Notes 2

The following notes come from this paper:

Ireland’s housing boom: what has driven it and have prices overshot?

David Rae and Paul van den Noord
OECD Economics Department working paper No 492


Abstract

The Irish housing market is very buoyant. The housing boom is driven by strong economic growth, dynamic demographics and low interest rates. However, large tax advantages and relatively lenient credit policies by banks have also played their part, and prices may have become overvalued. To the extent that high house prices reflect favourable tax treatment, they may lead to economic inefficiencies by drawing excessive resources into residential construction. While a soft landing appears the most likely prospect, a disorderly correction of house prices would pose risks for macroeconomic and possibly financial stability. In this context, one policy lever available to the government would be a phased removal of the tax advantages associated with housing. In addition, banks should remain cautious in their lending and provisioning policies.

House prices across the industrialised world have surged since the mid 1990s – with the notable exceptions of Germany and Japan which are both still grappling with the aftermath of real estate busts in the early 1990s. In many countries, housing demand is underpinned by an easy monetary stance (Otrok and Terrones, 2005), while over a longer period tight zoning regulations have exacerbated the upward movement in property prices in and around growth centres (Glaeser et al., 2005).


This paper argues that most of the increase in Irish house prices is justified by the economic and demographic driving forces. It should be remembered that in 1993 the average Irish house cost a mere € 75 000, which was extraordinarily low for a European country. Since then, remarkable growth in incomes, low interest rates, strong population growth, especially among the younger house-forming age groups, a surge in immigration and changing living patterns have all contributed to the boom.


Ireland’s house prices have risen dramatically since the mid 1990s. From 1995 to 2005 the price of second-hand houses more than tripled in real terms (Figure 1, left panel). House price inflation eased temporarily in 2001 but it has reignited since. Compared with other countries, the Irish housing boom has been extraordinarily vigorous: both in real and nominal terms the increase in house prices since the mid 1990s has been the highest in the OECD, with the United Kingdom and Spain ranking second and third respectively.

More favourable demand factors in comparison with developments elsewhere have surely played a role in shaping the buoyant price developments in Ireland. Growth in real disposable income since the mid 1990s has been stronger than in any other industrial country and real interest rates were among the lowest (Figure 2). The decline in inflation has also contributed by front-loading mortgage repayments. Furthermore, demographic trends were particularly favourable to housing demand in the 1990s, including strong population growth, a sharp fall in household size from a high level, a rapid acceleration in the growth of population in the household formation cohort and sizeable net immigration. Other demographic developments include the increase in the number of double income households and higher divorce rates. Another factor is the number of baby boomers investing in the buy-to-let market because of increasing worries about inadequate pension provisions for retirement.

In addition, the tax treatment of housing in Ireland has been more favourable for home ownership than in most other EU countries (van den Noord, 2005). This is reflected in a low user cost of capital. The user cost for homeowners is analogous to the cost of rental accommodation for tenants. It includes the after-tax mortgage interest rate net of capital gains, the opportunity cost associated with equity financing (usually the after-tax deposit rate), property tax (if any) and depreciation. There have been extended periods when the user cost has been negative, in particular in the late-1970s and from the mid 1990s onwards, implying a strong incentive to invest in housing. The main driving factor keeping the user cost negative has been the untaxed capital gains (on owner-occupied homes), whereas the importance of income tax deductions has diminished with the gradual decline in marginal income tax rates and a series of other tax reforms (Box 1). Since taxation of capital gains has an important negative influence on the user cost, its absence could have acted as a catalyst for the upward spiral in house prices.

Finland, Portugal and Spain are the only other countries which, like Ireland, give a tax deduction for mortgage interest payments but do not tax imputed rent or capital gains on the principal owner-occupied dwelling. However, all three have municipal taxes on property values ranging from 0.4% to 1%. The size of the tax bias in Ireland has been reduced over time as the ceiling on mortgage interest deductibility has not kept pace with the increase in house prices. Updating the estimates by van den Noord (2005) shows an overall tax wedge of –0.57% for the first seven years and –0.36% thereafter, giving Ireland the fifth-largest tax bias in the EU15.

The rise in housing demand triggered a strong response in supply, which again is unprecedented by international standards (Figure 3). House construction and residential permits per capita are among the highest in the OECD. Around a third of the housing stock is younger than ten years old. Half of the stock is detached houses, with apartments accounting for just 6%. The enormous increase in housing supply was accompanied by significant increases in real construction costs and land prices. The significant cost increases did not deter the supply of housing, which was aided by more relaxed zoning rules. Yet, despite the massive increase in the housing stock, it will almost certainly increase further in the medium term (even ignoring the effect of population growth) given that in Ireland there are significantly more adults per dwelling than in other OECD countries. If preferences in Ireland were similar to those in other EU countries, this would, ceteris paribus, lead to falling numbers of (adult) persons per dwelling. This gap has undoubtedly been a factor in the buoyant demand for housing and a driving force behind the escalation of house prices, and is likely to act for several more years. Indeed, the high cost of accommodation in Ireland may be discouraging people from forming an independent household (Fitz Gerald, 2005).

The demographic variable (the share of the population that is around the household-formation age) is included to capture the hypothesis that a younger population is likely to put extra pressure on the housing market.

It is difficult to compare prices across countries because the size, quality, location and amenities of houses can differ substantially. Comparisons are a little easier if they are restricted to the major cities, but this does not solve the problem entirely. Bearing this in mind, the available evidence suggests that average prices in Dublin are higher than in comparable cities. In a comparison of average sale prices in 2004 across a dozen European cities, the price per square metre was higher in Dublin that everywhere else (Figure 7, left panel). Some further evidence comes from cost-of-living comparisons conducted by various private-sector consultancies. These usually focus on prices or rents of inner-city apartments typically bought or rented by business executives. Here Dublin does not stand out so dramatically (Figure 7, right panel). This may be because rents are not especially high in Ireland but it may also reflect urban sprawl. Anecdotally at least, there is not a great deal of diversity in the housing stock. The centres of the main cities have not been taken over by apartment complexes and there is relatively little high-density in-fill housing. If preferences change and Irish people become more comfortable living in downtown apartments or in higher-density housing with no garden, then the distribution of prices may become more uneven: house prices in the central city may rise significantly relative to prices in the suburbs and city fringes. There is some evidence this may be happening already (Policy Exchange, 2005).

In a majority of countries, the ratios of prices to rents and prices to disposable income do not have strong trends when considered over long periods of time. The ratios may rise sharply during housing booms, but they usually fall back again through a combination of falling real house prices (i.e. a lower numerator) and rising rents or incomes (the denominator rising to catch up). In Ireland’s case, the increase in these two ratios far outstrips the cycles that have been seen in other countries before the most recent global housing boom (Figure 9), although the increase in the price-to-income ratio is in line with some other countries that have also enjoyed booming house prices in the last five years.

Affordability
The concept of housing “affordability” is popular in public discussions and with the real estate industry, perhaps because of its simplicity. While it is not particularly useful for assessing house price over-valuation, it is a useful measure of cash flow pressures. In 2005, the average mortgage repayment burden for a first time buyer was estimated to be 30% of disposable income (Central Bank, 2005), which is higher than in 1994/95, but is actually slightly lower than it was in 1991, when interest rates were much higher (Figure 10). Thus, the repayment burden is not out of line with past levels – provided, of course, that interest rates remain low.


The effects of increased housing wealth and equity withdrawal on household saving have never been strong in Ireland. The savings rate has been fluctuating around 9% throughout the housing boom. However, this does not imply that no housing equity is released, but rather that it may be recycled back into the housing market. This shows up especially in the buy-to-let market and in the rapid growth in the number of secondary or otherwise mostly vacant homes. This suggests that demand is driven, at least in part, by expectations of capital gains, which may confirm the impression of over valuation emerging from some of the quantitative indicators.

The buy-to-let market is small but has been growing fast. New buy-to-let mortgages constituted 20% of all mortgage transactions in 2004 while 30% of second-hand dwellings sold during the first half of 2004 were previously held as investment properties. The buy-to-let market is dominated by small, mostly inexperienced investors, whose primary objective is to provide for retirement. With property investors taking such an active part in the market, the question is to what extent they have driven up house prices. Attracted by the substantial capital gains and small carrying costs, many investors have entered the buy-to-let market, possibly displacing first time buyers and contributing significantly to housing demand and house prices. The main concern – and another indication of overshooting prices – is the growing divergence between property prices and rental income. Indeed, rents actually fell from 2002 to early 2005. The position of those in the buy-to-let segment of the market will continue to be sustainable only if interest rates stay low. However, if mortgage rates were to rise many of these investment positions would be loss making.

Demand for second homes appears to be another important factor in the housing market. Although housing supply has risen tremendously in recent years, a surprisingly large proportion of it appears to be satisfying demand for second-home properties (in 2005, around 15% of homeowners aged 35-54 owned a second home). As in the case of the buy-to-let market, some properties may have been acquired with the expectation that house prices would continue to grow at a fast pace for the indefinite future. More generally, an important element of the boom over the last decade has been the growth in the number of dwellings that are vacant, for whatever reasons, for most of the year. Fitz Gerald et al. (2003) calculated that the number of vacant dwellings in Ireland had increased by 80 000 from 2000 to 2003, which is equivalent to half the houses constructed over that period. On the basis of modelling work in that paper it was estimated that this additional demand would have added between 15 and 20% to house prices over the same period, which roughly corresponds to the overvaluation estimated in the econometric model above.


An over-valued housing market may have implications for financial stability, but that depends on many factors. The first point to note is that an overvaluation does not imply that prices will drop, at least if the degree of overvaluation is moderate. The housing market is unlike other asset markets in that house price dynamics are not symmetric. Prices rise quickly during booms, but in a market slump most people prefer to take their house off the market rather than sell at a loss. Hence, a small fall in prices followed by several years of a flat market is more likely than a sharp drop in house values. Put another way, the price level may remain fairly high as the market waits for the underlying fundamentals to catch up.

Even if house prices level off, there is a potential macroeconomic and financial stability issue that could arise from decline in residential construction. As noted in Chapter 1, the rate of house building will need to fall to some extent to return to its sustainable long-run level. International experience shows that this process is seldom smooth: when the investment rate turns down, it usually falls sharply (Box 3).

Definition of "booms"

Between 1960 and 2004, 49 residential construction booms have occurred in 23 countries for which data is available. A boom is defined (rather generously) as a rise in the level of real per capita residential investment of at least 15% over a five year period. In order to avoid identifying false peaks and data blips, a peak is defined as the highest point in a window of the preceding four years and the subsequent three years. By construction, the latest peak that can be identified is 2002; the analysis therefore omits the housing booms that are currently underway. In the cycles that have been identified, the average increase in real per capita residential investment from trough to peak is around 40%. The largest occurred in Korea from 1973 to 1978 (where investment rose by 160%). The trough to peak increase has exceeded 50% in 16 cases.

How common are soft landings? If a soft landing is defined as a relatively small reduction in the investment rate, they are not especially common. There have been only four cases where the decline in per capita residential investment has been smaller than one third of the increase that occurred during the boom years (these are the Netherlands after 1978, Belgium after 1990, the United Kingdom after 1998 and Finland after 2000). Soft landings are more common if they are defined as gradual declines, i.e. where it takes at least three years to hit the trough. There have been around 20 examples of these. But all of these were comparatively deep declines. If a soft landing is defined as something that is both mild and gradual, there has not been a single case out of the 49 boom bust cycles.