Subscribers to our Building Forecasts will have noted substantial upward revisions by Stats NZ to its estimates of “other construction” activity (ie, excluding residential and non-residential construction – so mostly infrastructure) over the last 2-3 years. This article examines why Stats NZ’s “other construction” estimates can be subject to so much revision, and takes an in-depth look at what we know about this segment of the economy.

Although there are regular monthly releases of residential and non-residential building consents, and a quarterly assessment of the amount of building work put in place for these two construction categories, there’s nothing similar to assess the levels of network infrastructure construction undertaken in New Zealand. Instead, GDP data provides a view on gross fixed capital formation (essentially, investment into plant, property, and equipment), including the “other construction” category, from which we derive our infrastructure estimates.

Chart 1 shows the revisions to Stats NZ’s “other construction” estimates between the GDP data releases published in September and December last year. The changes are massive, averaging a 14% upward revision in the March 2021 year and a 24% upward revision in the year to March 2022.

We’ve seen substantial revisions like this one before. In late 2020, published estimates of “other construction” for 2017/18 and 2018/19 showed investment levels that were higher than previously published data by 12% and 22% respectively.

The changes to the estimates come about because of the lack of comprehensive data for “other construction” on a quarterly basis. As noted earlier, this lack of coverage contrasts with data for residential and non-residential construction. “Other construction” is based on more limited data collected from a selection of companies that represent only a small portion of activity.

Annual updates to the national accounts, which underpin Stats NZ’s GDP figures, are published in November each year, with the potential for major revisions. These updates are based on far more comprehensive financial data from the Annual Enterprise Survey and, particularly pertinent for “other construction”, the Local Authority Census.

Our assessment of various datasets suggested that the revisions to “other construction” were largely due to higher levels of local government spending on water, sewerage, and drainage projects. Conversations with Stats NZ have highlighted the same trends. Central government spending on transport infrastructure projects was also revised upwards, with Stats NZ noting that smaller crown entities have the potential for sizable revisions.

Additional data from Stats NZ provides a small insight into what kind of “other construction” is being undertaken each year in this investment category. The gross fixed capital formation data is split up by production industry on an annual basis, and the shares for “other construction” activity are shown in Chart 2.

On average, electricity, gas, water, and waste services makes up the largest share of “other construction” activity (37% in 2021). Activity in this category can be relatively variable, influenced by major generation projects or significant upgrades by Transpower, for example. Water assets, and limited waste assets, will also be recorded in this group.

Transport, postal, and warehousing has the second-largest share of activity (36% in 2021), although it maintained the largest share between 2014 and 2020. The largest component of this category is likely to be roading, with rail making an increasing contribution in recent years, and airports and ports likely to be less important.

Information media and telecommunications represented 6.8% of activity in 2021, its lowest result since 2011. Activity throughout the last decade has been boosted by Ultra-Fast Broadband and associated government initiatives. Investment in mobile network upgrades will also be a key component of this segment.

Mining is also worth mentioning, if only to note that its share of “other construction” investment has fallen from 32% in 2008 to just 3.4% in 2021. The boom in oil and coal extraction in the mid-late 2000s has given way to a dearth of new activity throughout the last decade. Whereas previously referring to “other construction” as “infrastructure” risked overstating activity because of the potential significance of mining in the numbers, mining’s demise means that this potential issue is not currently a problem.

Our Infrastructure Pipeline Profile provides subscribers with a similar, but more in-depth, breakdown of planned infrastructure spending by asset type.

One of the consequences of the significant upward revision to “other construction” estimates is that activity now makes up its largest share of the economy since March 2010 (see Chart 3). When one considers the much larger share of activity represented by mining between 2007 and 2010, the implication is that New Zealand is currently investing in infrastructure at a faster rate than for many decades. Indeed, additional data stretching back to the early 1970s shows that, apart from 2009/10, “other construction” only previously made up a larger share of GDP in 1976 and 1978.

It is against this backdrop that we have taken a cautious, and arguably downbeat, view about prospects for infrastructure activity over the next couple of years. The capacity constraints and skill shortages that have affected the industry since the COVID-19 pandemic began are well documented. The confirmation of high activity levels is more consistent with the cost increases that have been recorded in the infrastructure industry during 2021 and 2022 than the previous estimates of activity were.

There is little doubt that more investment in New Zealand’s infrastructure is necessary and that there is much work to do. The government’s Three Waters programme, arguably inadequate funding for regional roading, as well as several significant projects that are planned across the roading, rail, and electricity generation networks are areas that spring readily to mind.1 However, the highly stretched nature of the construction industry means we are doubtful about whether further growth can be achieved.

Perhaps even more importantly, the cost overruns caused by rapid inflation over the last 18 months are leading to significant budgetary issues for local government and central government agencies. Looking forward, the higher costs to complete work effectively mean there will be less money to go around the planned range of projects. As a result, without big increases in revenue, we expect to see more prescriptive prioritisation of projects, with the less important initiatives being put on hold. In other words, the value of infrastructure spending is likely to run in line with projections, but the volume of work will be well below what might have been expected when budgets were initially prepared.

1 Some care needs to be taken when considering inadequate maintenance of wastewater pipes or roads, for example. Work to maintain or repair these assets is usually likely to be excluded from the figures discussed in this article, because gross fixed capital formation only captures spending on new assets and significant upgrades or expansions of existing assets. Complete replacement of a failing pipe would probably be captured in gross fixed capital formation, but repairs to that failing pipe would probably be excluded and classed as maintenance.