5.4 Short and long regressions: Three-way split in youth ethnicity
Column 1 of Table 13 reports the parameter estimates from the short regressions using the three-way split in ethnicity. Non-Maori continue to serve as the excluded or reference group (as they did in the earlier two-way split in ethnicity). Although the estimated coefficients are negative for both sole and mixed Maori, only the coefficient estimate for mixed Maori is statistically different from zero at conventional test levels (1.4% level in this case). Once other personal characteristics and circumstances have been held constant, there is no statistical difference between the accumulation of work experience by age 21 between sole Maori and non-Maori.[22] As a result, we focus in the remainder of this section on the observed gap in work experience between mixed Maori and non-Maori.
| Without Background Factors | With Background Factors | |
|---|---|---|
| Constant | 0.475**(0.155) | 0.847*(0.351) |
| Years Not Enrolled in Education or Training | 0.419**(0.030) | 0.436**(0.030) |
| Sole Maori | -0.119(0.154) | -0.041(0.152) |
| Mixed Maori | -0.360*(0.145) | -0.283*(0.144) |
| Female | -0.040(0.068) | -0.111(0.069) |
| Male · Number Children Born to Respondent | 0.026(0.163) | 0.121(0.160) |
| Female · Number Children Born to Respondent | -1.074**(0.129) | -1.049**(0.127) |
| School Certificate | 0.910**(0.100) | 0.792**(0.102) |
| 6th Form or Higher School Certificate | 0.209*(0.096) | 0.176*(0.095) |
| University Bursary | -0.214*(0.096) | -0.211*(0.097) |
| University Diploma or Degree | 0.022(0.169) | -0.010(0.166) |
| Vocational Qualification | -0.062(0.069) | -0.081(0.068) |
| Mother had School Qualification | --- | -0.155*(0.076) |
| Mother had Post-School Qualification | --- | -0.110(0.096) |
| Mother had University Degree | --- | -0.606**(0.160) |
| Father had School Qualification | --- | 0.022(0.076) |
| Father had Post-School Qualification | --- | -0.061(0.111) |
|
Father had University Degree |
--- | -0.281*(0.127) |
| Years in Single-Adult Family | --- | 0.070(0.226) |
| Maximum Number of Children in Family | --- | 0.004(0.027) |
| Years Family Received Benefit | --- | -0.472*(0.213) |
| Real Family Income (in $10,000 units) | --- | 0.081**(0.030) |
| Mean Conduct Problems Score | --- | -0.007(0.005) |
| Years Truant, Suspended or Expelled | --- | -1.026*(0.449) |
| Convicted of Criminal Offence | --- | -0.058(0.112) |
| Alcohol/Drug Abuse or Dependence | --- | -0.139*(0.071) |
| R2 | 0.300 | 0.347 |
| Adjusted R2 | 0.292 | 0.330 |
| Number of Observations | 973 | |
| Mean of Dependent Variable | 2.249 |
** Significantly different from zero at 1% level.
* Significantly different from zero at 10% level.
Notes: This information is taken from the 973 respondents in the CHDS who provided valid information for the purposes of this study. The dependent variable is the effective years of work experience accumulated by the individual between the ages of 16 and 21. Youth are defined as ‘Sole Maori’ in this table 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’. Standard errors are in parentheses.
We saw earlier in Table 5 that mixed Maori had accumulated, on average, 0.410 fewer years of work experience than non-Maori. This difference is statistically significant at better than a 1% level. The Neumark decomposition can be used to break this observed experience gap between mixed Maori and non-Maori into its explained and unexplained components from this short regression.
(12) Observed Gap = [(ΔPOTEXP)b + (ΔX′)c] + [Unexplained] = [Explained] + [Unexplained]
0.410 = [-0.159 + 0.220] + [0.349] = [0.061] + [0.349]
If non-Maori spent the same time as mixed Maori away from education and training, the ethnic gap in work experience would widen to more than one-half of a year (0.410 + 0.159 = 0.569). Yet, we know that mixed Maori have, on average, more children and fewer qualifications than non-Maori. If non-Maori had these same personal characteristics as mixed Maori, the ethnic gap in work experience would narrow by 0.220 years. The net effect of all of these regressors is to narrow the gap in work experience by only 0.061 years. This is almost identical to the explained component (0.060) of the gap in work experience between non-Maori and all Maori in the earlier short regression using the two-way split in ethnicity. Less than one-sixth (14.9%) of the observed gap in experience between non-Maori and mixed Maori can be explained by the covariates in this short regression (0.061/0.410). More than five-sixths (85.1%) of this gap cannot be accounted for with this specification (0.349/0.410).
If the sample is restricted to those without formal school or tertiary qualifications, we find that unqualified mixed Maori (n=13) had accumulated 1.104 fewer years of work experience than unqualified non-Maori (n=132). This gap in experience between these ethnic subgroups is substantially larger than the gap for these overall ethnic groups of 0.410. Where the results from this short regression are capable of explaining 14.9% of the gap in work experience among all youth in these ethnic groups, the same regression can explain 29.4% of the average work experience gap between unqualified youth in these ethnic groups.
Column 2 of Table 13 reports the parameter estimates from the long regressions using the three-way split in ethnicity. As with the short regression, although the estimated coefficients are negative for both sole and mixed Maori, only the coefficient estimate for mixed Maori is statistically different from zero at conventional test levels (5.0% level in this case). Once these personal and family background characteristics have been held constant, there is no statistical difference between the accumulation of work experience by age 21 between sole Maori and non-Maori. As a result, we focus in the remainder of this section on the observed gap in work experience between mixed Maori and non-Maori.
The Neumark decomposition can be used to break this observed experience gap between mixed Maori and non-Maori into its explained and unexplained components from this long regression.
(13) Observed Gap = [(ΔPOTEXP)b + (ΔX′)c] + [Unexplained] = [Explained] + [Unexplained]
0.410 = [-0.166 + 0.311] + [0.265] = [0.145] + [0.265]
If non-Maori spent the same time as mixed Maori away from education and training, the ethnic gap in work experience would widen to more than one-half of a year (0.410 + 0.166 = 0.576). Yet, mixed Maori have, on average, personal and family background that are negatively associated with the accumulation of work experience. If non-Maori had these same personal characteristics as mixed Maori, the ethnic gap in work experience would narrow by 0.311 years. The net effect of all of these regressors is to narrow the gap in work experience by 0.145 years. This is similar to the explained component (0.140) of the gap in work experience between non-Maori and all Maori in the earlier long regression using the three-way split in ethnicity. More than one-third (35.4%) of the observed gap in experience between non-Maori and mixed Maori can be explained by the covariates in this long regression (0.145/0.410). The remaining two-thirds (64.6%) of this gap cannot be accounted for with this specification (0.265/0.410).
If the sample is restricted to those without formal school or tertiary qualifications, we find that unqualified mixed Maori (n=13) had accumulated 1.104 fewer years of work experience than unqualified non-Maori (n=132). This gap in experience between these ethnic subgroups is substantially larger than the gap for these same ethnic groups of 0.410, when those with formal school or tertiary qualifications are included. Where the results from this long regression are capable of explaining 35.4% of the gap in work experience among all youth in these ethnic groups, the same regression can explain 47.8% of the average work experience gap between unqualified youth in these ethnic groups.
Notes
- [22]Recall from Table 5 that the observed mean difference in work experience between sole Maori (2.120) and non-Maori (2.278) was also not statistically significant at a 10% level.
