Household spending › Communication
Postal Services: what NZ households spend per week
Stats NZ publishes this category as Postal services, within Communication. It is called Postal Services here because that is what it is, but every figure on this page is the official category.
What this category covers
Sending things rather than communicating electronically. Stamps and franked postage, courier and parcel services bought directly by the household, freight on an item being sent, registered and tracked post, post office box and private bag rental, and packaging bought for the purpose. Both a single stamp and a business-sized courier account for a home-run venture are recorded here.
What it does not cover
Delivery charges bundled into the price of an online order are part of the item bought, not postal services, which is why this line understates what households actually pay to have things delivered. Telephone and internet services are telecommunication services.
Those boundaries are set by the survey's own classification, not by us. They matter because a figure read against the wrong definition is worse than no figure: it looks precise and it answers a different question. If you are comparing your own spending against this page, check the exclusions above first.
How to read the figure
A small and shrinking category recorded by under one household in twelve, with a sampling error wide enough that the figure is suppressed here. Its interest is entirely in the trend: it is one of the few lines in the survey that shows a long structural decline, and the reason is visible in the telecommunication services line next to it.
Every survey year
The Household Economic Survey runs every three years. This is postal services in each survey since 2007, for all New Zealand households.
Read the two money columns together. The published column is the money of each survey year itself, so comparing 2007 with 2023 across it compares two different dollars. The adjusted column restates every year in 2026.06 dollars using the Communication CPI series, which is the column to use for a real change. That adjustment is our arithmetic, not a Stats NZ statistic.
The reporting column is the share of surveyed households that recorded any spending in this category at all: the lower it is, the further the average sits from what a household that does spend here actually pays.
| Survey | Published per week | In 2026.06 dollars | Sampling error | Households reporting | Change |
|---|---|---|---|---|---|
| 2007 | $1.40 | $0.97 | ±20% | 19.3% | — |
| 2010 | $0.00 | $0.00 | — | 0% | -100% |
| 2013 | $1.50 | $1.28 | ±20% | 16.7% | — |
| 2016 | $1.20 | $1.13 | ±22% | 11.7% | -20% |
| 2019 | $1.20 | $1.28 | ±41% | 6.5% | 0% |
| 2023 | $1.60 | $1.82 | ±31.5% | 7.6% | +33% |
Postal Services over time
Each bar is one survey, restated in 2026.06 dollars so the six are comparable. In real terms, spending on postal services has moved up 88 per cent since 2007. In the money of the day it went from $1.40 to $1.60 a week, and most of that difference is prices rather than behaviour, which is exactly why the chart is drawn on the adjusted figures.
$0$2.49
Nothing to chart for 2010, because the survey recorded no spending in this category there at all. That is a measured zero rather than a figure withheld for uncertainty.
The dark line with a cap at each end shows the sampling error Stats NZ publishes with that figure: the survey is confident the true value lies within it. A long line means a less certain number. Bars in a pale fill are survey years whose sampling error is above the level at which this site normally withholds a figure. They are drawn here because the direction of a series is worth seeing even where the level is not reliable enough to quote, and the table above gives the error on each one.
By tenure
Whether a household owns or rents correlates with age, income and household composition, so a difference here is not caused by tenure alone.
$0$1.97
Not shown, because the survey sample was too small to report reliably: Owned, mortgaged, Owned, no mortgage, Not owned, Renting, pays rent, Family trust, Trust, mortgaged, Trust, no mortgage.
Nothing to chart for Renting, no rent, because the survey recorded no spending in this category there at all. That is a measured zero rather than a figure withheld for uncertainty.
The dark line with a cap at each end shows the sampling error Stats NZ publishes with that figure: the survey is confident the true value lies within it. A long line means a less certain number.
Breakdowns that cannot be reported
Too few surveyed households recorded spending in this category for these breakdowns to be published at a sampling error under 30 per cent: by household size, by region, by household income. They are named here rather than left out, because a missing section reads as an oversight and a suppressed one reads as what it is. The communication page reports the same breakdowns for the wider category, where the sample is large enough.
The rest of communication
These are the other components of the same top-level category. They add up, with this one, to the total on the Communication page.
Related calculators
Where these numbers come from
Every figure on this page is from the Stats NZ Household Economic Survey, retrieved through the Aotearoa Data Explorer, for survey years 2007 to 2023. Inflation adjustments use the Stats NZ Consumers Price Index to the 2026.06 quarter. Stats NZ data is used with attribution; the survey figures are theirs and the adjustments are ours.
Household Economic Survey figures are averages of weekly household expenditure and carry the sampling errors shown. They are averages rather than medians, so they are pulled upward by the highest spenders. Figures are rounded as published.