Ronan Lyons | Personal Website
Ronan Lyons | Personal Website

Blog

Housing and Ireland’s competitiveness jigsaw

Housing should always be primarily a social issue. If the country cannot house its own citizens, this should be disturbing enough for remedial action. However, most people would acknowledge that, with everything the last five years have revealed, when housing is only an issue of social justice, it doesn’t feature high enough on the agenda. What needed to happen was for housing to become an issue of international competitiveness. Unfortunately, this has now happened – but fortunately, this means that it is far more likely to feature on the radar of key Ministers, policymakers and the Cabinet.

Housing has traditionally been regarded as domestic issue. Just as firms based in Britain typically ignored EU bashing in the media, as they didn’t want to meddle in a local issue, so firms based in Ireland typically focused on more obvious inputs to competitiveness, such as corporate tax rates and membership of the European single market. No firm wants to get a reputation for meddling in domestic affairs. However, firms based in both countries have realised that ignoring a political issue can be precisely the worst course of action. In the UK, firms are now scrambling to respond to the referendum result on EU membership last year. The powerhouse of the entire British economy – the hub of financial services centred in the City of London – could have its foundations taken away in coming years.

And in Ireland, large employers here have become increasingly noisy on the issue of accommodation. To see why, you need look no further than the National Competitiveness Council’s ‘Cost of Doing Business’ profiles. The standard FDI project coming to Ireland now is a services, or perhaps R&D, facility and for such projects, labour costs make up three quarters of their overall costs. This is a sea change from the 1980s and 1990s, when the IDA’s main targets were large manufacturing facilities, where half of all costs were imported inputs.

And the single most important item of spending is housing, which typically makes up one third of disposable income in cities. So of the three quarters that relates to wages, one third is down to housing. This means, simply put, that housing is one quarter of Ireland’s competitiveness. Therefore, when rental prices in Dublin rise by 65% in less than six years – or sale prices rise by almost 50% in less than five years – this does not just put pressure on those on lower incomes. It also erodes a key source of wealth for this country: jobs serving foreign customers.

When Irish-based subsidiaries of foreign-owned firms have to take unusual steps, such as offering bonuses to existing employees to temporarily house new ones, while they find their own homes, it is only a matter of time before HQ finds out that Ireland has a problem. In a world where capital is chasing skilled labour, and where skilled labour wants to enjoy all the amenities offered by a vibrant city, it is the cities that can house growth which will win. Irish cities are hopeless at accommodating growth currently. Dublin has now been allowed to grow up and so, since the 1980s, has started to sprawl. Commuting, though, is consistently ranked as people’s least favourite use of time and is not a viable long-term way of life.

But it is not just Dublin that is suffering. Cork and Galway are home to very large employers, with thousands of workers, in particular in pharmaceuticals in Cork and in medical devices in Galway. But both cities are struggling to accommodate the growth. One home was completed in Galway City Council in May last year… and while this is an extreme example, only 52 new homes were started in the city in 2016. In a rapidly growing city of 80,000 people, it should be adding closer to 800 a year – over 15 times the current level of activity. In Cork, just 310 new homes were started. In a city with a population of 200,000 people, it should be adding 2,000 new homes per annum.

And the problem is not limited simply to the building of homes, although it is clearly at the heart of the issues. Access to schooling, childcare, public transport and infrastructure are all related to the decision of where to live and work. Back in Galway, traffic has become such an issue that there are stories of three-hour commutes after work from the east of the city, where most of the business parks are, to the west of city, where much of the housing is. Three hours to cover ten kilometres is not sustainable.

Housing is and always will be first and foremost an issue of human rights. In a modern, high-income country, access to housing should be guaranteed by a system of subsidies that top up a family’s means to meet their need, where appropriate. But housing is also a competitiveness issue. If we don’t figure out how to build enough homes quickly, it will start costing the country jobs.

Are tracker mortgages haunting Ireland’s mortgage market?

Why are mortgage interest rates in Ireland so high compared to other countries? The ECB collects information on the average interest rate of a new mortgage for all euro countries and, for December 2016, the average for the euro area was less than 1.8%. In France, it was 1.5% and in Finland it was just 1.1%. In Ireland, however, it was 3.2%, the most expensive of the 19 countries covered, quite a bit above the next most expensive countries, Cyprus (3%) and Greece (2.9%).

At one level, this seems to be nit-picking. Compared to the bad old days like the early 1980s, when mortgage interest rates were almost 20%, is there that much of a difference between 3.2% and 1.2%? Perhaps more importantly, is an average of less than 2% sustainable? If growth were to pick up in Europe – as has been happening over the last year or so – then the pressure would be on for the ECB to normalise its own rates, with knock-on effects in markets such as Finland and France.

Nonetheless, the “Programme for a Partnership Government” last May included a commitment to examine the structure of the market for mortgages in Ireland and see if there were policy options available to bring the cost of mortgages down. This week, the Competition & Consumer Protection Commission – the amalgam of the Competition Authority and the National Consumer Agency – launched a consultation on this topic, as its analysis starts. I have no doubt that there are many issues which will affect the cost of mortgage credit at the margin. This includes Ireland’s small market size, which may deter international banks from entering. This may also include regulations, which – for example – prevent credit unions from lending more than a certain fraction of their capital for more than ten years.

However, I suspect that deep down, the problem lies with the legacy of the bubble and, going deeper, Ireland’s predilection for variable, rather than fixed, rate mortgages. To explain, it wasn’t always the case that Ireland was an expensive place to borrow. The ECB figures go back to 2003 and for much of 2003, 2004 and 2005, Ireland was one of the cheapest places to borrow. And therein may lie the rub.

Ireland’s mortgage market is skewed towards variable rate mortgages and in the bubble, these were trackers. Irish banks fixed the margin that borrowers would pay above the ECB rate, for the lifetime of the mortgage. They then gambled – in spectacularly unsuccessful fashion – that they would be able to roll over their own debt at a cost close to the ECB rate. Of course, when the so-called “Great Moderation” came to an end, and the “Great Recession” started, Irish banks suddenly found that borrowing short to lend long was a recipe for wipe-out.

But the trackers that helped to destroy them still exist. So, with ECB rates currently at 0% (yes, zero!), there are plenty of Irish households who borrowed in the bubble who are paying 1% or less interest. Irish banks, however, struggle to borrow themselves at such low rates and cover costs associated with being a financial intermediary. And if their own lenders aren’t subsidising legacy trackers, there is only one other group that could: new borrowers. The reason Irish borrowing is expensive now is precisely because it was so cheap ten years ago.

Rolling the clock forward, bubble-era trackers will continue to exist right through the 2020s and into the 2030s. Bizarrely, an ECB president in 2035 will still be able to financially affect Irish households for decisions they made in 2005. A PhD in Economics is not required to see that this is hardly ideal.

In most other high-income countries, the typical mortgage product is a fixed rate mortgage. In Denmark, for example, Danish banks are required to borrow for as long as they lend. This turns them into warehouses, breaking up the journey from international capital markets to the individual borrower. Households borrow for rates that are fixed for the term. The best banks are those than can add on the slimmest margins.

This is hopefully where the CCPC’s study will take it. A good rule of thumb is that when Ireland sticks out in a graph, find out if it’s a good thing or a bad thing and then find out why. The answer might not be popular or a simple fix. But it’s better to be right than to make things worse by applying a quick fix.

==

This is a version of a piece that originally appeared in the Sunday Independent.

Will Rent Pressure Zones worsen the rental market logjam?

A point I’ve made a number of times before is that turnover in the residential sales market is a key measure of the health of the housing market. At one point, in the depths of the crash, the typical property in Connacht was changing hands roughly once every 100 years – which we know is far too long and thus the market was unhealthy. The same metric can be used to examine the rental market, particularly in light of the figures this week that rental inflation continues to climb, pushing average monthly rents to fresh new highs. Unfortunately, we don’t yet have the full figures from the 2016 Census but we can at least give a rough picture of turnover in the rental market.

For example, in 2006, there were about 145,000 homes in the country in the private rented sector. That year, just under 80,000 rental properties were listed on daft.ie, suggesting that the typical property changed hands slightly more often than once every two years. In Dublin, indeed, the typical rented home came back on the market after 18 months. By 2011, this had slowed down a little, and of the 300,000 or so private rented homes in the country, just under half were listed on the market that year. In both Dublin and in the country at large, the typical lease length was two years. What is interesting is that, despite the huge growth in the sector – the number of private rented homes more than doubled in just five years – the turnover was similar.

It is likely that when the figures for 2016 come out, they will show that the private rented sector has not shrunk since 2011. So in other words, we are still looking at a sector with at least 300,000 homes in it – and possible more. Even if it were just 300,000, though, what has changed dramatically is the fraction that comes on the market every year. In 2016, there were less than 70,000 rental listings nationwide. What this means is that fewer than a quarter of rental homes came on the market. Flipping that stat around, the average lease length is now more than four years, having been two years in length as recently as 2011.

On the face of it, this may seem like good news: Ireland is becoming a nation of long-term renters. This is the benign interpretation. There is nothing at all wrong with people renting for longer (As long as they are also putting money away for their futures). However, the figures are also consistent with logjam in the market.

It is reasonable to think that the typical landlord in Ireland does not fully review the rent as often as they are legally allowed. What this means is that many tenants in Ireland enjoy below market rents. This creates an incentive to stay put, even if in other circumstances you might have moved. So when a group of students goes away for the summer, they may keep paying rent, so as not to “lose their place”. Or where a group of young professionals are sharing a house, this group may keep rotating, despite couples being formed or people change jobs.

This has become a much more pressing concern, now that Ireland has moved from a “rent resets when empty” model to system where whatever level of rents held in late 2016 is now frozen in, for at least a few years. The Rent Pressure Zones risk creating – or re-creating, for those with long enough memories in the Irish rental market – a two-tier rental market. This would be the case even if the entire country were deemed a Rent Pressure Zone (as increasingly seems likely, given the figures in this week’s Daft Report).

On one tier are those properties whose rents were increased steadily by landlords throughout the period 2011-2016. These are effectively market rents, albeit from now on the increases will be slightly capped. The other tier consists of those properties whose rents were not increased by landlords during the same few years, believing that they would get the chance to in the near future, once their tenant ultimately moved on. But the act of bringing in Rent Pressure Zones itself has made moving on less likely: why give up a low level of rent?

Not only that, when they do ultimately move on – for example, if they are buying a home to live in – tenant and landlord are likely to cooperate and find a new tenant through word of mouth. What landlord wants the hassle of 200 people turning up to view something that’s being rented at 40% below market rates?

Again, word-of-mouth may sound benign, but those who lose out will disproportionately be those on lower incomes, on some form of State support, or who have just moved to the country. Like many other aspects of life in Ireland, the dice in the rental market have just been loaded in favour of insiders.

==

This is a version of a piece that originally appeared in the Sunday Independent.

Nasty Nick and Decent Dave: how do you measure a problem like rental inflation?

Late last year, new measures were announced to try to limit rent increases, primarily to give tenants – and the Government – some breathing space while issues inhibiting the supply of new homes are addressed. On the face of it, the measures were straightforward: in any area where rental inflation is above 7% for a year and a half, rent increases are now capped at 4% per annum.

There was a bit of a kerfuffle in the first few days, when there was confusion about the formula used to define 4% per annum. And there continues to be political football played with which areas were to be granted ‘Rent Pressure Zone’ status immediately.

Of course, when something becomes legislation, nothing is ever straightforward. What does “area” mean, for example? In everyday conversation, it is fine to talk about Dublin rents increasing by a certain fraction and to include Bray and Greystones in that definition of Dublin, but it an entirely different thing if the law has designated Dublin a rent pressure zone but not Wicklow.

Even if we were clear on what each area meant, things remain less than ideal. What if the Galway market is starved of apartments but has enough family homes? Rental inflation for apartments in Galway could be well above 7% but for houses it might be close to zero. Can rent pressure zones distinguish between the two markets in the same location?

But whatever about these technical worries, there are two more fundamental problems that threaten to undermine the system of Rent Pressure Zones just as it gets underway. The first is that it effectively attempts to control the price of apples while measuring the price of oranges. The motivation for this measure is to protect sitting tenants. However, rent inflation is benchmarked not against rents paid within leases, i.e. by sitting tenants. The RTB measures rents and rent inflation by looking at rents paid by new tenants. These are completely different beasts.

Secondly, what makes this policy such a substantial change from previous policy around rent control is the fact that it applies across tenancies as well as within tenancies. It has been standard, for more than a decade, that while landlords are limited in their ability to increase rents for sitting tenants, when those tenants move on, the rent can effectively ‘reset’ to the market level. This new system shatters that link.

The end result of both these features is best described by the parable of Nasty Nick and Decent Dave. Nick and Dave both own adjoining semi-detached properties as investments and rented them out in 2010 for €1,000 per month. Dave, being a decent bloke, has gone easy on his tenant – a family with young children – and only increased the rent twice in the following six years. The family now pay €1,200 a month for their home.

Meanwhile, Nick, being a different sort of character altogether, has increased the rent as often as the law allowed. The family living there in 2010 are long gone, priced out and living somewhere further out but more affordable. In their stead are four young professionals, who collectively pay €1,700 a month for the house.

Nick’s property, having been on the market a few times since 2010, is what is used to measure rent inflation. Dave’s should enter the calculations for cost of living – after all, his tenants are people too – but the way rents are measured in Ireland currently, the stable rent he offers is not reflected at all in the headline measures.

And the worse the situation gets on the open market, the more tenants in situations like Dave’s are going to stay in their home and enjoy their below-market rent under the radar. This means that, while it may have been the case 15 years ago that the typical rented property changed hands every 12-18 months, many renters are now staying put for 3-5 years. In other words, the more measured inflation in rents goes higher, the greater the pressure to stay put – exacerbating the problem with measured inflation.

It gets worse. Suppose the family living in Dave’s house are moving out to buy their own home. One of the reasons Dave was so lenient with his tenants was that he knew, once they were gone, he would be back to the market rent. This has now been taken away from him. Even though his tenants have moved on, and even though his next-door neighbour is charging a rent €500 more than he is for the same property, the most he charge his new tenants is 4% a year more.

We should not be surprised, then, if Dave decides to sell up. The problem is that, if he sells to another investor, they face the same level of rental income as Dave does. The new system of Rent Pressure Zones has effectively devalued Dave’s investment… unless he sells to an owner-occupier.

Now that Rent Pressure Zones have been introduced, there is likely to be a silent exit from the market of landlords who were muddling through and who thought they were doing the right thing by being nice to their tenants. If they stay in the market, their asset is devalued, so they have a strong incentive to leave by selling to an owner-occupier.

If Rent Pressure Zones are here to stay, the first step in addressing these problems is to better measure rent inflation, by reflecting not only “between lease” rent inflation in the headline stats but also “within lease” inflation. We should not be surprised if this dramatically changes our picture of what has been happening renters since 2011.

==

A version of this article appeared in the Sunday Independent Property Section on February 12th, 2017.

Would you rather be fat or tall? Time for Dublin to grow up

The figures from this latest Daft.ie Rental Report make grim reading not only for tenants but for policymakers also. The annual rate of rental inflation – at 13.5% in the final quarter of 2016 – was the highest in the history of the Daft.ie Report, which extends back to 2002. The rate of inflation in rents is high across the country, with only Connacht seeing inflation of less than 10% currently, of all the major regions. (Its rate of inflation is 8.9%.)

Indeed, in 45 of the 54 sub-markets analysed, there is double-digit inflation in rents currently. In Dublin, rents are now rising by almost 15% a year, the highest since the middle of 2014. It means that rents in the capital are now up almost 65% from their lowest point in 2010 and are a full 14% higher than their previous peak at the start of 2008. In the four other cities, inflation has eased back somewhat in recent months but remains between 10% and 13%.

Unsurprisingly, the issue is a severe imbalance between supply and demand. Fewer than 4,000 homes were available to rent across the country on February 1st. While this marks a 10% increase on the same figure a year previously, it is still well below the lowest point for availability during the Celtic Tiger. Then, stock on the market reached its lowest point in April 2007, when slightly fewer than 4,400 homes were available to rent nationwide.

The crucial difference between 2007 and 2017, however, is the size of the renting population. We won’t know for another couple of months how many households in Ireland rented at the time of the 2016 Census, but it is unlikely that the number then is less than the number in 2011. At the time of the 2011 Census, roughly 475,000 households lived in rented accommodation, just under 30% of the total. In 2006, there were just 145,000 households in the private rented sector, together with a further 155,000 in the social rented sector.

It is likely that much of the growth of roughly 150,000 extra renting households between 2006 and 2011 was the private rented sector, given the limited social housing available in recent years. Thus, the private rented sector doubled in just five years – and may have increased further since. And yet, instead of availability doubling, it has fallen, even compared to the worst days of the Celtic Tiger. And in Dublin, the problem is most acute.

Recently, the Government launched a consultation period for its National Planning Framework. This will be the successor to the National Spatial Strategy, which aimed to develop a system of gateway towns and hub towns around the country. Much of the media discussion around the launch focused on the idea of Dublin “eating up” the country and how policymakers should focus instead on ‘balanced regional development’.

While policy can help a city thrive, it is foolish to think that politicians have control over the economic forces at work in determining city size. Indeed, one of the most remarkable laws in economic geography is known as Zipf’s Law. It states that the biggest city in an economy is on average twice as big as the second biggest city, three times as big as the third biggest, and so on. This is true more or less all over the world, with – if anything – the world’s largest cities being too small, rather than too big.

Some respond to this with “Ireland is different”. In some ways, it is. Dublin is more than twice as big as Cork. But the same holds true for Vienna in Austria and Budapest in Hungary. What all these countries have in common is that their current borders do not match their past ones. The island of Ireland is a much more natural economic unit than the Republic and the North are individually. And sure enough, once Belfast and Derry are included, Zipf’s Law starts to look like a much better description of Irish cities.

What does this have to do with policy? It means that population is not a game of redistribution. Trying to redirect people from Dublin to other cities in Ireland merely stunts the growth of the country as a whole. If you want to stimulate growth in the fifth, fifteenth and fiftieth biggest cities in your country, you should encourage growth in your largest, not try to stop it.

This does not mean, of course, that sprawl should be ignored. There is absolutely no reason for Dublin to take up nearly as much space as it does. Greater Tokyo, with a population of over 27 million people, fits into 9,000 square kilometres. Greater Dublin, with a population of less than 2 million people, takes up almost 6,000 square kilometres.

Dublin can drive population growth across the country without “eating up” the countryside. For this, though, three things need to happen. Firstly, Dublin needs to be allowed to grow up. Height limits of between four and six storeys in one of Europe’s fastest growing cities merely translate into lost jobs and higher rents. After all, why be fat when you can be tall?

Secondly, homes need to be built on brownfield, as well as greenfield, sites. Looking around Dublin, it is littered with grossly underemployed land, in the form of army barracks, golf clubs and bus depots, built on the fringes of the city in the past but now in prime central locations. A land tax would help achieve this reuse of our scarce and valuable urban land.

But, even if these regulatory changes were brought in overnight, we still would not see the apartments being built that are needed to stem rental inflation. And the main reason for that is the extremely high cost of construction in Ireland, compared to other high-income countries. An audit has been promised – let’s hope it delivers.

==

A version of this appeared as the commentary to the 2016Q4 Daft.ie Rental Report, released on February 14th, 2017.

How to build enough homes

6,200 newly built homes were sold in Ireland last year, an increase of almost 100 on the total for 2015, and twice the number of new home sales seen in 2011, 2012 and 2013. Given the early part of 2015 was period of significant activity – as borrowers rushed to use their mortgage approvals from before the Central Bank rules – the fact that volumes increased again in 2016 can only be viewed as a good thing.

So far, so good. However, the bulk of what is being built still does not come on to the market. When the final figures come in, it is likely that 2016 will have seen 15,000 new dwellings completed. Thus out of every ten dwellings built, only four ever come on the market – the other six are self-builds. Indeed, of the 80,000 new dwellings completed since the start of 2010, just 32,000 have come on the market.

This huge gap is not normal – in most countries, the bulk of new homes built would be for the market, rather than self-build. It reflects a long-standing problem with land usage in Ireland. Those building for themselves often do so on family land, which is free, or else as one-offs on cheap land far away from urban centres. Such building imposes huge costs on the rest of society, with knock-on effects for services, such as healthcare and broadband, which rely on density to be cheap.

Overcoming this problem will require a fundamental rethink about how we use land in Ireland. One aspect of this is about regional development. A narrative has emerged that “Dublin is too big”. This misses the point completely that in any economy, and certainly in an economy the size of Ireland, leading cities determine the size of the rest of the economy: in simple terms, if we want our fifth, fifteenth and fiftieth biggest cities to be bigger, we need our biggest city to be bigger. This is not something we get to choose – these are the laws of economic geography.

In addition to rethinking the role of our cities, we also need bottom-up measures that encourage far better use of land. As I’ve written previously in this column, our cities and towns are riddled with “last use” rather than “best use” examples of land usage, with army barracks, industrial estates and bus depots taking up sites that would be far better used for residential or commercial purposes. To encourage public and private organisations to use land better, we need to introduce a land tax – a measure that would also penalise speculation, land hoarding and cynical vacancies.

But that is only part of the problem. A far bigger part of the problem is not that there are 9,000 self-builds but that there are only 6,000 dwellings built for sale. In a country growing as rapidly as Ireland, between 40,000 and 50,000 new homes are needed each year. Allowing 10,000 of those to be self-builds, this means that construction of homes for sale needs to be roughly five times the size it is now.

And that huge gap between what is needed and what is happening is due largely to the high level of construction costs in Ireland. This refers to hard costs, so issues around profit margins, VAT and site costs are contributing to this. Experts say that it costs roughly 50% more to build a home of 100 square metres in Ireland than in other parts of Europe. This is the nub of the problem.

The second half of 2016 saw two important changes to the housing market. Firstly, first-time buyers of newly built homes were given access in the budget to their past income tax, of up to €20,000, to lower the deposit needed. Secondly, that minimum deposit requirement for first-time buyers was lowered even further by the Central Bank. These measures will boost prices and, in so doing, the argument goes, stimulate new supply.

Unfortunately, such measures do not address the underlying problem in the new home segment today. Ireland – with its rapidly declining average household size – desperately needs new homes other than three- and four-bedroom semi-detached houses in estates. It needs student bedrooms near universities, central apartments in high-rise blocks for young professionals, and suburban mid-rise apartments for well-to-do downsizers, as well as many more types of home.

Until the high cost of building in Ireland is addressed, though, we are likely to see only baby steps towards a sector building at least 40,000 dwellings per year. The Minister for Housing announced an audit of construction costs in Ireland, compared to its peers, last October. The results of that study can’t come soon enough!

==

A version of this appeared in the Sunday Independent Property Section on February 6th, 2017.

Property in 2017 – the year ahead

2016 saw a number of important policy shifts in relation to property in Ireland, particularly in the final few months of the year. The first was creation of a Cabinet-level Minister for Housing, a sign that – with the change of government – the severe shortage of housing was finally being taken seriously by those in power.

The second substantive change was the long announcement, from about mid-summer until the Budget, of “help” for first-time buyers of newly built homes in the form of a tax rebate. This kicks off in earnest with the new year and is likely to combine with the third change. This was the revision of the Central Bank’s mortgage rules, which came through in November. These revisions mean first-time buyers no longer require a deposit larger than 10%, even if they borrow more than €220,000.

Combined, these two measures create something of a two-tier market. Take two otherwise identical families, on the same household income and with two young children. Their only difference is that one rents the two-bedroom apartment they live in currently, while the other owns it.

Both families are looking to buy a newly built family home in the Greater Dublin Area, at a cost of €400,000. The family that own their apartment will need to produce a deposit of €80,000 (20%), while the family that rent their apartment will – once the tax rebate is factored in – need a deposit of just €20,000 or 5%.

In terms of how this will affect the market, both the rebate and the change to Central Bank rules will have the effect of further stimulating demand and thus pushing up prices. In a way, they are complementary measures. The tax rebate will have the biggest effect on cheaper new homes (those costing between €200,000 and €400,000), while the rule change will have a bigger effect on more expensive homes, including second-hand ones.

Either way, the expectation for the year coming is for a return to house price inflation in Dublin after something of a two-year pause, with average prices in the capital increasing by just 4% between early 2015 and late 2016.

Increases of at least 5% and probably closer to 10% will also be expected in the rest of the country, as strong demand interacts with a lack of supply. The country needs at least 40,000 new homes a year – and probably closer to 50,000 once obsolescence and immigration are factored in. But the current hope is to get construction of 25,000 new homes by 2021.

There is an excessive focus on the “starter home”, however. In fact, when you look at Ireland’s demographic structure, there are close to enough family homes in the country to cater for our families. What Ireland lacks – more than any other high-income country – is apartments.

This shortfall is unlikely to be addressed in the coming year, however, as the cost of building apartments is prohibitive, compared to average incomes. The break-even monthly rent for a two-bed apartment – even with free land – is roughly €1,600 but in most parts of the country, a two-bed rents for less than half this.

So, while there will be a fuss about the vacant site registers (and in time the vacant site levy), until the hard costs of construction have been dealt with, expect little improvement in the chronic lack of accommodation for one- and two-person households. Government Ministers lodging complaints against developments in one of the small number of areas where apartments are viable certainly doesn’t help.

One area that has recovered somewhat in recent years and is likely to continue to strengthen in 2017 is purpose-built student accommodation. Ireland is in the middle of a long higher-education boom. This appears to have been missed by the Higher Education Authority: a 2015 report of theirs predicted student numbers to rise from 168,000 to 193,000 in the decade to 2024. Instead, there are likely to be 193,000 students as early as this September, seven years ahead of schedule!

In that context, all new purpose-built student accommodation is welcome, even if it will only really cater for better-off students. The reason this is the case is the same problem that bedevils residential construction in Ireland currently: how expensive it is to build. Looking at the pipeline of student accommodation, it is likely that this will be only just enough to meet new demand and will do little to take existing students out of the general private rented sector.

So 2017 is likely to be another year of very strong demand for all types of residential property: sales, rental, new, second-hand, urban, rural, houses, apartments, student living and assisted living. It is also likely to be yet another year of weak supply in all these segments.

Solving this would mean dramatically reducing the cost of building a home – by something like 35%. Even if the causes of this cost gap are identified, it is likely to take a year to bring costs down and then a further two years before these lower costs translate into anything like the level of construction the country needs.

So don’t be surprised if there’s a similar-sounding piece this time next year!

==

This piece originally appeared in the Sunday Independent on January 8th, 2017.

Supply, supply, supply: the new housing mantra

Below is my commentary to the latest Daft.ie Sales Report, which reviews the market in 2016. Its overall point is that Ireland needs roughly three times as many new homes to be built per year as is currently the case.

==

Nationally, the average list price rose by 8% in 2016, very similar to the 8.5% seen in 2015. Compared with static prices in 2013 – although this masked huge regional differences – and an increase of almost 15% in 2014, perhaps this, then, is the new normal. The graph below shows the number of markets (out of a total of 54) that fall into one of four categories: falling prices (in year-on-year terms), rising slightly (0-5%), strongly (5-10%) or unsustainably (above 10%). As you can see, the most common change has gone from falling (the blue line; pre-2014) to 10% increases (red line; 2014 and 2015) to 5-10% increases in the last couple of quarters. The green line (0-5% increases) only briefly emerged as “normal” before fading away in recent quarters.

House price changes, by market, quarter and inflation bracket
House price changes, by market, quarter and inflation bracket

Normal does not mean healthy, however. We know that in a healthy housing system, any extra demand for more housing is offset by more supply – in other words, the real price of housing should be stable, once general inflation is taken account of. In Ireland, general inflation has effectively been zero not just over the last 12 months but indeed over the last decade.

So Ireland is currently trapped in a situation where housing prices are increasing far faster than prices in the rest of the economy. This is not sustainable but the latest indications are that this high rate of inflation is embedded in the market, due to strong demand and weak supply.

We know from the initial Census results that the country added 170,000 extra people between 2011 and 2016. Given the likely composition of new households – between 2 and 2.5 people per household on average – this means that the country added almost 75,000 new households in those five years.

We know from the same source, the Census, that there were just 17,000 new homes added to the stock of dwellings in the last five years, once holiday homes are excluded. In other words, for every ten new families formed, just two new dwellings were built, for the entire period from 2011 to 2016. (Completions numbers were much higher than this, but this includes properties built during the bubble and only connected to the electricity grid more recently. It is also a measure of “gross” construction and doesn’t account for buildings going obsolete.)

Bad as that may seem, the picture is worse again. Firstly, the period 2011-2016 was largely one of net emigration, with 125,000 people leaving between 2011 and 2015. There is a clear move toward net immigration, though, emigration falling from 90,000 to 75,000 since 2013 while immigration has risen from slightly more than 50,000 in 2012 to 80,000 this year.

Migration is driven by those in their 20s and 30s, in other words the very groups forming households and starting families. Based on the 2011 Census, we know that every additional 10,000 migrants require on average 4,000 dwellings, so even if net migration remains relatively low – at say 20,000 a year over the next few years – that will add 8,000 to the number of new homes required annually.

This is in addition to the core demand resulting from “natural increase”, in other words a surplus of families being formed over families dying. A fast way of checking the size of this natural increase is to compare the size of the cohorts of women aged 30 and 80. There are roughly 35,000 women aged 30 in Ireland currently, which gives a good baseline of household formation – ultimately, the vast majority of these women are likely to be part of one household each. There are just 10,000 women aged 80. Thus, there is a natural increase in number of households each year of at least 20,000 and closer to 25,000.

On top of this, demographics are changing – not least, people are living longer. Coupled with other factors, including a greater fraction of people who do not have any children, separation and divorce, Ireland’s average household size has fallen from more than 4 people in 1971 to roughly 2.7 people today. However, it is still the highest in Europe, where the average is just 2.3.

This may sound like a small difference but it is hugely important for how many new homes are needed per year. For example, if Ireland’s population did not increase but the average household size fell from 2.7 to 2.3, an additional 300,000 dwellings would be needed. Realistically, that convergence will take time, but it is likely that declining household size will add at least 10,000, if not 15,000, to the number of new homes needed each year.

The last factor when figuring out how many new homes are needed each year is one that is most often forgotten: obsolescence. The Department of Housing and CSO estimate that roughly 0.8% of the housing stock goes obsolete each year: in other words, the typical dwelling lasts about 125 years. This means that, every year, about 16,000 dwellings fall out of use.

That figure seems somewhat high and, while 125 years may be an accurate guide for rural cottages, urban properties typically remain in use due to renovations. But even a depreciation rate of 0.5% a year would mean 10,000 dwellings are needed annually just to stand still.

Adding all these up, there are roughly 10,000 dwellings needed each year to offset obsolescence, a further 10,000-15,000 needed to accommodate Ireland’s smaller households, between 20,000 and 25,000 on top of that to house the natural increase – and to top it all off, likely a further 8,000 or so due to net migration.

In total, Ireland needs at least 40,000 new dwellings a year and probably closer to 50,000. These will be concentrated in and near the urban centres and will be disproportionately homes for one- and two-person households, such as apartments, downsizer homes and student accommodation. As the latest figures show, without this kind of supply, we will all have to spend more and more of our income just to have a home.

When does a housing bubble start?

Yesterday, former Minister for Finance Charlie McCreevy appeared before the Oireachtas banking enquiry. His refusal to answer whether or not he believed Ireland suffered a property bubble that burst in 2007 was not only great TV, it also brings up some important issues. For example, the Irish Independent reports:

The conflict arose when Mr Doherty asked the former minister if he believed there had been a property bubble in the previous 15 years before the financial crisis. Mr McCreevy insisted he would only answer for his time in office and there had been no property bubble during that time… [after legal advice] Mr McCreevy said from 2003 to 2007 house prices grew at an extraordinary rate. He supposed that was a bubble. But he said: “I don’t believe the policies I pursued helped to create that bubble.”

The clear implication is that Mr McCreevy believes that, if there was any housing bubble at all, its roots do not lie in decisions made in the period 1997-2004, and that in reality there was no bubble at all. Given the title of my doctorate at Oxford was called “The Economics of Ireland’s Housing Market Bubble”, you might not be surprised to learn that I disagree.

First, I think it is important to note that there are two ways of diagnosing bubbles. They can be thought of as statistical bubbles and economic bubbles. A statistical bubble is one where the growth rate in the price of an asset, such as housing, grows at a rate that is unsustainable for any reasonable period of time. Between 1995 and 2007, house prices in Dublin increased by 300% in real terms (i.e. stripping out inflation), or 12.2% a year. Between 1997 and 2004, McCreevy’s term in office, the increase was 136%, or 13.1% a year. (Nationwide figures are comparable, although slightly lower for the period as a whole, although not necessarily in every year.) Thus, by any statisticians metric, it was a bubble – put another way, if 12% growth had continued for 25 years, a house costing €100,000 in 1995 would have cost €1.7m by 2020.

Economists like to get at causes, though, and a 10% increase due to – for example – a lack of supply has very different implications for what policymakers should do than a 10% increase due solely to first-time buyers needing a smaller deposit and thus being lent more. To economists, a bubble in asset prices is not just any old increase in prices, it’s an increase in prices due to excess capital/money. In the housing market, this means too much mortgage credit. Of course, to sustain people borrowing and lending too much, you need expectations. So the two ingredients for an economic bubble are over-optimistic expectations and excessive credit.

The graph below is, in effect, the one-chart summary of one of my D.Phil. chapters: what drove real house prices in Ireland during different market cycles (measured in changes per annum). Falling income (measure here relative to supply), pushed down house prices in the 1980s, together with higher real interest rates (a term that includes house price expectations). This reversed somewhat during the period 1987-1995, which income, as well as demographics (fewer people per household) pushing up prices by nearly 5% a year. Note, however, that credit conditions – measured by the ratio of mortgage credit to deposits – were not pushing up house prices as this time.

Irish annual house price growth, by driving factor, 1975-2012

The period 1995-2001 saw very strong house price growth, driven by a combination of tailwinds, including incomes growing proportionately faster than housing supply. By the time these supply constraints were removed – through the follow-up to the Bacon reports and other measures – borrowers and lenders now expected rapid house price growth. These unrealistic expectations were facilitated by rapidly easing credit conditions. Crucially, almost all house price growth from 2001 to 2007 was driven by a relaxing of credit conditions.

What this means for Mr McCreevy is that it is simply not credible for him say that there was no housing bubble on his watch. Bubbles, driven by asset factors in particular expectations and credit, grow out of booms, when demand outstrips supply. The 1995-2001 boom created the 2001-2007 bubble. A Minister for Finance in 2004 could have tried to burst the bubble, but not prevent it. To do that, the Central Bank mortgage rules would have had to have come in not in the mid-2010s but in the late 1990s.

A post-script. Mr McCreevy has come to be known as a man who strongly believes in pro-cyclical fiscal policy. As he clarified yesterday, as Minister, he believed “When you have it, you spend it.” Exhibit B below is a graph I show my first-year Economics students. It is the average all-in tax rate paid by a household on an average income, by country and year from 2000 to 2007 (source: OECD). At a time when the Irish economy was growing more rapidly  than ever before, the state took a declining share of these higher incomes. I think a strong case can be made that much of the austerity undertaken by Ireland in the period since 2007 would not have been necessary if tax rates had been in line with other developed countries and that Ireland sorely missed a Minister for Finance able to spot that Irish fiscal policy was increasingly unsustainable and take the steps necessary to correct the path.

Average all-in tax rates, by country and year

 

Expectations, credit and house prices

Happy new year to all readers – after an-almost two-year hiatus (or at least severely restricted service), I hope to return to regular blogging this year and have revamped the site to reflect how times have changed since 2008, when “Version 1” launched.

Of course new year means new quarter and new quarter means house price reports… The latest Daft.ie House Price report is out this morning. The PDF is available here. For me, the key takeaway is as follows: house prices fell in the final quarter of 2014 and it seems very unlikely to have been statistical noise or a seasonal effect.

35 areas are analysed in each report. For each of the first three quarters of the year, an average of 32 showed quarterly gains in asking prices. For the final quarter, this flipped, with 30 of 35 regions showing a fall. For Dublin, this was the first quarterly fall since mid-2012. (Given the size of increases earlier in the year, a one-quarter fall still leaves the year-on-year change large and positive: 20% in Dublin and 8% elsewhere.) Broadly speaking, a mix-adjusted analysis of Price Register transactions shows the same. While it is only one quarter, it seems more than just a statistical blip.

For me, the check-list of what matters for house prices contains five items: [1] household incomes, [2] demographics and [3] housing supply (“the fundamentals”); and [4] credit and [5] expectations, these last two being the “asset factors” that can create and destroy housing bubbles. None of the fundamentals changed dramatically in the final three months of the year (the only thing you could argue was a slightly higher volume of listings in Dublin), so the change after September must be due to asset factors.

The Central Bank proposed in October to cap residential mortgages as early as January 2015, although this could not affect prices directly in 2014. So the last remaining candidate is expectations.* The quarterly Daft.ie report includes findings from a survey of housing market sentiment. This survey indicates that, yes, those active in the housing market did revise downward their expectations about future house price growth, particularly in Dublin. Whereas those surveyed in September expected a 12% increase in Dublin house prices over the next 12 months, this had fallen to less than 5% by December. I expect that the Central Bank would be happy if it were the case that their proposals strengthened the link in people’s heads between fundamentals (in particular people’s incomes) and house prices.

As for my opinions on the Central Bank guidelines themselves, I submitted a response to the Central Bank’s Consultation Paper, which is available online here. The TL;DR version is “max LTV good, max LTI bad”. I made similar points at an Oireachtas hearing on this and related topics in late November.

* Some have argued that the end of Capital Gains Tax relief was what drove trends in the final months of 2014. The theoretical reasoning behind this is unclear – it is not obvious that this would affect supply more than demand – while practically speaking, it is also not clear how this would have managed to infiltrate the vast bulk of the market which is not of interest to investors. When asked what they thought was driving house prices, those active in the housing market rarely mentioned tax factors, instead picking credit and supply as the main factors.