Occupation mix accounts for about a quarter of Australia's 15.9% Indigenous pay gap

You can check the professional figures against each other. Divide each group's average annual pay by its average weekly hours and Indigenous professionals come out at about $2,452 per weekly hour, against $3,046 for all professionals. That is a ratio of roughly 0.80 ^[1]. It matches the researchers' own hourly figure of about 80 cents in the dollar for this group, which includes tech sales professionals and pharmacists ^[7]. The annual shortfall is $23,064, or about 21.6%. The hours differ by 0.9 a week, about 2.6% ^[2].

The analysis covers 227 occupations ^[4], and the widest gaps are in higher-skilled work, among professionals and managers ^[7]. That tells an employer where to look first. The researchers' decomposition also puts a limit on what changing the job mix can do. Concentration in lower-paid jobs accounts for 27% of the average hourly gap, and hours for 2% ^[12]. Applied to the 15.9% headline, occupation is worth roughly 4 percentage points and hours about a third of a point, which leaves about 11 points unexplained ^[3]. I think the occupation map is most useful for deciding where to audit pay and promotion. A substantial gap remains even after hours, occupation, age and education are accounted for ^[11].

That unexplained 70% is simply the part of the gap those measured factors cannot account for. The researchers write that it likely reflects structural factors, and they name three: racial bias in human resources practices, unpaid cultural work and inflexible workplace practices ^[13]. A decomposition like this measures how big the unexplained share is. It cannot test which cause fills it. The thing this doesn't tell you is how the 70% splits among those three causes, or whether variables outside the model contribute to it.

The comparison groups affect the headline number. The 15.9% is measured against all workers, a group that includes Indigenous workers, but the researchers also express it as 84 cents for every dollar earned by non-Indigenous people ^[1]. Indigenous pay pulls the all-worker average down, so the gap against non-Indigenous workers alone would be at least as wide ^[6]. The 21%, 18.4% and 12.5% figures work the same way, since each is measured against all men or all women ^[5][6].

The researchers make five recommendations ^[19]. Paying for cultural work speaks most directly to one of the causes they name. They write that Indigenous employees' mentoring of colleagues, advising of leaders, support for reconciliation and work building community relationships often sits outside job descriptions and goes unpaid. They want employers to count it in workloads, roles and pay ^[18].