Public-PRIVATE Pay Differentials
299
market. Of these, roughly 22% of men and 41% of women are employed in the
public sector. Appropriate cross-section weights have been applied to enable
separate analysis for Scotland only.
The total number of females employed in the public sector follows a u-shape,
the number for men fell since 1996, peaked in 1999 before dropping in the following
year and has risen again since. As one would expect, the latest increase in male
public sector employment is mainly driven by increases in civil service and local
government jobs in the aftermath of devolution and the instalment of the Scottish
Parliament. This development is also present among females although it is less
pronounced. Unsurprisingly, the percentage of women working in health care and
higher education is far higher compared to men. On the other hand, men dominate
jobs in the central governments as well as in the civil service. However, the largest
share of public sector employment for both men and women stems from local
governments.
Furthermore, public sector workers are selected differently into occupations.
For both men and women, public employees are more likely to be found in
managerial occupations compared to their private counterparts. This is especially
pronounced for females where 63% are employed in this occupation compared to
only 31% in the private sector. Additionally, public workers are more likely to be in
professional rather than in unskilled occupations. However, whether the above
patterns are supply or demand driven is a priori not clear.
Figure 1 depicts the unconditional log wage distribution in the public and
private sector by gender. Following the usual convention in the literature (e.g.
Booth at al. 2003) wages are expressed in log hourly gross pay.2 As Figure 1
shows, the male pay distribution in the public sector is less spread compared to
the private sector distribution though there is a significant overlap. Furthermore,
the mean wage in the former is higher than in the latter. The respective mean log
wages in the public and private sector are 2.2 and 2.0. A simple “difference in
means” test reveals that the pay in the public sector is indeed significantly higher
than in the private sector even on the 1% significance level. In contrast, the overlap
in the female public and private earnings distribution is much smaller compared to
males. Public pay is systematically higher in the upper end and much flatter at the
lower end of the distribution compared to the private one. Once again, the difference
in means between public (2.15) and private (1.74) sector is positive and significant
2 The log hourly gross wage rate is defined as w = ln[ Paygu / ((30/7) (Hs+αHot)) ], where
Paygu is the monthly gross pay in the current job, Hs is standard weekly hours, Hot is paid
overtime hours per week and α is the overtime premium set to 1.5.