiles. On the whole, unions boost wages of the median male worker in
the union sector and reduce wage inequality within that sector. The same is true for
public sector employment and employment by a large firm. Employment in the Canal
Zone increases wages of men and women uniformly except at the top quantiles. Higher
education and experience for men have higher effects at higher quantiles implying that
higher education and ability and on-the-job training and ability are complements. The
returns to secondary education for men have a u-shaped pattern across quantiles. The
effects of education and experience on wages do not differ across quantiles for women.
Union presence does not increase women’s wages but it reduces wage inequality.