Page 35 - Contributed Paper Session (CPS) - Volume 7

CPS2021 Noor Ismawati et al.
= A series of dichotomous variables indicating marital
status of the respondent;
= A series of dichotomous variables indicating level of
educational attainment;

= A dichotomous variable indicating whether the
respondent lives in urban area;

= A series of dichotomous variables indicating state of
residency;
= A series of dichotomous variables indicating occupation;
= A series of dichotomous variables indicating industry of
employment;
= The disturbance terms.

3. Result
In total, 32 regressions were generated for employees in public and private
sector by gender. Due to limitation of the paper, Table 1 shows the regression
output for female employees in both sectors for each separate year. In 2010,
the model indicates that all 10 predictors explain 56 per cent of the variance
in the sample of 6 218 female employees working in public sector. Age and
Age Squared are highly significant (p-value<1%) determinants of earning.
Chinese earns significantly (p-value<10%) as compared with Bumiputera. On
the other hand, Indians earn lower and other ethnic earns higher but the
difference is not significant. Data in 2010 of the female employees in public
sector also indicates that those who are not married earn significantly lower
(p-value<1%) as compared with the married employees. However the
coefficient of never married is higher as compared with
widowed/separated/divorced. The model also shows those having secondary
or tertiary education significantly (p-value<1%) earn more as compared with
those with no formal education or with primary education. The coefficient of
secondary is lower than tertiary. This suggests that higher education
attainment positively affects earning.
Comparison between years shows that age, age square and educational
attainment remain significant determinants (p-value<1%) of earning among
female employees in public sector. The coefficients of the Chinese’s female
employees in public sector are positive every year but the significant level
varies throughout the year. The coefficients of Indians and other ethnic groups
also vary throughout the year. It is observed that the Indians and other ethnic
groups earning are significantly different as compared with Bumiputera in
2013. As shown in Table 1, being unmarried is significantly (p-value<1%)
negatively effecting earning. Those who are never married remain to earn

24 | I S I W S C 2 0 1 9