3.3  Variables used in the Analyses

The variables explaining the hours of work are wage and other income, which enter the utility function in Section 2 directly, and personal and household characteristics, which affect the hours of work through the preference parameters for income and work. For example, women with young children are expected to have a lower preference for paid work. Table 1 gives summary statistics of the independent variables used in the analyses (averaged over the whole period). The selection of variables is based on findings in previous studies, (a priori) expectations and the available data. The background characteristics used to specify preferences in the utility function are listed below.

Age is observed in years. Many studies include age and age squared to allow for a non-linear relation between age and the preference for leisure. Van Soest (1995), Aaberge, Colombino and Strøm (1999), Duncan and MacCrae (1999), Euwals and Van Soest (1999), and Van Soest, Das and Gong (2002) either find that age reduces the preference for leisure or a U-shaped relationship with age where the preference for leisure is reduced at first, followed by an increase in the preference for leisure after a certain age. Thus, younger and older persons are expected to have a higher preference for leisure.

Note a: These averages are somewhat different from those reported in Kalb and Scutella (2003), because a few additional observations have been excluded in the selections for this analysis. The largest difference between the two sets of summary statistics is observed for couples, because complete information on all variables is needed for both members of the couple whereas this was not necessary in the wage and participation rate analysis.

b: 2.0753 when only couples with children are included.

Education is divided into the following categories:

• no qualifications
• school certificate
• bursary, scholarship or University Entrance (UE), which all three refer to qualifications obtained by 18yr olds (usually) at the end of secondary school.
• vocational or trade qualifications
• part degree or other qualification
• bachelor’s degree or diploma
• postgraduate qualification

Comparing the education levels of the five demographic groups, we observe that married men and single women have the largest proportion with at least a bachelor’s degree and the lowest proportion of persons without qualifications. Sole parents have the lowest level of education. Education is expected to increase the preference for work, because time and money have been invested in human capital. Apart from the financial rewards, one would also expect a high-skill job to be more interesting than a low-skill job and hence more desirable. In accordance with the above expectation, Duncan and Harris (2002) find that having some qualifications is associated with an increase in the preference for labour supply and Duncan and MacCrae (1999) find that leaving school at 16 years of age is associated with a decrease in the preference for labour supply. Murray (1996) similarly finds that sole parents with some form of post-secondary school qualifications have a higher preference for work.

The number of dependent children in each income unit consists of the number of dependent children from 0 to 18 years old. This variable is expected to be of particular importance for females, married or single. Children are likely to increase the value of time at home, which is reflected in a higher preference for leisure in the model.

From the surveys, the age of the youngest dependent child in the income unit can be derived. The effect of dependent children on the preference for time spent in work is likely to be bigger when young children are present.

The expected effects with regard to children are found in several studies. The effects are strongest for women. Van Soest (1995) finds effects for both men and women, where the female effects are somewhat larger. Van Soest, Das and Gong (2002), Aaberge, Colombino and Strøm (1999), Fraker and Moffitt (1988), Hagstrom (1996) and Hoynes (1996) find effects for married women. Duncan and MacCrae (1999) find strong effects for sole parents (mostly women) and married women of both the age of the youngest child and the number of preschool children. Much lower (and often no) effects are found for men. Similar effects are found for sole parents in Australia (Murray, 1996).

Region of residence are location variables for where the individual/household lives in New Zealand. It is expected that the fixed cost of working is different for people in or outside the larger cities and in or out of more remote areas, in particular for people with children who may need childcare services (Duncan and Harris, 2002).

Living with parents is an indicator variable used for singles and sole parents to indicate that they live with their parent(s). For sole parents, this could indicate a childcare opportunity that would enable them to work. For this reason, we expect this variable to influence the preference for work. In the case of singles, the cost of living is much lower when living with their parents and additional income may be received from the parents. As a result, it is expected that the preference for income is lower for singles living with their parents.[14]

The survey period includes an interesting policy change. In 1991, at the start of the survey period eligibility for the New Zealand Superannuation started at 60 years of age. During the ten years this age of eligibility gradually increased until it was 65 at the end of the survey period in April 2001. We expect this delayed eligibility to affect the preference for work. For this reason, an indicator variable eligibility for New Zealand Superannuation is included in the preference for work parameter. This change in eligibility only affects individuals who are aged between 60 and 65 years. Although a linear and quadratic term for age are already included, an additional dummy for age is introduced for those aged 60 or over to avoid picking up the effect of being in this age category when examining the effect of the changed eligibility.

Given the long time period over which data are available and the changes, in particular for female labour force participation, which occurred over this time, the inclusion of a time trend in the work preference parameter may capture these changes. It may also allow us to more accurately measure the effect of eligibility for New Zealand Superannuation. However, at the same time business cycles are occurring which may influence the availability of work. Although strictly speaking the unemployment rate does not belong in the labour supply model, it can be argued (particularly for sole parents and married women) that high unemployment discourages individuals trying to find work. We believe that including an unemployment measure (by gender and quarter) assists to identify a purer time trend.

Finally, men and women are expected to have different preferences for “leisure” time. In the models for two-adult income units, person 1 is male and person 2 is female. The two-adult income units containing two adult men or two adult women have been excluded from the analysis. In the single-adult income units, models are estimated separately by gender. For sole parents the male group is too small to estimate separate models, therefore a dummy variable for gender is included in the preference for labour supply and income, and in the fixed cost parameter, to explore whether gender affects the preferences in this group.