Can zoning reform increase construction productivity? Suggestive evidence from New Zealand.

Economists frequently emphasize the importance of productivity growth, as it’s the main driver of living standards and prosperity. A rise in productivity means we’re able to do more with less; we can have more and better products with fewer hours of work and resources.

Economists define productivity as the output produced for a given level of input. To make this concept a little less abstract, in this post we’re going to focus on labour productivity, which refers to how much a worker can produce within a given time frame (e.g. an hour, or a year).

The construction sector presents a global productivity puzzle. Around the world, while we have gotten better at making all kinds of stuff – cars, phones, pharmaceuticals, etc. – we haven’t gotten better at building houses. Indeed, a recent paper shows that the productivity of a construction worker in the U.S. has remained unchanged since the 1960s.

This productivity divergence mirrors a diverge in price: houses have become more expensive over time, while cars and other goods have gotten cheaper. The improved quality of houses don’t explain these trends, nor do the techniques we use to measure construction output.

Australia is no exception to this phenomenon: our construction sector is about as productive as it was at the turn of the century, and labour productivity has been declining over much of the past decade. For comparison the ‘market sector’ – an aggregation of most major non-government industries – is about 25% more productive than 20 years ago.  

This issue is particularly concerning for Australia, as the construction sector makes up a much larger share of GDP than other rich countries. In most OECD nations, construction contributes around 5% to economic output, but in Australia, it accounts for over 9%. As a result, productivity in the rest of the Australian economy needs to grow faster just to keep pace with global standards.

So why we have we gotten so much better at building goods which you can pick up, play with, or drive but not things fixed on a block of land?

An emerging literature has pointed the finger at zoning regulations. The aforementioned paper noted that a) construction productivity plummeted after modern land use regulations came into effect, and b) areas with stricter zoning regulations have smaller and less productive firms. Zoning may reduce the capacity for construction firms to grow, achieve economies of scale, and invest in new technology.

Lower productivity could be a significant factor in how zoning regulations drive up housing costs. When firms are inefficient, they may struggle to build enough houses each year to prevent prices from rising. Additionally, the homes that are built are more expensive because they require more capital and labour to construct. Higher costs can squeeze profit margins, discourage new firms from entering the market, and means many projects are unfeasible.

A case study that tests these theories is New Zealand’s experience following major changes to land use regulations in the 2010s. Auckland, home to around a third of the country’s population,  adopted widespread upzoning through the Auckland Unitary Plan (AUP) from 2013 to 2016. The AUP upzoned residential land to allow for medium density development throughout much of the city, and high-density around transit corridors.

Australia and New Zealand have the same definitions for categorising industries (Australian and New Zealand Standard Industrial Classification (ANZSIC) Codes) so we can make direct comparisons in construction productivity growth on both sides of the Tasman. On all three metrics of productivity – Labour, Capital, and Multifactor – the two countries track each other closely until 2013. This isn’t unexpected, given the global stagnation in construction productivity and the lack of any structural factors that would lead to differences in technology adoption or capital accumulation between the two nations.  

However, following the implementation of the AUP construction productivity in New Zealand began to rise, while Australia’s continued to stagnate. Indeed, in 2020, New Zealand’s construction multifactor productivity was around 15% higher than 2000, while Australia’s was 10% lower.

You can note that much of these gains were rapidly eroded over COVID-19 pandemic – a period I’ve shaded in grey. This reflects New Zealand-wide, rather than construction industry specific, factors. Lockdowns across New Zealand reduced productivity in goods producing sectors in 2021, followed by recessionary periods in 2022 and 2023, with multiple quarters of negative growth and high inflation. Indeed, labour productivity in all goods producing industries are 8% lower than pre-pandemic levels, according to the latest measurements. This serves as a reminder that poor macroeconomic conditions can undo a lot of the benefits from microeconomic policy reform.

Given this drop doesn’t reflect factors in the construction sector specifically, rather is an macroeconomic shock generally, I’m going to focus on the 2013-2021 period. You should caveat the below analysis appropriately and note that basically every line will see a sharp decline from 2022. Whether there will be a bounce back in future years as the economy recovers remains to be seen.

While the above charts suggest that something happened that made NZ’s construction sector more productive around the time of zoning reform, correlation is not causation. While I won’t make any causal claims, a closer look suggests zoning reform is a highly plausible explanation for the increase in construction sector productivity.

First, let’s rule out some alternative hypotheses. Perhaps NZ was experiencing a general productivity boom over the 2010s while Australia wasn’t. This doesn’t appear to be true: productivity in other goods producing industries was flat during the 2010s, while construction productivity spiked, and - in a sharp contrast to the global story - outperformed other sectors.

Construction is also a broad sector. Perhaps New Zealand got better at building infrastructure, rather than houses, and our sector-wide measure is picking that up instead. This doesn’t seem to be the case either. “Heavy and Civil” construction productivity was flat, while “Building Construction” and “Construction Services” grew.

Second, we can estimate labour productivity by region to see if growth was driven by upzoned areas. Stats NZ publishes estimates of industry GDP and employment by region. We can combine the two to obtain an estimate of construction productivity by region. (Note that this variable is slightly different to the published productivity measures I was using above).

We will focus on two regions in New Zealand. Auckland which, as we noted above, adopted major zoning reform from 2013. And Canterbury, which includes Christchurch.

The story in Christchurch is complicated, and not as clean as in Auckland. In 2011, Christchurch experienced a major earthquake which damaged around 167,000 buildings. In response to the earthquake the NZ Government used emergency powers to override local land use plans, rezone large amounts of farmland for housing, and allow townhouses in areas previously zoned for detached housing. You can read an excellent summary of these changes here.

So, there are many factors influencing construction productivity in Canterbury over this period. The earthquake may have spurred investment and capital into the sector, and damage to the built environment may have opened the door to more efficient construction techniques. It’s important to keep this in mind when interpreting the below analysis.

NZ construction productivity growth was almost entirely driven by Auckland and Canterbury from 2013 to 2021. In both areas, output per worker grew by around 13%, while the rest of New Zealand – much like Australia – barely saw any productivity growth.

Auckland’s turnaround is quite remarkable. From 2000 to 2013, Auckland’s productivity fell by around 0.2% per year. Once the AUP was adopted, it instead grew by 1.6% per year. This was faster than economy-wide productivity growth in Australia over the same period.

The increased size of Auckland’s construction sector also provided a double dividend. Auckland has the highest construction productivity in New Zealand in levels —reflecting a well-established fact that large cities are the most productive— and saw its share of NZ’s construction output increase from about 30% to 37%. This meant there was more construction activity in the country's most productive region.

If we decompose New Zealand’s productivity growth over this period, we can indeed see that most of it was driven by Auckland and Canterbury. Sector productivity grew at a rate of 1.1 per cent annually once the AUP was adopted, compared with effectively no growth from 2000-2013. Comparatively, from 2014 to 2021 regions outside of Auckland saw 0.76% annual growth, dropping further to 0.5% when we exclude Canterbury.

More analysis is needed to understand the impacts of zoning regulations on construction productivity, and the mechanisms at play here. But, if upzoning did increase productivity, how could it have done so? I’ll conclude with a few theories.

First, as noted earlier, a more flexible construction market could have enabled firms to grow and achieve economies of scale, and invest in new capital and technology. While this is plausible, a preliminary look at the data suggests that the average Auckland construction firm grew at about the same rate as other regions over this period. More detailed analysis using firm-level microdata will be needed to provide further insights.

Second, and contrastingly, the legalisation of the "missing middle" housing market may have created opportunities for smaller firms to enter the market and compete. Firms may need large amounts of capital to compete in the high-density market (e.g., needing cranes), and large amounts of land to compete in the low-density market. The “missing middle” requires neither of these. The smaller scale of projects may also have allowed for less overhead or more efficient construction techniques.

Third, there may be gains from a reduction in bureaucracy and red tape. "As-of-right" zoning could reduce uncertainties and delays from planning decisions. Firms may need fewer staff to navigate government processes. It may also allow firms to manage and time their capital more efficiently. Moreover, a shift from greenfield to infill may have eliminated the need to coordinate construction with infrastructure development.

Fourth, there is a direct effect on land productivity, as each square metre can now hold more floorspace. Further, upzoning gives builders a greater selection of sites from which they can feasibly choose to develop. Instead of being limited to the subset of sites planners have said they can use (some of which might be on hills, etc.), they can go out and simply find those that can be developed at least cost.

Overall, these points suggest that economists should take seriously the idea that zoning reform can improve residential construction productivity. A common critique of upzoning is that it is only a partial solution and might not be useful in a tight construction market. It is argued that in poor macroeconomic conditions the construction sector may not be able to scale up sufficiently to meet demand. However, if upzoning can both legalize new construction and improve the efficiency with which it is carried out, it could address multiple aspects of our housing affordability challenge.