Ronan Lyons | Personal Website
Ronan Lyons | Personal Website

June 2018

Good news or bad news? Housing and the latest population figures

Five years ago, as many wondered whether the green shoots were actually emerging, the Central Statistics Office published its population projections for the future. The numbers were based on a detailed analysis of the 2011 Census figures.

When forecasting population, there are three main moving parts: births, deaths and net migration. So to plan ahead, the CSO used scenarios for these factors. For example, they had two fertility scenarios.

The optimistic scenario was that Ireland would maintain its 2010 fertility rate of 2.1, just enough to replace the population. The more pessimistic scenario was that this rate would fall to 1.8 by the mid-2020s and remain there into the 2040s. This would leave Ireland on a par with North European countries, whose fertility rates have actually risen since the 1980s.

The flip-side of the natural increase is death. For mortality rates, there were no scenarios – simply an assumption that life expectancy would increase by 7 years for men between 2010 and 2046 (to 85 years) and by almost 6 years for women (to 88.5).

By far the trickiest element to forecast for Ireland, though, is net migration. Whereas the annual natural increase has been quite steady, net migration jumps around. Ireland’s people are unusually mobile.

Five years ago, the CSO adopted three migration scenarios. In each of the three, Ireland was expected to lose population each year to 2016 – by 19,000 a year in the most optimistic scenario and by 25,000 a year in the most pessimistic. After that, the three scenarios diverged – ranging from a loss of 5,000 persons a year to a gain of 30,000 as the average annual change from the 2020s through to the 2040s.

The net effect of all this was that Ireland’s population in the low-fertility medium-migration scenario would rise from 4.6 million in 2011 to 5.6 million by 2046. These numbers in turn underpin Eurostat’s bolder predictions of population out to 2080. Eurostat’s central forecast is that Ireland’s population would reach 6.2 million by 2080.

Fast forward to 2018 and the latest population projections have been released by the CSO. Ireland’s fertility rate is already at 1.8 and the optimistic scenario now is the pessimistic one from five years ago: that it stays there. The pessimistic scenario, in turn, has been downgraded and plans for fertility to fall to 1.6.

Projections around mortality are largely unchanged – with the expert group behind the report still seeing a 1.5% improvement in in mortality rates each year, as they did five years ago.

But the joker in the pack for Ireland’s population is – as it always has been – net migration. Remember that Ireland was expected to lose 19,000 people in 2016? It gained 16,000 that year instead.

This has led the expert group to go back to the drawing board. Its most optimistic scenario has been bumped up. Instead of having to wait to the 2020s for 30,000 people a year will come here, this will happen from 2017. This change may seem small – but it’s an extra 50,000 people by 2021.

The medium and pessimistic scenarios are more radically changed. The medium scenario has been doubled – from net immigration of 5,000 a year to 10,000. And the pessimistic scenario has changed entirely: even in this scenario, Ireland will be a net recipient of people – 5,000 per year.

What is the effect of all these changes? Substantial, it turns out. In the “baseline” scenario – lower fertility and medium migration – Ireland’s population by 2046 would be 6.1 million, not 5.6 million.

Think about that. Our economic circumstances have changed so much in the last five years, that the thirty-year outlook has gone up by 50%!

Or to put it another way, Eurostat thought – based on how things looked back in 2011 – that it would take Ireland over 60 years for its population to go above 6 million. Now, that milestone is just 25 years away – or so our best guess goes.

How does this relate to housing? One of the clearest ways I can find to summarize our housing shortage had been that 2080 target. Thinking that far ahead allows people to stop worrying about elections or their own circumstances and focus instead on the bigger picture.

With a population of 6.2 million by 2080 – and with something approaching normal levels of urbanization and household size – Ireland would need to build roughly 1.7 million apartments over six decades. (Its need for family homes is, contrary to popular perception, negligible as almost all population growth will come from households of 1-2 persons.)

Dublin alone would need an apartment block of 200 homes every week for decades. To be clear, by ‘apartment block’, I mean urban housing for 1-2 person households. This includes student accommodation and co-living spaces, core urban high-rise and suburban low-rise for downsizers.

It also includes independent and assisted living complexes for Ireland’s older residents. And that is a topic in and of itself – two thirds of Ireland’s 1.5 million extra inhabitants will be aged 65 or over.

Amazingly, though, Ireland’s timeline to build 1.7 timeline just got halved. The country is now expected to reach a population of 6.2 million by 2050, not 2080.

And even more amazingly, the expert group restrained themselves when thinking about net migration – due to housing shortages. So if housing weren’t an issue, we would get there faster.

If there were any doubt that housing is now the single biggest bottleneck facing the Irish economy, this week’s CSO report puts that to bed. The country needs to get building – in particular building urban homes for 1-2 person households. And twice as fast as we thought we needed to.

==

An edited version of this post was originally published in my column in the Sunday Independent.

Brisk business at the top end of the housing market once more

The latest Daft.ie Wealth Report shows that – as prices continue to rise  – so does Ireland’s housing wealth. There were over 800 transactions of at least €1m during 2017 and by all accounts there will be even more seven-figure purchases this year.

When a property is transacted, of course, that wealth is – at least for an instant – liquid cash. Most of those selling a property soon buy another so the wealth becomes embodied once more in real estate form.

Another equally important question, though, is who is sitting on a property worth a million. Daft.ie’s Wealth Report is the only one of its kind to estimate the number of property millionaires in the country. It does this by using a detailed form of analysis, looking at the value of 25 different property types in almost 400 micro-markets around the country.

Of these nearly 10,000 segments, just 58 have an average value of over one million euro – from four-bed terraced homes in Sandycove to five-bed detached homes in Ballsbridge. These property segments cover an estimated 4,600 dwellings.

This is an increase of almost 20% in the number of property millionaires in the space of just one year. A year ago, there were just 3,800 property millionaires. In other words, rising property prices over the last year have created more than two new property millionaires every day!

How is the top end of the market holding up? The figure accompanying this piece shows the average annual change in property prices across Ireland’s top 20 markets. All are in South Dublin, with the exception of Howth and Enniskerry.

The pattern is clear: before the Central Bank rules came in, these markets were taking off, with prices in late 2014, just before the rules were announced, running at more than 20%. The rules coming in saw prices stabilise, with almost no change between the third quarter of 2014 and the first quarter of 2016.

Since then, though the lack of supply and strong demand have caused inflation at the premium end of the market to tick up again. And in early 2018, for the first time since the Central Bank rules were introduced, inflation at the top end of the market was higher than the national average: 8% compared to 7% nationally.

At least some of this push at the top end of the market is Brexit-induced. Anecdotally, architects will tell you stories of couples returning from London with a seven-figure budget buoyed by the extraordinary level of prices in the British capital.

On top of that, it seems clear that any Brexit dividend in terms of FDI jobs will be at least as large in the tech sector as in financial services. The on-going uncertainty around the nature of Brexit has meant that the ‘born online’ firms – most of who have large operations in both Dublin and London – are favouring the Irish capital for new projects.

This stems from an uncertainty about whether their highly-skilled and very multinational workforce will be able to enter and stay in the UK, once it leaves the EU. Ireland has remained explicitly open for business – issues around housing and visas aside – and that certainty appears to be reaping rewards.

The premium end of the market is always going to be the most illiquid. If you have a €5m home in Temple Gardens or Eglinton Road, your pool of buyers is always going to be far smaller than if you have a €500,000 home in Rathfarnham or Raheny.

That lack of liquidity can trick some commentators into thinking the market at the top is in trouble. The figures here show that this is not the case – at least not currently. The market for Ireland’s most expensive homes is booming currently.

What underpins the top end of the market? Looking through the list of areas that make up Ireland’s Top 20 markets – Killiney, Dalkey, Sandymount, Sandycove, Enniskerry, Howth and Rathmichael, for example – it is clear that nature is a luxury.

Natural features such as mountains and coastline are over-represented at the top end of the housing market. The richer we get, it seems, the more we want to be able to enjoy what nature provides for free, either as a picture out our window or as a playground outside our front door.

Nature may provide these things for free – but they don’t come cheap in the housing market. New research I have been working on, jointly with Stephen Hynes and Tom Gillespie at NUI Galway, finds that natural amenities add significantly to property values. The best located properties – those with not only the views but on the coast too – are on average a third more expensive than similar dwellings, at least in structure, but without the same amenities.

If nature is one of the luxuries at the top end of the Irish housing market, history is the other. Many of the areas I mentioned above – together with others in the Top 20 such as Ranelagh, Rathgar and Ballsbridge – are home to some of Ireland’s finest older homes.

By definition, you can’t build Victorian or Georgian property now, so this limited supply gives those homes added attraction in the market. One of the contradictions of the housing market is that both energy efficiency and age are prized in the market. Joint research with Sarah Stanley and Sean Lyons, of the ESRI, found that “vintage premiums” are understated if you don’t take into account energy efficiency. Compared to similar properties built in the 1990s, properties built before World War 2 is worth 20% more – but if you left out energy efficiency from the model, you might think the premium was just 15%.

Scarcity, in other words, is what drives the top end of the market. And with lots of demand – and based on Brexit uncertainty, this seems set to continue – it seems the coming year will create even more property millionaires.

Delivering an extra 20 homes an acre in Dublin

Ireland’s housing challenge over coming decades is substantial. Within that, though, the need for new housing supply is not spread evenly. As things stand currently, Dublin will be, regardless of how tough the politicians talk, the centre of Irish economic and population growth in the 21st century.

Let’s put some numbers on it. Over the coming half-century, the country will need almost two million new apartments. This is needed to accommodate changing demographics and increasing urbanization, as well as regular population growth.

Dublin will need about 600,000 of those apartments. For apartments, think everything from purpose-built student accommodation and urban medium- to high-rise right the way through to suburban low-rise for downsizers, as well as independent and assisted living complexes. The number includes both market and social housing need – ideally integrated in the same developments.

Over half a million new apartments – principally to accommodate households of one or two persons – is a housing challenge unlike anything the city has faced before. It is effectively an apartment block of 200 homes every week in the city for the next 3,000 plus weeks… in other words, every week until after most of Ireland’s adults currently are long dead.

An obvious concern you might have, upon hearing these numbers, is where these apartments are going to go. The contiguous city of Dublin houses 1.2 million people currently in its 11,500 hectares, or a little under 30,000 acres in ‘old money’.

Adding an extra 600,000 homes to the city’s 30,000 acres means an average of 20 new homes per acre. But this kind of abstract maths hides the rather important detail that many of these acres already have homes on them, and policymakers can’t simply kick out existing residents and bring the builders in.

This means that the ‘Parisian dream’ of high density without tall buildings – espoused by, for example, the Green Party – is simply not possible. In this dream, Dublin can accommodate all its new residents by simply building to a height of six storeys.

This might work – but only if you could CPO existing homes and demolish them. I don’t see that happening.

What the city needs to do instead is start a detailed plan of where the 600,000 new apartments will go over the coming decades. To do this, the four local authorities will need to work together and identify sites where 10 or 12 storeys – and higher closer to the centre – can work.

There are lots of options, not least Dublin Port, which is over 600 acres of prime land that could house as many as 60,000 of the homes needed. But that is just 10% of the total needed – and it would come with a fight, as Dublin Port has longed resisted any attempts to move again.

The green fields around Dunsink are an extraordinary opportunity for the city to grow inside the M50. Given it’s not in a central location, though, densities would be lower. The 1,000 acres or so there could add another 60,000 homes.

But that’s still only 20% of the city’s growth need.  A further 20%-30% is likely to come from density near the urban rail infrastructure outside the core city.

But that leaves city planners trying to find space for hundreds of thousands of apartments on brownfield sites all around the city. We need to think creatively, as a city, about re-using space. This means army barracks set up by the British two hundred years ago – and golf clubs established a hundred years ago – may not be the best use of space today.

In addition to thinking creatively, Dublin needs to grab opportunities when they present themselves. DIT’s move to Grangegorman frees up a bundles of city-centre sites for redevelopment. The college on Cathal Brugha Street is very likely to become part of a renovated Gresham hotel and add much needed hotel space in the city.

But other sites could help address the city’s chronic shortage of apartments. Take the Kevin Street campus. The block bounded by Bride Street, Camden Row, Liberty Lane and Kevin Street is roughly six acres and is located in one of the most fashionable parts of the city, just west of Camden Street and just north of Portobello.

Granted, one acre of the six is St Kevin’s Park, a hidden gem of the city. But five acres becomes seven, as there are also two acres just across Camden Row, currently split between a cash-and-carry and another DIT building.

So the Council should consider zoning the lands for high density mixed use, including residential, and imposing a land tax instead of development levies and commercial stamp duties. This way, those seven acres could easily act as a test case for integrated social and market housing, at relatively high densities, and using a cost-rental model.

“If I were you, I wouldn’t start from here” runs the joke about the Irishman’s reply to a tourist asking for directions. If we could start all over again, the high-density low-height model in Paris and Barcelona could work for Dublin too. But that option is not open to us – so we will need to rethink and reuse land, and build up.

The need for Dublin is an extra 20 dwellings on every acre of the city. What is the plan?

==

An edited version of this post was originally published in my column in the Sunday Independent.

Will Irish house prices fall?

A couple of weeks ago, Philip Lane, the Governor of the Ireland’s Central Bank, suggested while appearing before an Oireachtas committee that Irish house prices could fall in coming years. The comments generated a lot of interest, given his role. In particular, he said that a combination of factors could combine to not only halt price rises but even bring down sale and rental prices. These factors, he suggested, were increased housing supply, including of student accommodation, higher interest rates, and potentially higher unemployment due to Brexit.

How much weight should we put on the Governor’s comments? Some thought the Governor even mentioning price falls means that he sees this as likely. Others, however, see in his remarks the typical economist’s way of seeing the world: many possible outcomes – one of which is falling prices – so why deny that?

To get a sense of just how much weight to put on the potential of housing prices – either sale or rental – falling in the coming years, it is important to understand the forces that push them up and down. In brief, there are five forces that affect average house prices, three of which also affect rents. The first three factors affect both sale and rental prices and are what you might term fundamentals, as they reflect the real economy. The first is household income. This captures both average income per worker and also the number of workers per household (i.e. unemployment).

When the Governor spoke of Brexit being a force for lower house prices, he was referring to this channel: that due to the British economy cutting itself off from its closest trading partners, the Irish economy – as its single closest trading partner – would be affected enough that incomes would fall and unemployment would rise.

The second factor, again a fundamental one affecting both sale and rental prices, is demographics. This is best measured as the number of persons per household. In Ireland, there has been a steady fall in the average household size over the past half-century, although one that has paused in recent years, due to a lack of housing.

But in big picture terms, it is hard to see how Ireland will do any other than slowly follow all its peers towards a smaller average household size. If over coming decades Ireland’s average household size falls from 2.75 people to something like 2.25 people, this creates a need to produce over 10,000 new homes – mostly urban apartments – each year.

The third and final fundamental is housing supply. This is best thought of in euro terms, rather than in number of dwellings, as a rural cottage and an urban apartment may have very different values.

But for simplicity, let’s think of housing supply as the number of dwellings. Construction is simply the addition to the stock of dwellings. Although it is important to remember that at least some construction merely offsets the obsolescence of other, older properties. In Ireland, a country with roughly 2 million dwellings, at least 10,000 – and perhaps as many as 15,000 – homes are likely to fall into disrepair each year.

Adding obsolescence on top of the various sources of demand, the country needs at least 40,000 and closer to 50,000 new homes a year. Granted, this is certainly less than the 80,000 or so new homes built at the peak of the Celtic Tiger. But it is more than twice the number of new homes started in 2017.

So for an over-supply to contribute to price falls, rather than a shortage contributing to price rises as has been the case since 2012, the building of new homes would need to more than double in the coming two years. And this itself is assuming that none of the backlog in unmet demand since 2012, either sale or rental, needs to be met for prices to fall.

Various commentators believe such an increase in housing construction is simply not possible, given capacity constraints in the sector. Where would all the new construction workers live, for example?

The final factor mentioned by the Governor is interest rates. And this brings up the difference between sale and rental segments. Unlike rental homes, owner-occupied homes are assets as well as consumer goods. Given the favourable tax treatment of the family home, it can be very profitable to concentrate your wealth into housing.

This potential for capital gain – or indeed capital loss – looms large in the minds of would-be buyers and sellers. In surveys over the last 15 years, those active in the Irish housing market have expected everything from 10% gains a year for the next five years to 10% falls per year.

This are huge numbers compared to the equivalent changes in interest rates over the same time. Mortgage interest rates have barely budged since 2008 and have been between 3% and 5%. This is not to say that mortgage rates are irrelevant. Far from it.

But mortgage rates are only one part of the ‘user cost’ of housing and have to be offset by potential capital gains. Since the 1980s, Ireland – and Dublin in particular – has been building far too few homes relative to the underlying need. This has created an expectation of capital gains that may prove tough to shift, even with the bubble and crash we’ve witnessed.

The fifth and final factor affecting housing prices is the one Professor Lane did not mention: credit conditions. This was the one that was most responsible for the bubble and subsequent crash in property prices in Ireland 1995-2012. Think of it as the typical deposit required by the first-time buyer.

The typical deposit fell from over 25% in 2000 to just 5% in 2006. But since the introduction of the Central Bank’s mortgage rules in 2015, this property price lever has largely been removed from the equation. Banks may have a greater appetite for lending now than in 2013 but the rules prevent them from repeating the mistakes of the past.

The Governor is right to highlight that property prices are not a one-way bet. However, looking into the detail of the Irish housing system, it is tough to argue – given the tight control the Central Bank itself has over credit conditions and the large gap between supply and demand, that prices will fall soon.

==

An edited version of this post was originally published in my column in the Sunday Independent.