Infometrics
Infometrics
PUBLIC ACCESS:
Min-Wage-Cash-Image
Has the minimum wage helped to lower wage inequality?

Disclaimer:

In relation to HLFS and Income Survey data:

Access to the data used in this study was provided by Stats NZ under conditions designed to give effect to the security and confidentiality provisions of the Statistics Act 1975. The results presented in this study are the work of the author, not Stats NZ or individual data suppliers.

In relation to the versions of the data accessed through the IDI:

These results are not official statistics. They have been created for research purposes from the Integrated Data Infrastructure (IDI) and Longitudinal Business Database (LBD) which are carefully managed by Stats NZ. For more information about the IDI and LBD please visit https://www.stats.govt.nz/integrated-data/.


In 1894, New Zealand introduced the world’s first minimum wage. Today, the minimum wage is still well and alive, growing faster than median wages since 2000.1 Over the same 20+ year period, there has been a substantial compression of the lower half of the wage distribution, leaving those at the bottom closer than ever to those at the top.

But how much of this increase in lower wages can we attribute to the rise in minimum wage over time? Many other factors could be making low-wage workers hours’ more valuable; for example, New Zealand’s labour force is notably more educated than 20 years ago, and hence likely more skilled. The Support Workers (Pay Equity) Settlements Act 2017 boosted the wages of 1% of the employee workforce. Further, teenage employees, who in 2008 made up 7.3% of employees, had their wages increase at least 33% by the 2008 abolishment of the youth wage. In this article, I explore how the minimum wage and wage inequality have evolved since 2000, and whether the minimum wage can be credited for improving wage inequality – an appraisal that could prompt further increases to it in the future.

A rising minimum wage

Minimum wage increases have been particularly pronounced in real terms since 2011, increasing 69.7% since 2000 (2.4%pa on average), 34.2% since 2011 (2.7%pa), and 19.1% since 2016 (3.0%pa). For comparison, the real median wage has increased 30.0% since 2000 (1.1%pa), 15.3% since 2011 (1.3%pa), and 6.9% since 2016 (1.1% - see Chart 1).

A good measure to understand the relative level of the minimum wage is taking its ratio to the level of the median wage, known as the Kaitz Index. This Index tells us how ‘close’ the minimum wage is to the median wage, and hence how relatively high the minimum wage is to the wage earnt by the middle earner across all wages.

In 2020, our calculation of New Zealand’s Kaitz Index was amongst the highest in the world at 0.72, compared to Australia at 0.53, Germany at 0.51, Canada at 0.49, and the world averaging at 0.49. The only country not below our Kaitz Index was Chile, at a matching 0.72. Therefore, if New Zealanders closely follow minimum wage laws, those who are working at the lowest wages in our economy are a lot closer to those on the middle wage, relative to the rest of world.

New Zealand’s minimum wage is also peculiar in that we abolished our youth minimum wage rate in 2008, whilst most countries (such as Australia and the UK) have staggered their minimum wage rates depending on age. The abolishment meant that all workers aged 16 and above in New Zealand could earn the adult minimum wage rate.2 Teenagers were no longer the ‘cheapest’ workers and had to compete for jobs against older applicants, who may have had more skills or experience.

In fact, the group most impacted by minimum wage changes by magnitude are teenagers. Out of all employees in 2021, teenagers comprised 5.0% of the workforce. However, out of the employees earning at or below the minimum wage, teenagers make up a quarter of all workers (24.9%). This much larger concentration means that, compared to all workers, minimum wage workers are almost five times more likely to be a teenager. Minimum wage workers are also 14% more likely to be female, 28% more likely to be of ‘Other’ ethnicity, and 33% more likely to be Māori or Pasifika, as compared to the entire working population.

The rise from below

New Zealand’s peculiar minimum wage context and large increases in the minimum wage rate spring the question of whether the rising minimum wage (and Kaitz Index) has impacted on wage inequality, specifically for the lower deciles. It is relevant to note that raising the minimum wage does not have predictable effects, with a range of impacts possible, including:

  • Making each hour of work more valuable
  • The number of hours worked, and offered each week
  • The propensity to be hired
  • The increased supply of labour (because working more hours becomes more attractive)
  • The decreased supply of labour (because the same earnings can be had with less hours)
  • The demand for labour (as costs to businesses increase)
  • The likelihood of continuing education (because working has a lower opportunity cost)

and many other factors.

Here, we will look at whether the increase in New Zealand’s minimum wage has impacted wage inequality, as the wage distribution gives us an indication of the market value of labour and hence of the activity of labour markets.

Since 2011 New Zealand has had a marked compression of the bottom half of the wage distribution, as well as declines in various measures of bottom half wage inequality. As Chart 2 shows, the wage distribution has been successively shifting right. There is also increasingly a ‘bunching’ of wages approximately around the level of the minimum wage. Looking at 1998-2000, the distribution of wages was almost a bell curve, indicating the minimum wage likely did not have a strong effect on determining wages. However, as the years go by, this left slope slowly disappears as the minimum wage seems to become a binding constraint on lower wages, leading to a wall of people whose wages appear to have been lifted by the minimum wage.

Additionally, as seen in Chart 3, wages in percentiles 5 and 10 (the lowest 5% and 10% of wages, respectively) have seen impressive growth, increasing by 67% and 46% respectively since 2000, well above all other percentiles.

In fact, in 2021, the boundary of the 10th percentile was equal to $20 – the level of the minimum wage – meaning any changes to the minimum wage directly affected at least 10% of all employees, all else equal.

We can also look at how the different percentiles have fared compared to each other using ratios of percentiles. For example, a 90-10 measure is the ratio of the 90th percentile boundary to the 10th percentile boundary (in other words, the highest 10% of wages versus the lowest 10% of wages), indicating how close the boundaries of the top 10% and bottom 10% are to each other. As shown in Chart 4, the biggest change since 2000 has been in the 50-10 measure, where the distance between the 50th and 10th percentile of the wage distribution has decreased by 26%. This decrease begins steadily around 2005, and really takes off from 2017 as minimum wage increases got larger in dollar term.

Note that the 90-10 measure has also decreased, but it’s a smaller decrease than the decrease in the 50-10 measure, indicating that the 90th percentile of wages has grown faster than the 50th percentile since 2000, such that we don’t see any clear reduction in wage inequality for the upper half.

However, the groupings shown in Chart 4 only show us what is happening at each of the grouped percentile boundaries, and may not give the whole picture. For example, if within a certain decile, the distribution of wages has changed, the change may not show up in the decile boundary, but is relevant to our question, as ‘within decile’ inequality has changed.

In Chart 5, I examine how every percentile in the wage distribution has had its boundary changed between 2000 and 2021, and between 2017 and 2021. In both time periods, all percentile boundaries have grown, but the magnitude of growth increases as one goes down the percentiles. For example, the orange line shows that percentile 10 (the wage for the lowest 10% of pay rates) has grown by about 40% since 2000, whilst the 90th percentile (the highest 10% of pay rates) has grown about 25% since 2000. It is striking how much the lowest 10 percentiles have grown in both time comparisons, indicating that much of the reduction in wage inequality may be from pulling in the lower tail of the wage distribution into the middle.

Indeed, a particularly interesting aspect of wage inequality has been the continuous and somewhat sudden contraction of the standard deviation of real log wages, particularly since 2015. This continued and immediate contraction indicates that the wage distribution has gotten much tighter, meaning employees’ wages have come increasingly closer together. This contraction would especially be coming from the tails of the distribution, as the standard deviation is most sensitive to the extremes.

More to wages than a minimum wage

So far from my analysis, wage inequality looks to be explained by changes in the minimum wage affecting the lower deciles. However, wage inequality at the bottom half is not only reliant on the level of the minimum wage. For one, the bottom decile (which contains all minimum wage workers) has become increasingly educated since 2000, and hence likely more skilled - that is, producing work that is more valuable per hour.

If we look at what percentage of the workforce has no qualifications, this figure has decreased 18 percentage points since 1997. Meanwhile, the percentage of those with no qualifications who are earning the lowest 10% of wages (decile 1) has dropped 23 percentage points, suggesting that workers in wage decile 1 have had a higher change in education composition than the overall population (see Chart 7).

Indeed, looking at the data, more decile 1 employees had high school qualifications, post school qualifications, and degrees in 2021 than in 1997. Therefore, we can infer that decile 1 employees – and hence minimum wage employees – are more skilled than they were 20 years ago, which may mean that their output per hour is higher or more valuable, hence raising their price of labour.

Health workers leaving the minimum

Another factor which may have compressed lower tail wage inequality is the Support Workers (Pay Equity) Settlements Act 2017. This agreement led to successive wage increases for health and care workers, as shown in Table 1. The result was that this group of workers, who in 2015 comprised 4.5% of all minimum wage workers above the age of 25, soon had no minimum wage workers by 2017. Indeed, the average real wage for care workers went from $18.70 in 2015 to $24.50 in 2021, a 31.0% increase.

Where does this leave us?

We have seen that wages have been increasing in real terms since 1997, and especially for the lowest paid workers, who are with time getting even closer to the middle wage. This greater increase in lower wages might be due to New Zealand’s minimum wage policies, which are also amongst the most generous in the world – but the results are not definitive.

The minimum wage seems to have now cut into the wage distribution, creating a spike of people who earn the minimum wage that wouldn’t have been otherwise. An interesting point that this article did not discuss is also whether any workers have had spill over effects, i.e., workers who earn just above the new minimum wage getting a rise in pay, which would also help explain how the minimum wage could be improving wage inequality. But the wage distribution may also have been compressed because of increasing qualification levels for lower wage workers or the Support Workers (Pay Equity) Settlements Act 2017 that affected 4.5% of minimum wage workers and boosted their hourly wage substantially.

The overarching conclusion is that although the improvement in wage inequality has been significant, it is not clear that the minimum wage can predictably and conclusively cause this trend. In other words, we cannot yet claim that the minimum wage is a hero for wage inequality.


1 All calculations of wages are based on a sample of employees aged 16-64 with positive hours and positive earnings. All samples have been person-weighted according to Stats NZ sampling weights.

2 After 6 month’s work, if they are under 18 and working for the first time.

Related Articles