Compensating differentials for nurses

Lenke til artikkel:



Sæther, Erik Magnus




HERO Skrifter
Nummer i serie: 10

Sammendrag (engelsk)

When entering the job market registered nurses (RNs) face job alternatives with differences in wages and other job attributes. Previous studies of the nursing labor market have shown large earnings differences between similar hospital and non-hospital RNs. Corresponding differences are found in some of the analyses of shift and regular daytime workers. In the first part of this paper I analyze the wage differentials in the Norwegian public health sector, applying a switching regression model. I find no hospital premium for the shift RNs and a slightly negative hospital premium for the daytime RNs, but it is not significant for the hospital job choice. I find a positive shift premium. The wage rate is 19% higher for the shift working hospital RNs and 18% for the sample of primary care workers. The shift premium is only weakly significant for the shift work choice for the sample of hospital RNs, and not for the primary care RNs. I identify some selection effects. In the second part of the paper I focus on the shift compensation only, and present a structural labor supply model with a random utility function. This is done to identify the expected compensating variation necessary for the nurses to remain on the same utility level when they are “forced” from a day job to a shift job. The expected compensating variations are derived by Monte Carlo simulations and presented for different categories of hours. I find that on average the offered combination of higher wages, shorter working hours and increased flexibility overcompensates for the health and social strains related to shift work.


Registered nurses, compensating variations, switching regression, random utility models, discrete choice, shift work, labor supply


C25, I10, J31, J33

Prosjekt info:

Oppdragsgiver: Norges forskningsråd
Oppdragsgivers prosjektnr.:
Frisch prosjekt: 4101 - Økonomi og helse






Norges forskningsråd