6  Regression results on hourly earnings at age 21

We know from Tables 4 and 5 that Maori workers in the CHDS receive higher hourly earnings, on average, than their non-Maori counterparts at age 21. Yet, these ethnic differences in mean wages are not statistically significant. In this section, we use regression analysis to control for personal and family background factors that we know, in many cases, vary significantly between Maori and non-Maori. Once we control for work experience, formal qualifications and other factors, do Maori workers receive significantly higher wages than non-Maori? Does the ethnic wage gap widen when we control for other relevant factors? Could the relatively higher wages of Maori workers be the result of a sample selection process, where Maori facing lower wages are unable to find employment?

Empirical findings in the previous sections suggest that Maori accumulate less work experience by age 21 than non-Maori, and that these differences can only partly be explained by observable differences in personal and family background characteristics. We now ask whether Maori face systematically lower wage rates in the labour market than their non-Maori counterparts, once we control for factors that proxy for individual differences in productivity.

6.1  Distributions of hourly earnings by ethnicity

The software package STATA is used to estimate the probability density functions of hourly earnings for workers at age 21 using a kernel smoother. Figures 5 and 6 display the results for the two and three-way splits in youth ethnicities. Note that only the youth working and reporting hourly earnings at the time of the survey at age 21 are used in these calculations. This subsample consists of 671 individuals.

Table 4 shows that the mean wage of Maori workers ($9.44) is higher than the mean wage of non-Maori workers ($9.26). This difference is not statistically significant. Figure 5 shows that these ethnic differences in hourly earnings come from a relatively flatter density function for Maori. The non-Maori distribution has a sharper peak at around $7.50. The middle of the Maori distribution is placed slightly further to the right compared to the non-Maori distribution. Both density functions have similar slowly tapering right-hand tails beyond a wage of about $15.00. These skewed right-hand tails are common in wage distributions. The numbers at the bottom of the table, for example, show that 14.1% of Maori workers receive hourly earnings in excess of $12.00. The same is true of 13.8% of non-Maori workers.

Figure 5 – Hourly Earnings of Workers at Age 21: Two-Way Split In Youth Ethnicities

Notes: These data are taken from the 671 respondents in the CHDS who provided valid information for the purposes of this study. The probability density functions are estimated using a kernel smoother in STATA. These are the estimated distributions of hourly earnings reported by workers at the age 21 survey for the two ethnic groups. Youth are defined as ‘Maori’ in this figure if they identify Maori as at least one of their ethnicities at age 21, and had at least one parental figure claiming Maori ancestry by age 14. All other youth are considered to be ‘non-Maori’.

The differences in the hourly earnings density functions for sole and mixed Maori (relative to non-Maori) are depicted Figure 6. Table 5 shows that only the mean wage of mixed Maori ($9.71) is higher than that of non-Maori ($9.26). Yet, none of these ethnic differences in hourly earnings are statistically significant. Figure 6 shows that this relatively higher wage for mixed Maori is the result of the central section of this distribution being positioned slightly further to the right relative to the other two distributions. Given the relatively small sample sizes for sole and mixed Maori, differences in the right-hand tails of the three density functions are difficult to distinguish.

Figure 6 – Hourly Earnings of Workers at Age 21: Three-Way Split In Youth Ethnicities

Notes: These data are taken from the 671 respondents in the CHDS who provided valid information for the purposes of this study. The probability density functions are estimated using a kernel smoother in STATA. These are the estimated distributions of hourly earnings reported by workers at the age 21 survey for the three ethnic groups. Youth are defined as ‘Sole Maori’ in this figure if they identify Maori as their only ethnicity at age 21. Youth are defined as ‘Mixed Maori’ if they report both Maori and any other ethnicity, and had at least one parental figure claiming Maori ancestry by age 14. All other youth are considered to be ‘non-Maori’.