Ireland’s housing bubble and bust has become emblematic of what not to do in housing debates around the world. The only problem is nobody agrees what actually went wrong.
Ireland had arguably the world’s largest housing bubble and crash in the 2000s, with prices quadrupling in the decade to 2007, even while supply soared, before crashing by more than half between 2007 and 2012. Unsurprisingly, this extreme experience has been the subject of much research. Housing has become a critical economic, social and political issue in many cities across the high-income world. At its worst, it even threatens the very concept of living standards in high-income countries, gobbling up a third or even half of the disposable incomes of individuals and households in some locations. But it wasn’t always like this. Adjusting for inflation, the price of housing in high-income countries underwent ups and downs in the century to the 1960s but the trend was largely stable. Though the timing varies by country, it has only been in the last half-century or so that the price of housing has shot up like a hockey stick.
As the world’s largest economy, the United States has been the highest-profile market to make this transition, along with a number of other countries that have followed the same patterns. Ireland is at the extreme end. Like a Rorschach test, people look at Ireland and see whatever suits them most in making arguments about housing and economic policy.
But many of these arguments rely on simplistic myths about what happened. Contrary to many of these claims, Ireland was not a story of overbuilding caused by laissez-faire policy, or an experience that defied standard economics. Ireland built very few ghost towns – housing excesses, where they occurred, were a product of government tax policy, rather than irrational markets. And supply and demand perform very well in explaining the trends. Failing to understand these basics will mean we are susceptible to making the same mistakes all over again.
I have spent much of the last fifteen years studying the Irish housing system, following it from the heights of the Celtic Tiger bubble to the following crash and the subsequent decade of rising prices. There are, to my mind, three myths that have emerged about the Irish housing market that muddy the waters in our understanding of housing markets not just there but everywhere.
Myth #1: Ireland built too many homes
One myth I meet again and again – even in Ireland – is that Ireland built too many homes, creating ghost towns of thousands of empty homes. But this is based on a misunderstanding of the data. To understand why, we first need to understand what Ireland’s ghost estates were – and were not. Over the last decade, journalists I have spoken to, from many different countries, have invariably asked about the ghost estates they heard about. Often, particularly in the earlier years, they would ask if they could go visit one within a short distance from Dublin. Unfortunately, I had to let them all down.
Don’t get me wrong. As of 2011, Ireland had nearly 3,000 developments classed by policymakers as ‘Unfinished’. They were enough to hold 180,000 dwellings. In a country that had 1.7 million occupied dwellings in the 2011 Census, that is certainly a large number. But while the number 180,000 conjures up images of streets of pristine homes slowly disintegrating for a want of people, the truth is quite a bit different.
As figures based on comprehensive surveys show, nearly half of those homes – just under 86,000 – were not only complete but already occupied. Irish policymakers called them ‘unfinished’ rather than ‘ghost’ estates for reasons other than marketing. From the off, Ireland’s ‘ghost estate’ problem was much more about some people living in developments that were partially complete than no one living in fully complete ones.
And of the nearly 95,000 homes that weren’t occupied by 2011, nearly two thirds simply didn’t exist – they were homes for which planning permission had been granted but no work had ever started, while a further 9,000 were in the early stages of construction when work halted. Just one in ten of the units said to be in ghost estates – 18,600 – were complete and sitting empty in 2011. Roughly the same number were neither fully complete nor abandoned in the early stages of construction – around half were classed as ‘near completion’ and the other half earlier in the build process.
In Dublin, if you were looking for large developments of empty houses or apartments, in 2011 you would find two developments with more than 200 empty apartments – Clancy Quay near central Dublin and Tallaght West in the south-west of the metro area. The largest number of empty houses in the city was Stocking Wood, which had fifty-three empty houses as well as eighty-six occupied ones. In a city of 1.5 million residents, these are hardly shocking numbers.
By the mid-2010s, these developments and the others in the survey were all almost entirely fully occupied, and the strength of demand for Irish housing meant that the annual survey of unfinished developments was wound down.
So with just a bit more understanding of the information available, ‘ghost estates’ shrink from 10% of Ireland’s housing to just 1% – or maybe 2% if you count the units that could be completed for relatively low cost. Given that most countries would look to add something like 1% to their housing stock each year, it is clear that the myth of Ireland as the poster child of ghost estates is way wide of the mark.
Myth #2: Ireland’s bubble was a case study in laissez-faire
A second myth about Ireland’s housing bubble is that it was the product of the market left to its own devices. Once again, though, this is at best only a secondary part of the story. In reality, policy failure – as much as market failure – drove the bubble.
Of the 850,000 or so homes built in Ireland between 1996 and 2015, fewer than one in six were urban apartments. But these are precisely the homes that Ireland needed to build as it converged to European patterns of urbanization and household size. In contrast, over four times as many – 60% of all new homes built – were built in rural areas, and the vast majority were for larger households, which were a declining share of Ireland’s population.
Ireland’s exports have shifted from being dominated by agriculture, a land-intensive activity, to being dominated by tradable services and sectors such as pharmaceuticals, which are prone to clustering in and around cities. In other words, if housing had followed the economy, Ireland would have urbanized rapidly in recent decades. But building urban homes was an unpopular message in a political system skewed towards rural areas.
Again, the devil is in the detail. In the years leading up to 2008, Ireland’s rural locations did add far too many homes than would be needed in the long-term, under any realistic measure. Between 2000 and 2008, for example, the north-west rural County Leitrim, a three-hour drive from Dublin and a two-hour drive from Galway city, added nearly 7,700 new dwellings. This was a remarkable feat given that in the previous Census of 1996, Leitrim had fewer than 8,400 households.
Similar volumes were built in Leitrim’s north-west neighbours, County Longford (82 new dwellings per 100 households) and County Cavan (83). Dublin, on the other hand, added just 38 new dwellings between 2000-2008 for every 100 households it had in 1996.
The concentration in the Upper Shannon region in Ireland’s north-west is telling – and also gets to the heart of why blaming Ireland’s bubble on laissez-faire leads one badly astray. This region was subject to what were known as ‘Section 23’ tax reliefs. These incredibly generous tax reliefs created something of a no-lose situation for prospective home builders. They could build new rental homes and write the construction costs off the tax bill for their rental income.
The key detail in small print, however, was that the rental income against which you could write the costs off did not have to come from the property itself. Not only could it be from other properties in Ireland that year, if you didn’t have enough rental income from your portfolio that year, you could roll it forward to other years.
In other words, policy created a bizarre situation where the optimal tax strategy was to build homes in designated areas such as Ireland’s Upper Shannon region, but it didn’t really matter if you found tenants or not – the important thing was to spend the money on building, in order to reduce your tax bill on other rental income.
Unsurprisingly, given this distorted playing field, the sector responded with gusto, building tens of thousands of homes in places designated by politicians as ‘in need of renewal’ (read: votes) as they didn’t need to worry about long-term need. As can often happen with heavily planned systems, the construction of new homes followed the preferences of planners and politicians, rather than the emerging needs of households.
Unsurprisingly, all this extra supply has made homes in areas subject to Section 23 very affordable, even over a decade on from the scheme being wound down in 2008. The average price of a home in Leitrim at the start of 2020 was €125,000, effectively unchanged from at the start of 2000 (€121,000).
Dublin, however, and Ireland’s four other main cities suffered from rocketing prices, with construction diverging away from the areas of greatest need. In Dublin, the average price of a home at the start of 2020 was €375,000, over 50% higher than the figure for the start of 2000 (€228,000).
Myth #3: Ireland’s experience refutes the laws of supply and demand
The third myth is that recent trends in Ireland prove, somehow, that there is no link between prices and supply. One argument is that both construction rates and prices were rising rapidly during the 2001–2007 period – and this correlation disproves the effect one would expect to see, where new supply would lower prices.
The problem with that line of reasoning is that it is just that – simply a correlation. A core part of my doctorate at Oxford entailed disentangling the various factors that drove housing prices in Ireland up and down. Indeed, I started my research wondering whether Ireland was a case where more supply fuelled the fire, rather than dampening it down. What I found was the importance of credit conditions.
At a fundamental level, any housing market is a race between demand, which can be measured by income per household, and supply, which can be measured by the housing stock per person.
Between 2001 and 2007, these two forces largely cancelled each other out in Ireland: demand rose by 23%, while supply increased by 25%. Demographics – in particular the ratio of people per household – pushed prices up by an average of 1% per year while the ‘user cost’ – borrowing costs less expected increases in prices – pushed prices down by 1% a year.
How, then, did housing prices manage to increase 8% per year above inflation in this period? The answer lies in the mortgage market, outside of price conditions. Between the late 1990s and 2007, regulators did nothing as credit conditions relaxed dangerously. First-time buyers went from typically being offered a loan of three times their savings (in other words needing a deposit of 25%) to being offered a loan of nineteen times their savings (a 5% deposit) or more.
Indifference by policymakers is perhaps the wrong description – they had actively dismantled the Building Society system that had lasted for over a century in the late 1980s. The intention had been to free mortgage lending up from relying on local savings, which was a good idea in a growing country. But Ireland’s transition to having financial institutions being able to tap into global capital markets required a clear set of rules, which were entirely lacking.
Related to this, Ireland experienced a fall in interest rates, as the economy entered the European Union’s Economic & Monetary Union in the 1990s and ultimately joined the euro. The cost of capital in the country went from being high and volatile, as the country was a small and unimportant capital market with exchange rate risk, to being low and stable, now part of a larger global currency zone. Mortgage interest rates fell from above 10% in the early 1990s to below 5% a decade later.
The dramatic change in mortgage market conditions meant that the ratio of mortgage credit to household deposits rose from roughly 60% to over 175% in the space of a decade.
A similar episode had happened in the UK in the 1930s, but in that case the massive influx of cheap credit led to a building boom of such scale that prices were largely flat. In Ireland, the price of rental accommodation was largely unaffected by credit conditions and expectations. Although prices responded to the new supply, they reflected both the glut of credit supply and the constraints on housing supply. As described earlier, policy had a large role to play in determining the level and composition of new supply.
Unsurprisingly, tapping into all this extra leverage meant that the regular battle between supply and household income faded into the background. Of the 7.9% average annual increase in real housing prices 2001–2007, my research estimated that almost all of that increase each year (7.5 percentage points) was due to looser credit conditions – as captured by the ratio of credit to deposits.
The key point is that all the extra homes built in Ireland still absolutely meant that housing prices were lower than they would have been if those homes hadn’t been built – as the case of Leitrim shows.
For the country as a whole, one would need to think of supply not in number of homes, but instead in the value of housing – such as the quality and location of new homes – to see the bigger picture in a way that was meaningful. By this measure, the stock of housing per person in Ireland rose from €25,000 to almost €40,000 between 1995 and 2007. Since there was a 10% increase in the housing stock, which was estimated to lower prices by 12.6%, the overall effect of this supply was to lower prices by over 40%.
The current challenge
Another variant of the argument that ‘supply and demand doesn’t work’ is about more recent times. Surely, the argument goes, if supply responds to prices, it would have responded to the chronic and growing shortages that emerged first in Dublin’s rental market in the early 2010s and then spread around the country.
By way of background, based on estimates of Ireland’s population mid-century, there need to be between 45,000 and 50,000 new homes built every year – across all tenures – for decades. This is based on Ireland having a population of 6.1 million by mid-century, compared to just 5 million now, and household size in the country converging over coming decades to its European peers, from roughly 2.75 persons per household now to something like 2.25 in the 2050s.
As Ireland also converges in relation to urbanization rate, from a low level for high-income countries now, we know also that the bulk of new housing should be in the country’s main cities and towns and built for smaller households. Unfortunately, the average number of new homes built between 2015 and 2020 was just 15,000, one third of what is needed. Why would the market miss such an obvious opportunity?
As all good students of Econ 101 know, however, supply does not depend on prices alone – rather supply depends on the ratio of prices and costs. And this is where Ireland’s misguided tax reliefs have delivered a double-whammy. In addition to skewing the delivery of new homes away from where they were needed, they also created a ratchet effect in costs. According to the Society of Chartered Surveyors in Ireland, the cost of building a home has trebled since 1990, from €70 per square foot to over €210 three decades later. Of this increase, fully half of it came in just four years between 1997 and 2001. The tax reliefs unleashed demand far in excess of capacity, and the sector struggled to respond to its impact.
Unfortunately, however, wages and other costs can be a lot more rigid in adjusting downwards to a negative shock than upwards to a positive shock. Indeed one of the principal reasons the Great Depression was so severe stemmed from an inability of wages and other costs to fall in response to the collapse in demand. In Ireland’s case, the same phenomenon transpired: the extraordinary rise in costs during the early 2000s when the construction sector was booming was not followed by a substantial fall in costs in the later 2000s. Instead, the persistently high costs – especially once tax reliefs on those costs expired – made it unviable to build new homes for years after the end of the bubble.
Put another way, nobody worried about the cost of building a home rising from €125,000 to €225,000 when credit had pushed prices up from €150,000 to €350,000. But when prices crashed back down to €175,000, the system had a real problem on its hands – one that policymakers have been loath to touch. Put simply, the case of Ireland shows that where housing becomes unviable to produce, due to higher costs, then we as a society are missing out on the benefit that additional housing would bring.
Supply and the solution: another myth?
Scarcity is now Ireland’s biggest housing problem – especially for smaller households and around its largest towns and cities.
In general, the policy system has responded with small steps in the right direction. The Central Bank nipped an incipient bubble in the bud in 2014 with new mortgage rules, limiting how much households could borrow relative to the value of the property and to their own incomes. In 2016, the Department of Housing recognised that excessively high minimum standards could price out lower-income households and set limits on what could have been a harmful ‘race to the top’ by local authorities. And in 2018, the Department of Finance introduced a new planning code for purpose-built rental housing. Policymakers linked building regulations to the length of ownership by a single owner, which will help Ireland fund the homes it needs.
These steps in the right direction continued this year with reforms of social housing, away from a reliance on the market. The Irish system is moving towards a model where social housing is based on cost-rental, in which the supply of subsidised housing is linked to the cost of provision rather than market prices. Such a system is largely agreed to be the most effective use of providing housing for lower-income households.
Similarly, a newly-formed Land Development Agency has been given the remit to take state-owned land with out-of-date uses and bring them into the twenty-first century. And October’s budget saw much needed reform of land use, with a tax on land that is zoned for residential use, serviced by infrastructure but not being used or built on.
But, despite this progress, the outlook is far from certain. A number of vocal commentators – either blissfully or wilfully ignorant of the overwhelming academic consensus on the matter – have claimed that new purpose-built rental accommodation will make the situation worse, not better. And their arguments have found a receptive audience among some politicians, including those who sit on Dublin’s City Council. The first draft of the city’s new Development Plan, which will cover the period 2022–2028, now seeks to strictly limit the construction of rental housing – in a city desperately in need of tens of thousands of more rental homes!
Thankfully, a pipeline is already underway that should see the city get between 40,000 and 50,000 new rental homes over the next few years. While welcome, though, they will not be enough on their own to solve the rental shortage.
But there is a remaining challenge in the State’s housing apparatus, the newly established Office of the Planning Regulator (OPR).
The OPR’s job is to review the Development Plans of each of Ireland’s thirty-one local authorities and ensure that they comply with national policy. To do this, though, they are upending the system used in other countries. England’s Housing Delivery Test, for example, is a housing floor, which penalises councils which don’t meet certain supply targets over a set period of time. Ireland’s OPR, however, has interpreted the numbers as a ceiling for local authorities.
How on earth, you might ask, has Ireland ended up with almost all parts of its policy system trying to get lots more housing built – but the key cogwheel doing its utmost to hold new housing back? The answer, ironically, is Ireland’s own policymakers falling for the myths of the last bubble. It seems that the key personnel of the OPR believe the north-west of the country built too many homes in the 2000s because of state inattention and a wayward market, rather than as the result of extraordinary state effort to bring about that outcome. Without those reliefs, there is now little risk that new homes will be built where there is no long-term need.
Unfortunately, myths about Ireland’s housing market haunt not only international views but also the views of local Irish policymakers. As demonstrated above, there is no evidence that Dublin or Ireland’s other main cities suffered from a structural overhang of housing in the early 2010s. And in fretting about the potential for the market to build housing where it’s not needed, local policymakers have misunderstood the key role played by public policy in driving housing away from where it was needed most. And lastly, while a simplistic analysis might point to the correlation of prices and supply in the mid-2000s, rigorous analysis shows that lots of new supply had a big impact on prices that only became apparent once credit conditions normalised. As much as anyone else, Ireland’s policymakers need to learn the lessons of the country’s recent past.