Matrix > Toolkit: Pay and Conditions > GEM Pay Gap Calculation Recommendations
GEM Pay Gap Calculation Recommendations
GEM recommends calculating pay gaps using methods that reveal intersectional inequities, and which do not understate the size of gaps in real life. However, organisations calculate pay gaps in a range of different ways. Whichever method is chosen is better than not calculating or reporting Māori and Pasifika pay gaps at all. Click here for Doing Data Right, the Toolkit’s landing page for all data advice.
Contents
The GEM approach to pay gaps
Organisations calculate pay gaps in a range of different ways. Any of these methods are better than not calculating or reporting Māori and Pasifika pay gaps at all. However, where possible, GEM recommends:
Analyse pay gaps or pay comparisons by a combination of ethnicity and gender.
All large businesses in the private sector who had been preparing or considering Equal Pay Act gender pay gap reporting as per government proposals in 2023 should report their gender pay gaps and horizontal/equal pay data using combined ethnic and gender pay gaps as the gold standard.
Use European men (the most highly paid group nationally) as the comparator group for pay gaps, or if there are not enough of them, just ‘European’. This is preferable to using ‘non-Māori’ for the Māori pay gap and ‘non-Pasifika’ for the Pasifika pay gap.
Avoid using overall workplace average pay as the comparator.
In smaller workplaces with not enough people in each ethnic category, or where there are not enough Māori or Pasifika staff to protect individuals' privacy, using overall ethnic group (all genders) instead of ethnicity by gender, is okay (We take this approach in our worked examples with a small imaginary sample).
Everyone with any Māori ethnicity is counted as Māori in calculating the Māori pay gap, and everyone with Pasifika ethnicity is counted as Pasifika when calculating the Pasifika pay gap, regardless of how many ethnicities an individual identifies as.
The European comparator group should ideally be ’single ethnic group’ European only. This is particularly important for calculating the European-Māori pay gaps in businesses with very large numbers of Māori employees.
Currently “Pākehā” is not a level 1 or 2 ethnic category used by Stats NZ, although 'write-ins' in the Census are counted and back-coded to “NZ European” at level 2. The level 1 category is just “European”.
Combined ethnic and gender pay gaps
Calculating combined gender and ethnic pay gaps is important to ensure an organisation has an intersectional understanding from the outset of its journey toward closing those gaps. Measuring pay gaps can be done in two ways, as outlined below – we recommend the first approach. However, the analysis for both methods is essentially the same – calculating the median or average pay for different groups and comparing them.
1. Calculating pay gaps for Māori and Pasifika compared with European men. This is useful for:
Establishing specific pay gaps and what increases are needed to close them, e.g. “in this organisation, Pasifika woman earn 25% less than European men. Over time how do we close this gap?”
Tracking progress toward targets over time.
2. Comparing median or average pay for each group in a bar chart, by ethnic group and gender.
This is useful for communicating the overall picture in a visually digestible way, and is the approach taken by the Public Service Commission in reporting on combined gender and ethnic pay gaps.
Click here to read the example from Kai Toipoto - Te Tohu Guidance (see page 10).
We recommend that the bar chart comparison method is only used in combination with calculating the specific pay gaps for each group compared with European men as a complementary visual tool.
The bar chart approach is useful for capturing the breakdown of ethnic and gender pay gaps at a moment in time. However, tracking progress over time – crucial for meeting targets – is easier if the individual pay gap percentage figures for each group is tracked.
Recommended pay gap calculation formulae
GEM recommends using median pay for the best analysis of pay gaps, as per the MindTheGap approach to the gender pay gap, but many organisations following the Kia Toipoto methodology may also be using average pay to start with. Medians better capture where the true ‘middle’ of the wage distribution is for each group compared with averages (means). However, the most important thing is to measure Māori and Pasifika pay gaps at all. If an organisation is consistently measuring and reporting combined gender and ethnic pay gaps using averages, this is still ‘Good’ according to the GEM. In our worked examples at The Example Factory, we use averages in our small imaginary sample. However, to become ‘Great’, GEM recommends analysing and reporting medians as well as (or instead of) averages.
Using European men as the comparator group
The Human Rights Commission Pacific Pay Gap Inquiry and the MindTheGap campaign have used European men as the comparator group for calculating combined gender and ethnic pay gaps. Kia Toipoto has a slightly different approach, using a binary split between Māori and non-Māori, or Pasifika and non-Pasifika, and using averages (means) instead of medians. They are all valid methods with different pros and cons.
Using 'non-Māori' or 'non-Pasifika' comparator group instead of 'European' can underestimate the scale of ethnic inequality
Using the data from The Example Factory, we can see how the use of a single-ethnic group ‘European’ comparator group gives quite different results to ‘non-Māori’ or ‘non-Pasifika’. Although this is a fictional example with a small sample size, it is common to have workplaces with high numbers of Māori and Pasifika staff, but the management is largely Pākehā. Using ‘European’ clearly identifies a common comparator for minority groups affected by pay gaps.
We first met The Example Factory here. Here’s how The Example Factory’s spreadsheet (backcoded to Level 1 ethnic group) would look with their pay added:
Here is how The Example Factory’s Māori and Pasifika pay gaps would differ using different comparator groups. (This example uses averages due to the small imaginary sample, although GEM recommends medians as the ideal approach).
Why 'single-ethnicity' European as the comparator group?
GEM specifies ‘single ethnicity European’ to make it clear that the comparator group should be Pākehā with no other ethnic group (including if reporting also on European women’s pay). For example, for those acknowledging both Māori and European whakapapa, counting their pay twice as both ‘Māori’ and ‘European’ pay using the Total Ethnicity approach will technically drag down the European figure, potentially understating the reality of the Māori pay gap in a single business. This would particularly affect Māori pay gap calculations in businesses with a high proportion of Māori staff, as 43% of working age Māori also reported Pākehā whakapapa in the 2018 Census. The same issue technically affects Pasifika and other ethnic pay gaps, but to a far lesser degree.
The Example Factory has a high proportion of Māori and Pasifika staff, who also have a high level of multiple ethnic groups, including Pākehā ethnic group, but there is only one single-ethnicity Pākehā, who is paid more than the rest of the staff. Below, we can see that using the single-ethnicity European comparator has very different results from using the Total Ethnicity approach, reducing the appearance of Māori and Pasifika pay gaps. This is because the lower-paid Māori and Pasifika staff have their pay double-counted as European pay as well.
How to prioritise ethnic group according to the pay gap being calculated
When categorising people with multiple ethnic groups for pay gap calculations, GEM recommends:
‘Total ethnicity’ for deciding how many pay gaps apply to each person, i.e. include people with multiple Māori or Pasifika ethnic groups as Māori when calculating the Māori pay gap, and as Pasifika when calculating the Pasifika pay gap.
‘Prioritised ethnicity’ within each ethnic pay gap calculation so that no-one is counted more than once in each pay gap formula, e.g. someone with both Māori and Pākehā ethnicity would be included as Māori in the Māori pay gap calculation, but not counted again as part of the non-Māori/Pākehā comparator group in that calculation.
Back at The Example Factory:
Maxwell is Māori, Pasifika and Pākehā – an increasingly common combination.
He gets counted as Māori in the Māori pay gap, and Pasifika in the Pasifika pay gap but is not counted as European for the comparator group.
Maxwell’s pay gets counted only once per pay gap calculation.
In the Māori pay gap calculation, his pay would not be counted as non-Māori pay or European man pay for the comparator group, despite also having Pākehā ethnic group. He just gets counted once – as Māori.
In the separate Pasifika pay gap calculation, again, he would not be counted in the European man pay or for the comparator group, despite having Pākehā ethnic group. He just gets counted once – as Pasifika.
See the glossary for short explainers on total and prioritised ethnic group.
Prioritisation for pay gap calculation example
The ‘prioritisation’ process here means allocating people with multiple ethnic groups to a single ethnic category for the purposes of the relevant pay gap calculation. For the Māori pay gap, ‘go down the list’ following the standard order of prioritised ethnicity used in the health sector (Māori, Pasifika, Asian, MELAA, Other, European), swapping the Māori and Pasifika order for the Pasifika pay gap.
At The Example Factory, to calculate the Māori pay gaps, Māori with multiple ethnic groups are assigned to ‘Māori’ only. Any Pasifika who are not Māori will be assigned to ‘Pasifika’ only. Those with European ethnic group only will be left in the European column.
To prioritise Pasifika to calculate Pasifika pay gaps would just mean swapping the position of Māori and Pasifika in the standard prioritised order for this one purpose. John remains the only single-ethnicity European.
