There’s an outfit called Turner & Townsend that measures construction costs in 33 cities around the world every year.

They measure two categories of costs: inputs, and outputs. Construction inputs include wages, concrete, glass and paint. Construction outputs are the all-in cost per square metre to develop different types of buildings: houses, apartments, offices and warehouses, for example.

The most recent Turner & Townsend report was released recently. What the report shows is an odd disconnect in Dublin construction costs. In Dublin, inputs such as wages and glass are cheaper than in the three other European cities studied – Amsterdam, Munich and Paris.

But in Dublin, outputs – the all-in cost of building different types of structures – are higher than in peer cities. Houses and apartments are particularly expensive to build in Dublin, relative to peers.

Why do construction costs matter? They matter because they feed through to house prices. When construction costs are low, it’s easy to build more houses, which keeps a lid on prices. When they’re high – as they are now – it’s hard for the market to bring prices down.

The following chart is from a Central Bank of Ireland report on housing from 2019. What it shows is the relationship between house prices and the number of new units per year between 1995 and 2018. 

As you can see, the data forms two distinct patterns. One pattern, which is coloured dark blue, shows the relationship between house prices and construction in the years 1995 to 2007. The other pattern, in pink, shows the same relationship after 2010. 

This indicates something changed after 2010. Before 2007, a real house price index of 100 induced the construction of 60,000 homes or so. After 2010, the same level of house prices induced only 20,000 new homes. 

This is why construction costs matter: higher costs mean less construction, fewer homes, and higher house prices.

The following chart shows how the cost of building 17 categories of buildings differs from the European average. In all but four categories, Dublin is more expensive. On average, it’s 10.2 per cent more expensive to build in Dublin.

Of the 17 categories, the two where the cost difference is most pronounced is the two residential categories: high-rise apartments and townhouses. It costs an average of 30 per cent more to build an equivalent high-rise apartment in Dublin than in peer cities.

And it costs 32 per cent more to build a townhouse than in peer cities. 

Dublin is obviously an expensive place – but then so is Paris, Amsterdam and Munich. The following chart shows the difference between Dublin and the average of the other three, for 18 different construction inputs. Across the 18 inputs, Dublin was an average of 5.5 per cent cheaper than the average.

In 12 of the 18 categories, Dublin is cheaper than the other-city average. These include important categories like concrete, concrete block, steel and rebar. 

Dublin wages were lower in five categories out of six — an average of 15.7 per cent lower overall. 

The story so far is that construction costs are higher in Dublin, but inputs are lower. What might be going on?

There's more to construction costs than wages and raw materials. There are preliminaries: things that don't go into the finished product, but that need to be paid for regardless. They include site service costs, health and safety documentation and site running costs. In Dublin, preliminaries are a bit higher than the average at 14 per cent of the total, compared to a European average (among 13 cities this time) of 12.3 per cent.

Then there's the profit margin. You might think, if inputs are cheaper and outputs are more expensive, that profits must be big. But Irish construction profits are lower than the average for 13 European cities — 4.0 per cent versus 5.5 per cent. 

The addition of preliminaries and margins doesn't clear things up. Dublin preliminaries are a bit bigger than peer cities, but margins are a bit smaller. Overall it's a wash.

So, high Irish construction costs are a puzzle. Thanks to the Turner & Townsend data, we know what doesn't cause them: it's not wages, margins/prelims, or raw materials.

What else could it be? I see three suspects. 

1. Is the data wrong?

I asked Turner & Townsend about their methodology. The company acts as a consultant on projects all over the world, and that's how they gather data on finished costs per square metre for different types of builds. For Dublin residential schemes, Turner & Townsend's MD in Dublin Mark Kelly told me they had advised on 15 to 20 schemes in the last five years.

For input costs, Turner & Townsend surveys contractors twice per year. It surveys tier one and tier two contractors in the 33 markets it covers.

As much as possible, the company said it tries to make apples-to-apples comparisons. But Kelly told me, occasionally there are discrepancies in how the survey is interpreted. For example, he said that the difference in the price of concrete blocks between Dublin and peer cities probably came from a different interpretation of the question.

Another thing the survey doesn't capture is shortages and delays. Wages and prices are different from quantities. Maybe Irish sites are more prone to shortages than in other places. 

A separate report by Core Research found 63 per cent of Irish construction companies were finding it difficult to recruit. Though, to be fair, this is also a global problem — 70 per cent of Turner & Townsend respondents globally said they were experiencing skills shortages.

There's also the fact that Dublin has moved up the rankings over time. You wouldn't expect this if it was just a figment of the data. The following chart shows cumulative cost inflation in Dublin relative to peer cities. Dublin had been ranked second most expensive, now it's ranked most expensive.

2. What about productivity?

Low productivity might explain it. 

Productivity basically describes how efficiently companies turn inputs into outputs.

The construction industry has a problem with productivity. And not just in Ireland. A 2017 McKinsey study found global labour-productivity growth in construction has averaged only 1 per cent a year over the past two decades, compared with 2.8 per cent for the rest of the economy and 3.6 per cent for manufacturing.

How does Irish productivity compare? Well, the way they measure these things is to divide gross value added by the number of workers and number of hours worked.

But that number doesn't work here. Because that number mixes up two things.

It mixes up productivity from operations — ie how efficient and automated the processes are, and the impact of planning or regulations. And that deserves a category of its own.

It is plausible that Irish contractors are less efficient than those in other countries. Ireland has only 14 contractors that employ 250 people or more, and then 47,000 small contractors. Perhaps Irish builders don't have the economies of scale.

3. What about planning and regulations?

The last explanation for high Irish construction is that planning rules and building regulations are tougher in Ireland than in other places.

Regulations include things like minimum apartment sizes, the number of building cores per unit, dual aspect ratios, green space per unit, and affordable units per market price unit. Planning costs include things like uncertainty over planning decisions that are subject to judicial review. 

Ronan Lyons made this case in a piece last February. In the following chart, he showed Irish apartments are legally required to be bigger than in other European cities.

We know Ireland is building far fewer homes in the second boom than in the first one. We know construction costs have gone up a lot, and help account for the difference. We know construction costs aren't driven by wages, or profits and preliminaries, or raw materials. 

That leaves three possibilities: bad data, low productivity, or onerous planning rules and building regs.