A housing system has two components: the market component and the non-market component. A healthy housing system needs its market and non-market elements to be responsive, in aggregate, to housing need – and not just market demand.

In my last piece I wrote about housing need, as opposed to housing demand. Given changes in the household-forming population as well as household size and obsolescence, the scale of that need in Ireland over coming decades could be as high as 60,000 homes a year. I also described a simple framework – shown again below – to guide policymakers as they seek to deliver a housing system that is based on underlying need and not just demand. 

Policymakers should target “Group C”: those with incomes too low for new market construction to be viable, and too high to be covered by subsidised housing schemes. They should be targeted by two complementary sets of policy actions. The first is moving some of that group into “Group A”, in other words including them in the group of households and individuals covered by housing supports. The second is to move the rest of that group in “Group B” – those households with sufficient income for the construction of market housing to be viable.

Theory is, of course, free from the constraints of reality. At first glance, the solid vertical line is the easier one, at least administratively if not politically. The government currently sets the income thresholds that apply for supported housing and if you are above those thresholds – typically €35,000 in the Dublin area and €25,000 a year elsewhere – you do not qualify. Of course, in practice, it’s not as simple as that. Firstly, there are a plethora of different schemes that exist, each with their own criteria. And secondly, the systems that exist have their own biases, most obviously a bias against households of one or two persons – more on that in my next column.

But those elements aside, the critical issue for housing policymakers – in Ireland and elsewhere – is the dashed horizontal line – the line of viability superimposed on the income distribution. One of the first questions a policymaker would have is where that dashed line is, relative to incomes. Viability reflects the costs of providing new homes, which has two main components, the cost of land and the cost of structures. Land costs are certainly an issue, in Ireland and in many other countries. 

But, even in systems with widespread restrictions on land use, the transaction value of a site will reflect viability as much as determine it. Take the example of a home that costs €300,000 to build, excluding site costs. The cost of building a set dwelling will be largely the same anywhere within a city and, even across Ireland, build costs do not vary substantially. But the value of the same kind of home does vary hugely across Ireland and even within Dublin. If that home would sell in one area for €310,000 and in another area for €450,000, then the site value in the first area will be roughly €10,000 and in the other area it will be €150,000: for all their inefficiencies, land values reflect viability as much as determine it.

All roads lead to construction costs

The main heading of interest, therefore, is construction costs. There is widespread, although not universal, agreement that Ireland is a particularly expensive place to build. To take one example of many, the 2021 Turner & Townsend International Construction Market Survey, released late last year, had Dublin as one of the most expensive cities in the world in which to build. The cost per square metre of building apartments in Dublin is almost 15 per cent more expensive than Amsterdam and one quarter higher than in Munich or Paris. When converted into dollars, as the report does to enable comparisons, the cost of building apartments in Dublin was given as close to $3,200, well above costs seen in Toronto and Houston and higher even than Chicago or Vancouver. Not only that, there is very little fat to trim. Of sixteen European cities, Dublin had not only one of the largest preliminaries, at 13 per cent, but also the smallest margins for contractors (just 3 per cent). 

A year ago, the Society of Chartered Surveyors in Ireland calculated the all-in average cost of building apartments – by far the housing type exhibiting the greatest shortages in Ireland – using data from almost 10,000 apartments built in Dublin. For apartments in buildings of 5-8 storeys, the all-in cost of construction ranged from €435,000 to €520,000. Even taking the cheaper suburban option, this still gives an all-in cost of €411,000 – although this does include site purchase and contributions. Even being optimistic, therefore, it is unlikely that new homes for one- and two-person households will be able to be built in this country for anything less than €400,000. 

How can we go from this figure to an estimate of the “forgotten middle” of Irish housing policy? To start, we need to convert this up-front cost into a monthly burden and then compare it to the income distribution, taking into account what is a sustainable share of income that can be spent on housing on an on-going basis. 

An overall figure cost of a new home of €400,000 can be converted into a break-even monthly rent, using two key figures. The first is the required net annual income for capital-holders. With the rise of PRS as an investment class in Ireland, this is now closer to 4 per cent per year than 6 per cent per year, as would have been the case in the mid-2010s when more opportunistic and thus less patient capital was investing in building homes here. In short, by shifting the policy landscape away from investors with a shorter horizon to others, such as pension funds, with a longer-term investment horizon, the same up-front costs are spread over a longer period of time, reducing the break-even monthly rent. 

The second parameter needed to convert up-front costs into a monthly rent is the cost of management. While this varies by operator and type of development, this is usually something like 20 per cent of gross rental income. This means that, for every €5 earned in rent, €1 is spent in managing the property, leaving €4 as net income. Where the owners of a rental development require a 4 per cent return, this pushes the gross yield required to 5 per cent per year. A two-bedroom apartment costing €400,000 to build, therefore, would need to generate €20,000 in annual rental income for it to be viable to build and thus a breakeven monthly rent of roughly €1,650 for the home. 

Already, it is clear that this figure is high relative to incomes enjoyed by average-earning households in this country. But we can be more precise about just how out of line it is. The second-last key parameter we need is the maximum sustainable housing burden, as a fraction of disposable household income. There is no one number for this but, in line with the practice of leading Approved Housing Bodies (AHBs) in Ireland, I am going to take a limit of 30 per cent: households spending more than 30 per cent of their net disposable income on housing are, in other words, spending ‘too much’ and market providers of housing will not be comfortable building new homes if inhabitants are stretched.

The last element is the Irish tax system. A gross annual income of €100,000 yields a net monthly disposable income, for a married couple, of €5,400 per month. If we take 30 per cent of this monthly income, it gives €1,620. In other words, given prevailing construction costs and yields, as well as the tax system, the construction of new rental homes is only viable for households with a gross annual income of at least €100,000.

To overlay this stark finding on the income distribution, we can use the CSO’s ‘Geographical Profiles of Income in Ireland’ to figure out the key parts of the income distribution by tenure and by local authority, as of the 2016 Census. That report gives, for example, the 10^th and 75^th percentiles of the income distribution by local authority – in other words, how much income would you need to be richer than 10 per cent or 75 per cent of households in the same area. For example, it reports that in Fingal County Council, the median income of households in the private rental sector was €49,337; €22,089 was the value at the 10th percentile, meaning that 10 per cent of market-renting households in Fingal had a gross household income of no more than €22,089 in 2016. 

The forgotten middle

This can then be combined with information on the number of renter households by local authority to give us a sense of the scale of effective housing demand at each point in the income distribution. This exercise is applied to each of the Greater Dublin Area’s eight local authorities and summarised in Figure 1. Combined, this gives an estimate of the number of private-renting households in the Greater Dublin Area, by thousand-euro income brackets, where the total number of private renting households in the GDA is just over 150,000, in line with the CSO total for 2016 – the last year for which we have this kind of information. (Don’t forget to fill out your Census forms!)

Overall, of 675,000 households in the GDA in 2016, roughly 615,000 had annual incomes between €10,000 and €200,000, the range presented in the figure, of which nearly 140,000 households rented privately. Almost half of these households earn less than €50,000 per year. 

(By comparison, almost half of social housing renters have household incomes of less than €25,000 per year, although a significant number of households in social housing have incomes well above this threshold – over 24,000 households in the Greater Dublin Area in social housing had incomes above the official €35,000 threshold in 2016. This reflects the large number of social housing schemes, of varying vintages and criteria for rental payments, as well as the importance of household composition (at time of entry into social housing).

With these four parameters in place – the maximum income at which a household can qualify for social housing, the maximum fraction of disposable income that can sustainably be spent on housing costs, the break-even monthly rent for newly-built rental homes, and the income distribution of market renters – it is possible to identify the “forgotten middle” group, i.e. households too rich for social housing but with insufficient incomes for newly-built market housing.

The chart below summarizes all this. It puts the 139,000 households in the Greater Dublin Area’s private rental market into three categories. On the left-hand side are nearly 39,000 households whose gross income is between €10,000 and €35,000 – the upper bound reflecting official cut-offs for eligibility for social housing in the GDA.

On the right-hand side are the 22,000 households whose income in 2016 was high enough (above €100,000) that they could sustainably afford the break-even rent of a newly built two-bedroom apartment in Dublin.

But look at what a small fraction of renters this is. Construction costs are so high in Ireland currently that new market rental accommodation is viable only for the top 16 per cent, or one-sixth, of renter households in the Greater Dublin Area. At the other end, 28 per cent of private renter households are, in principle, eligible for social housing support – although the fact that they are in the private rental sector perhaps tells us about the gaps between the stated thresholds and coverage.

But these two groups aside, this leaves a “forgotten Middle” that comprises more than half (56 per cent) of all current renting households in the Greater Dublin Area. These are households whose incomes are inadequate for new housing to be viable but who are ineligible, even in principle, for social housing. In a healthy housing system, this group would not exist: all households, existing and prospective, would be covered either by the market or by the social housing sector. 

This conclusion is an extraordinary one – and unfortunately not in a good way. It is incumbent on policymakers in Ireland to come up with a housing system that brings this half of renters back into the housing system, whether it is the element of the system that is for-profit or the rest. Clearly, a strategy relating to cutting costs is key – moving the line on the right further to the left. But the biggest chunk of renters are not just below the line of market viability, they are above the line of eligibility for social housing. We will need lots more from both components – the market and the supported housing sector – to bring about a healthy housing system in Ireland.

The analysis presented above has one obvious constraint – by necessity given its reliance on accurate data relating to household income, it can only include households that have already formed. But, especially considering the likely composition of housing need over coming decades, it is just as important to consider households that would form, were housing is more available and affordable. In my next piece, I will look at how to estimate this other part of Ireland’s forgotten households.