The statistics we have examined so far have gone down into some fine details, but a correlation matrix is also useful for giving us information about high-level categories, such as men and women. What can the elementary gender division tell us about Kiwis?
Some points that stand out are ones that were already fairly well known. Men are slightly wealthier than women – the correlation between being a man and net personal income was 0.23. Also, to continue the general theme of minor social advantage, the correlation between being a man and voting in the 2014 General Election was 0.29.
Perhaps less well known is that men really like the National Party. The correlation between voting National in 2014 and being male was 0.35, which was significant. This was mirrored on the centre-left: the correlation between voting Labour in 2014 and being female was 0.31.
Neither of those statistics is surprising if the reader is aware of the many parallels between masculinity and conservatism, in particular the desire for the maintenance of a relatively high degree of order.
Likewise, there are clear parallels between femininity and social democracy, in particular the desire for a relatively egalitarian distribution of wealth and social status.
There are small, not significant correlations between voting Green and being male (0.10) and between voting New Zealand First and being female (0.21).
Of some interest, women smoke slightly more than men – being female has a correlation of 0.19 with being a regular smoker, although this is not significant. Possibly this reflects the value of nicotine as a treatment for certain anxiety and depression-related mental disorders, which women tend to suffer from at a greater rate than men.
Looking at gender differences in personal income and choice of employment, several interesting patterns reveal themselves.
One is that women are significantly more likely to be on any of the four benefits this study looks at. Although the correlation between being female and being on the pension was not significant (0.03), the others were much greater. Between being female and being on the student allowance the correlation was 0.21, with being on the unemployment benefit it was 0.39, and with being on the invalid’s benefit it was 0.26.
Being male was not significantly correlated with net personal income – the strength of this was 0.23, which was on the boundary of significance. However, looking at the next level down reveals a few patterns.
The personal income band most strongly correlated with being male was the $50-60K band. Here there was a correlation of 0.22. The female equivalent was the $5-10K band. There was a correlation of 0.21 between being female and being in this band.
Despite that males are generally slightly wealthier than females, this is not reflected in either of the $100K+ income bands. In both of these bands there is no correlation with gender.
This suggests a complicated pattern, but the general trend is that the higher the social status of any given line of work, the closer to gender parity the pay will be. This could reflect a lot of things.
Perhaps the most notable clue to answering this question comes from the fact that more men are managers – the correlation between being male and working as a manager was 0.49 – but more women are professionals.
This is an interesting division because it suggests that there is a difference in how men and women get to the highly compensated jobs.
Men are more likely to rise up to the top jobs from a lower starting point, a path not as easy for women because of the demands of childrearing. However, women are more likely to get a good education, valuable skills and therefore a high starting point, from where further advancement is not necessary or desired, or as heavily impacted by taking time off for children (many family GPs are women who fall into this category).
This might explain why there is no gender gap for the top income brackets, but explaining why there is a gender gap for the lower income brackets is a different matter.
Most of the reason is that men, whether by will or fortune, tend to choose industries that pay better than the ones women choose. Being male is significantly correlated with working in agriculture, construction, accommodation, and rental, hiring and real estate services, and these jobs tend to pay better than jobs in education and training and healthcare and social assistance, which correlate significantly with being female.
Perhaps the statistic that all gender warriors will find the least objectionable is that the people in the truly plum industries of professional, scientific and technical services, information media and telecommunications and financial and insurance services have the weakest correlation with being either male or female.
Generally there were no strong correlations between men and women in New Zealand, which one might expect from a generally free and liberal post-industrial secular democracy. The strongest correlation of all in this study was the completely unsurprising one between being female and being a single parent, which was 0.52.
This article is an excerpt from Understanding New Zealand, by Dan McGlashan, published by VJM Publishing in the winter of 2017.