SAYit Blog
Are incomes keeping up with inflation?

We've done a lot of work recently on the issue of inequality.  As part of that, in August 2013 a SAYit survey asked a representative sample of n=1000 New Zealanders whether they believed that household incomes had risen by more than prices over the last year, or alternatively if prices have grown faster than household incomes.  We asked people to comment based on what they had experienced themselves and anything they have seen or heard, and the results are emphatic.

  • 91% of New Zealanders say that prices have risen faster than household incomes over the last year, with 5% saying that household incomes have outstripped price rises and 4% being unsure.

The view that prices have risen faster than incomes is shared by every demographic group we break our results down by, although there are some differences:

  • 96% of women think that prices have risen faster than incomes, compared with 85% of men.
  • 93% of 30-44 year olds and 92% of 45-59 year olds think that prices have risen faster than incomes, compared with 88% of under 30s and 88% of over 60s.
  • 86% of Aucklanders and 89% of Wellington residents think that prices have risen faster than incomes, compared with 94% of people living in other parts of the North Island.
  • 90% of Canterbury residents take this view, compared with 95% of people living in other parts of the South Island.
  • 95% of Labour voters take this view, compared with 85% of National voters.
  • 93% of people with household incomes below $100,000 think that prices have risen faster than incomes, compared with 85% of those with household incomes over $100,000.

This data suggests that, although people on medium to low incomes and those living in provincial areas are feeling the pinch more, rising prices are being felt by New Zealanders across the board.

I thought I'd look into that using the actual statistics.  I've used Statistics NZ's average hourly earnings data, plus the Consumer Price Index from Q1 1989 (which is as far as the hourly earnings data goes) to Q2 2013 (the most recent data available).

  • Between Q2 2012 and Q2 2013, average hourly earnings went up by 2.0%, while prices went up by 0.7%.  In other words, over that specific year, earnings have in fact risen faster than wages, at least on the official measures.
  • Similarly, since 1989, average hourly earnings have increased by 111% (from $13.07 to $27.55), while prices have gone up by 79%.
  • Even over the last 5 years, average hourly earnings have gone up by 14% while inflation has gone up by 11%.

I'm not 100% sure that those are the right statistics to be using but it does suggest that perceptions are not necessarily in line with reality.

Another issue may be that the 'bundle of goods' people use to develop their own perceptions of price movements is different from the 'bundle of goods' actually used by Statistics New Zealand.  According to Statistics NZ:

The CPI consists of a basket of goods and services that represent purchases made by households. The goods and services in the basket, and their relative importance, are reviewed every three years to ensure the basket remains up to date. There are about 690 goods and services included in the basket. They are classified into 11 groups:

  • food
  • alcoholic beverages and tobacco
  • clothing and footwear
  • housing and household utilities
  • household contents and services
  • health
  • transport
  • communication
  • recreation and culture
  • education
  • miscellaneous goods and services.

Perhaps​ what's going on is that people placing more emphasis on things like food, housing and household utilities (e.g. power bills) than Statistics NZ do, and less on clothing and technology.  While housing affordability is often seen as an issue that is rapidly getting out of hand, we don't seem to put as much emphasis on the fact that technology in particular is often massively cheaper in real terms than it used to be.  For example, I remember paying around $2000 for my first laptop in the mid 90s and used to regularly pay $30 for videos, whereas now I could get perfectly decent laptop for under $1000 and many decent movies are available on DVD for $10.

On the other hand though, perhaps the public perception is perfectly reasonable.  After all, we need food and housing, and we need to pay our household bills, but we can often make do without buying a new laptop or DVD.  Could it be that incomes are rising faster than prices for the things we might quite like to have, but failing to keep up with the rising prices for the things we actually need?

EDIT: Gerard Herir makes a very good point below - it definitely would make a difference if it was the median as opposed to the mean.  There's certainly evidence that incomes have increased more quickly at the top end of the scale than at the bottom.  The specific statistics I've used, however, only specify that it is the average hourly earnings, with no information on that page showing whether it's the mean or the median - the term 'average' can apply to either.  Even though it's often assumed that the term 'average' implies the mean rather than the median, I'd have expected Stats NZ to realise the problems inherent in using means for this sort of data and would instead use the median.  On closer inspection, however, it turns out that, when they use the word 'average' for these stats it's actually the mean.  Definitions are available here:  

For those who don't know stats or need a reminder: 

  • to work out the mean you add up all the incomes in New Zealand and divide by the number of people
  • to work out the median, you rank all the incomes in New Zealand from top to bottom and work out which one is in the middle.

If your data has a lot of extreme values, then it's generally better to use the median rather than the mean.

Stats NZ's otherwise excellent Infoshare tool only allows breakdowns by 'average hourly earnings' and not by 'median hourly earnings'.  If incomes at the top end have in fact increased by greater percentages than the incomes of people in the middle (and I think most of the evidence does point that direction), then Gerard is absolutely correct. Unfortunately, however, those are the only stats that Stats NZ is providing in this format - which I think is really bad.