Spain’s housing boom was supported by rapid economic expansion, strong employment growth, an immigration boom, and low real interest rates. With the abrupt drying up of funding since mid-2007, these factors have eroded quickly. Through 2010, employment and value added in construction are projected to halve as peak housing starts are completed. The authorities have launched efforts to help limit foreclosures and to activate the underdeveloped rental market. In the medium term, housing market cyclicality could be reduced by fading out generous home ownership incentives.

Abstract

Spain’s housing boom was supported by rapid economic expansion, strong employment growth, an immigration boom, and low real interest rates. With the abrupt drying up of funding since mid-2007, these factors have eroded quickly. Through 2010, employment and value added in construction are projected to halve as peak housing starts are completed. The authorities have launched efforts to help limit foreclosures and to activate the underdeveloped rental market. In the medium term, housing market cyclicality could be reduced by fading out generous home ownership incentives.

I. Developments in the Spanish Housing Sector1

A. Summary

1. Improving fundamentals created a decade-long housing demand boom. Spain changed significantly with democratization in the late-1970s and EU integration in the mid-1980s. In the 1990s, incorporation into EMU further improved confidence and lowered interest rates. Ample access to global liquidity spurred corporate investment and employment, and crowded in females and the unemployed. Increasing household incomes, lower interest costs, and longer mortgages increased housing affordability. Jobs in construction and services attracted large immigrant flows while Spain's baby boomers increased household formation. Thus, a number of positives for housing demand aligned, driving Spain’s housing boom.

2. Supply reacted forcefully but with delay to rising prices. House price appreciation was comparable to that in other countries experiencing housing expansions. The notable difference has been the accompanying major construction upswing in Spain—thus both prices and volumes boomed. However, gestation periods for new supply are long. Land laws provide incentives for municipalities to keep approved land scarce,2 building permit processing is lengthy, then it takes 3 months to start construction, and another 18–24 months to finish. Various authors (Garcia-Montalvo 2007, Callau and Pac 2008) suggest that speculative demand kept prices high despite the supply response later in the cycle.

3. Now fundamentals are deteriorating while inventory is accumulating. The tightening of funding has stopped construction in its tracks. Slowing growth is moreover reducing immigration, household formation, employment, and household income. With the long gestation lags, inventory is accumulating while interest rates first rose after the funding dry-up in mid-2007 just as indebtedness was peaking before easing again in late 2008 with the decline in the euribor interest rate index. Most mortgages are at variable interest rates.

4. The adjustment could run deep. Many young people are in fixed-term jobs and hesitant to commit to home ownership under current conditions. The rental market is underdeveloped so houses stay vacant longer and, as noted, the demographics could be reversing. With high and growing inventory, house price undershooting may occur, creating its own adverse dynamics. Indeed, the strong dependence on construction for employment is now ricocheting into rapidly growing unemployment, and the boom is turning to bust.

5. Absorbing inventory into the rental market is one option to limit undershooting. With inventories driving short-run house values, their absorption is paramount to stabilize prices. Given price expectations to the downside, fostering development of the small rental market holds potential, because it can tap into large unsatisfied demand from credit-constrained clients. Other useful measures aim at forestalling foreclosures, but increasing construction of subsidized housing is likely counterproductive. In the medium run, policies should aim at preempting boom-bust cycles. Cutting red tape could make supply more elastic, while curbing (fiscal) distortions favoring home ownership could result in a more efficient and transparent market.

B. The Boom

6. The Run-up to EMU entry (1995–2000) bolstered nominal stability and confidence (Figure 1). Short term interest rates were over 10 percent at the beginning of the 1990s. After devaluations of the peseta through 1995, increasing confidence and EMU entry cut nominal rates by over one-half. Real rates decreased even more as Spain did not eliminate its inflation differential with euro partners. Lower rates made many corporate investment projects attractive, resulting in high capital formation and job growth. Fiscal retrenchment to meet the Maastricht criteria crowded in the private sector.

Figure 1.
Figure 1.

Spain: Run-up to EMU and Convergence

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: Bank of Spain; Spain Ministry of Housing; and IMF staff calculations.

7. New-found prosperity led to a desire to upgrade housing. Investment and employment creation reduced unemployment by 9 percentage points. The dynamic economy increased female participation and household incomes. Households already servicing mortgages benefited from lower interest rates because 98 percent of mortgages are indexed to the 12-month Euribor. Consumers prioritized upgrades of lower quality (older) housing.3

Spain: Key Indicators, 1995-2000

(average annual percent change)

article image
Sources: IMF World Economic Outlook; and Eurostat.

8. Spain’s baby boomers moved out to set up their own households (Figure 2). The baby boom in Spain happened about 10 years later than elsewhere in the EU. Thus, many baby boomers reached household formation age in the 1990s. Furthermore, strong declines in youth unemployment from 1995 made it viable to move out, and household formation took off toward smaller households.4 Housing starts increased after 1997. Still, the fraction of young people (18–35) living with their families remained high at 63 percent in 2002.

Figure 2.
Figure 2.

Spain: Demographic Developments

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: INE; Ministerio de Vivienda; IMF, World Economic Outlook; and US Census Bureau International Database.

9. Housing prices first recovered from their 1996 trough. Real house prices fell by 20 percent in the early-1990s recession. They recovered in the later half of the decade but did not surpass their previous peak until 2002.5

10. A jump in immigration after 2000 accelerated housing demand. Spain’s upswing attracted many immigrants from Latin America, Eastern Europe and Northern Africa. Immigration’s contribution to population growth increased from 0.3 percentage points in 1999 to over 1.2 percentage points after 2002. Additionally, increases in income were partly related to immigration, which accounted for 20–25 percent of gains in GDP per capita.6

11. The supply response was slowed by zoning regulations, reinforcing house price increases. House price inflation in Spain deviated significantly from construction and land costs. This is related partly to costly land use regulations.7 Land approved for building saw average price increases of 30 percent in 2000–01, while agricultural land increased only 5 percent.8 Application processes for building permits are lengthy.9 Furthermore, Spain’s land law entitles local governments to 5–15 percent of rezoned sites (for roads etc.). Until 2007 this provided municipalities incentives to keep prices high to benefit from sales of excess land later on (OECD, 2007).10 Bureaucracy, segmentation and uncertainty induced by zoning processes aggravate scarcity of developable land further.11 Thus, relatively tardy supply translated the sizable demand shock into a doubling of real housing prices between 1999–2007.

12. Supply is also subject to long building times (Figure 3). Average time between building permit and house completion is around 2 years. Such delays can cause large swings in house prices, in both directions. On a structural basis, Ayuso and Restoy (2006) estimate that 2004 prices exceeded long-run equilibrium values by 24–32 percent. However, prices were only marginally overvalued compared to their short-term equilibrium, which takes supply rigidities into account. At the current juncture, supply sluggishness implies peak housing starts of 2006/07 reach completion in the recessionary period 2008/09. Thus, inventories will keep increasing for some time. This is exerting downward pressure on prices and transacted volumes, because price expectations are now to the downside.

Figure 3.
Figure 3.

Spain: Supply Rigidities, House Prices and Debt Dynamics

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: Thomson Datastream; OECD; Bank of Spain; Spain Ministry of Housing; Ministry of Finance BDSICE database; INE, Garcia-Vaquero and Martinez (2005); and IMF staff calculations.1/ The economic profit share is defined as unity minus the labor compensation share.2/ In years of an average household’s annual income.3/ The user cost assumes perfect foresight of households regarding house price changes as in Hilbers et al (2008). Fiscal reductions relating to mortgage debt service are considered in the calculation using the results of Garcia-Vaquero and Martinez (2005).4/ The real interests rate applicable to mortgages is calculated as the Euribor xb (MIBOR before 1999) plus 50 basis points minus HICP inflation.

13. The absence of a rental market exacerbates house price swings because it eliminates a cushioning reservoir of home use. Spain’s rental market is underdeveloped containing only 12 percent of residential properties or one-third of the OECD average.12 Incentives work against renting in supply and demand. In supply, slow court proceedings for eviction, and inflexible lease contracts with initial durations of 5 years discourage landlords (Matea, 2006).13 In demand, generous income tax deductions for mortgage payments and low real interest rates lower the user cost of house ownership.14 With the rental market underdeveloped, swings in housing demand quickly translate into house prices.

14. Fiscal deductibility of mortgage payments likely amplified house price increases. Fiscal incentives to home ownership tend to fail relieving homebuyers as sellers fix home prices at households’ payment capacity with the fiscal deduction taken into account (Garcia-Montalvo, 2007). The difference between house prices and land/construction costs—and thus the income-tax deduction—is then captured by construction companies, landowners and municipal governments. Strongly rising profit shares of construction companies and buoyant revenues for subnational governments during the boom reflect this.15

15. Housing affordability was gradually eroded. Over the last decade price-to-income and price-to-rent ratios of Spanish housing increased substantially. This was counteracted by lower financing costs, brought about by currency union. The user cost of owning a house, which subtracts expected capital gains on the property from the net financing cost, resulted even lower (and negative in many years) due to expectations for high price increases. This environment may have led to significant speculative demand in the later stages of the housing cycle. However, increased price-to-income ratios implied that average homebuyers needed to take on more leverage.

16. As a result, households became highly indebted at variable interest rates and longer maturities. Households’ willingness to take on debt rose for two reasons. First, fast per-capita income convergence elevated households’ perceived permanent income. A desire for consumption smoothing then helps explain higher indebtedness (Bank of Spain, 2006). Second, lower interest costs could be attributed to joining monetary union, and hence were seen as permanent. With the Euribor hovering around 2 percent and a persistent positive inflation differential of ½-1 percentage points with euro partners, real mortgage interest rates were around zero during the peak boom years. Nominal stability allowed lengthening of mortgage durations from 10–15 years in the late-1980s to 25–30 years in the 2000s. With collateral effects working in their favor, households took on mortgage liabilities to acquire higher-priced homes. They thereby increased their indebtedness from 70 to 130 percent of disposable income between 2000–07.16 This leverage has left households vunerable to increases in the mortgage index (12-month Euribor plus spread); this vulnerability is further exacerbated by longer mortgage terms. The Bank of Spain (2008a) estimates that a one higher-priced houses.17 They thereby increased their indebtedness from 70 to 130 percent of percentage point increase in interest rates leads to a loss of 0.7 percent of household disposable income.18 During most of 2008, therefore, household finances got increasingly squeezed, with relief only coming through Euribor declines at the end of the year. However, the recession has lowered expectations of future incomes and keeps housing demand low.

17. Households are cutting consumption to increase saving, yet defaults are rising. Spanish households are stretched: the amount of “free” household savings (not required for debt amortization) has steadily fallen from 8 percent of disposable income in 1996 to under 2 percent since 2000 and even negative territory in 2006.19 Given personal liability for mortgages,20 households are cutting back on consumption. However, most households (66 percent) in Spain own their main residence outright. Thus, the mortgage burden is unevenly distributed, and will likely prove excessive for those households that acquired homes recently at high loan-to-value (LTV) ratios, or that become unemployed.

18. Adjustment will feed through the real economy as the large construction sector needs to shrink. The size of the construction sector has become unsustainable at 13 percent of employment and 9 percent of GDP in residential investment.21 Thus, more so than in other countries, in Spain the negative consequences of the bust will be felt through unemployment as construction sheds workers. Other sectors will feel the knock-on effects. This will weaken housing demand fundamentals. Moreover, with roughly 25 percent of construction workers being foreigners,22 immigration has already started to slow.

C. The Correction

19. How will current dynamics settle down into a longer-run sustainable housing market equilibrium? To answer this question, we analyze demand, supply, inventories, and house prices. To gauge high uncertainty of population and consequently housing demand developments we set out three scenarios: low, high, and central.

Demand

20. Population growth may have peaked. Population growth surged from 0.2–0.3 percent during most of the nineties driven by native fertility, to 1.6 percent during the last five years owing to immigration. Immigration is now slowing. The 1.4 percent population growth in the first half of 2008 is already below the most pessimistic short term projections of the National Statistics Institute (INE).

21. INE’s long-run projections anticipate annual population growth to drop to 0.4–0.8 percent (Figure 4, Table 1). A low population growth scenario postulates that immigration will slow to 100,000 persons a year, resulting in population growth of 0.4 percent. A high scenario allows for a somewhat higher influx to yield 0.8 percent population growth, still well below recent peaks. A central scenario averages these two.23

Figure 4.
Figure 4.

Spain: Demographic Housing Need

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: Instituto Nacional de Estadística; Ministerio de Vivienda; and IMF staff projections.1/ Actual value for 2008 is the annualized growth rate of 1.4 percent during the first half of the year2/ Our projections assume a fall of the growth rate for the year 2008 as a whole to 1.2 percent in light of rapidly deteriorating economic conditions.3/ Value for 2008 is computed using growth rates of households and population available data through November 2008. The High scenario assumes a rate of change as in period 2000-07; Central scenario assumes a rate of change as in period 1987-2007; Low scenario assumes a rate of change as in period 1991-95.4/ Scenarios are calculated by combining assumptions of previous scenarios on population growth and persons per household. Actual data point for 2008 is upto third quarter.5/ In addition to previous assumptions, the high scenario assumes that 25 percent of home construction can be sold as vacation homes. The low and central scenarios assume 10 and 20 percent respectively.
Table 1.

Spain: Demographic Housing Needs, 2007-2015

article image
Sources: Instituto Nacional de Estadística; and IMF staff projections.

Actuals for 2008 are based on growth rates of data available through November 2008.

Numbers in bold are IMF staff projections to establish a transition to long-term INE growth projections.

Assumes a -1.3 percent annual change in persons per household, the sameas the average from 2000-07.

Assumes a -1.2 percent annual change in persons per household, the sameas in the sample average 1987-2007.

Assumes a -1.0 percent annual change in persons per household, the sameas the average in the last housing downturn 1991-95.

These scenarios are calculated by combining assumptions of respective scenarios on population growth and persons per household from above.

Assumes that the historical average of 25 percent of residential construction can be supported by demand for second homes.

Assumes that 20 percent of residential construction can be supported by demand for second homes.

Assumes that only 10 percent of residential construction can be supported by demand for second homes.

22. Alongside population, household formation is set to slow drastically. From a peak of 530,000 in 2006, we expect household formation to stabilize in a range of 240,000–390,000 per year. To obtain estimates for household formation, we combine the population projections with assumptions on the number of persons per household. The latter has followed a steady decline towards the European average. The high scenario envisions the number of persons per household to fall as rapidly as during the boom years 2000–07. The low scenario decreases changes in household size to the pace experienced in the last economic slump in 1991–95. The central scenario again describes the middle ground and translates to 310,000 households being formed. The range for household formation is between 240,000 (low) and 390,000 (high), thus ¼ to ½ less than the 530,000 new households recorded in 2006.

23. Vacation homes increase the sustainable long-run level of housing slightly beyond household formations. Since the late-1980s, it has been the case that housing starts and completions outstripped new household formation owing to vacation homes. Vacation (and empty) homes make up one fourth of all units built, with some 40 percent sold to foreigners.24 The high scenario assumes that 25 percent of output can be vacation homes, even during the downturn. The low scenario assumes this share to be 10 percent and central scenario 20 percent. Under these assumptions, the 310,000 household formations in the central scenario translate to almost 400,000 units in long-run sustainable housing demand. The range spanned by the low and high scenarios is 260,000–520,000.

Supply

24. Buoyant housing starts of 2006–07, substantially exceeded sustainable levels and are now adding to inventories. Property developers and construction companies were slow to react to first signs of slowing in 2006, starting 760,000 in that year and 620,000 in 2007. The implications are twofold. First, inventories will keep growing for some time even as Spain proceeds through the downturn. Second, owing to the long and relatively stable lag times, the adjustment path can be traced reasonably well (Figure 5). However, most recent data hint at actual completions falling short of those expected based on lagged starts. Hard-pressed developers may increasingly choose to suspend projects midway.

Figure 5.
Figure 5.

Spain: Building Permits, Housing Starts and Completions

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: Ministery of Finance BDSICE database; and IMF staff calculations.

25. Housing starts have halved since 2007 with permits foreshadowing a further drop. Housing starts in 2008Q3 have slowed to a pace of less than 300,000 per year—comparable to 1996 levels. Since then, building permits have dropped off further—to 1992–93 levels. In 1992–93, 200,000 housing units were started. Given increased tensions in financial markets in the fourth quarter of 2008, we expect some further deterioration going forward. Housing starts are expected to bottom out at 150,000 per year in 2009 and 2010 before gradually recovering to demographically sustainable levels by 2015.

26. Lengthy completion times suggest that home finishes will exceed demographically sustainable demand through mid-2010 (Figure 6). Houses started at end-2006 were completed in 2008H2, at a pace of 760,000 a year. Given that starts after the peak slowed only gradually, finishes are expected to remain above demographically sustainable levels until the middle of 2010, thus adding to inventory. This new inventory, in turn, is expected to weigh down home starts and prices.

Figure 6.
Figure 6.

Spain: Projected Scenario for Housing Starts, Completions and Value at Work

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: Ministery of Finance BDSICE database; and IMF staff calculations and projections.1/ Normalized to a construction time of one year to make the numbers of units under construction directly comparable to startsand finishes. True number of units under construction is 21/12 times as high, as it takes 21 months on average to complete a residential construction project.

27. Value-added in housing construction is projected to bottom out in mid-2010—at 1/3 of 2007 levels. Value at work (national accounts concept for value added) for residential construction can be proxied via a perpetual inventory method by adding new housing starts and subtracting housing completions.25 The results show that adjustment in the construction sector is well underway.26 Nevertheless, the largest adjustment is still to come in 2009 as housing completions pass their peak. Value at work will stop falling only in mid-2010, when the now low levels of starts will have fully fed through. Then, on an annualized basis, about 230,000 housing units will be under construction—compared to 730,000 in 2007.

28. Consequently, housing investment and construction employment will more than halve from their 2007 peaks (Figure 7). The housing sector’s value at work measure approximates closely both housing investment as well as housing sector employment (Table 3).27 At their projected trough in 2010, both will have returned to pre-boom levels last registered in 1997–98.

Figure 7.
Figure 7.

Projected Scenario for Housing Investment and Employment

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: Instituto Nacional de Estadística; Ministry of Finance BDSICE database; EU Klems; and IMF staff calculation and projections.1/ The housing sector is defined as residential construction and real estate service activities. Employment is national accounts based. Real residential and total construction investment are used to construct an estimate of residential construction investment. The ratio of employment in real estate services relative to total market services from EU Klems is employed to yield an estimate of employment in real estate activities.
Table 2.

Construction Sector Dynamics

(Thousands, unless otherwise indicated)

article image
Sources: INE, Ministry of Finance BDSICE database, EU Klems and IMF staff calculations and projections.

January through November values used in projection for 2008 are actuals or projections based directly on forward-looking building permit data. Throughout 2009 and 2010 housing starts are assumed to mimic the pattern observed in 1993. Thereafter starts increase by 14 percent per year.

Projections through mid-2010 are based on already observed housing starts.

These numbers are normalized to a construction time of one year to make them comparable with starts and finishes. True number of units under construction is 21/12 times as high, as it takes 21 months on average to complete a residential construction project.

Actual housing starts and completions are used in the computation of units under construction. Computation of the number of units under construction uses a perpetual inventory method starting at end-1983.

Calculated using a demographically sustainable absorption level of 390-400 thousand units/year, as set out in the central scenario of Table 1

The housing sector is defined as residential construction and real estate service activities. Employment is national accounts based and includes both salaried employees and self-employed. Real residential and total construction investment are used to construct an estimate of residential construction investment. The ratio of employment in real estate services relative to total market services from EU Klems is employed to yield an estimate of employment in real estate activities.

Table 3.

Construction Activity: Regression Results

article image
Source: IMF staff estimates.Notes: Standard errors in parentheses. *** indicates significance at the 1 percent level.

Inventory

29. Housing inventory is difficult to estimate, particularly given the high incidence of second and empty homes. Availability of housing statistics in Spain generally has tended to lag the sector’s importance in the economy. This is evident in the continued absence of official inventory statistics. Therefore, inventories have to be estimated. These estimates utilize data on housing starts, finishes, sales, household formation and the 10-year census of the housing stock. The last census shows that in 2001, 15 percent of the housing stock sat empty, while another 16 percent were second homes.28 With vacation homes constituting about 25 percent of residential construction, they complicate the relationship between household formation and residential construction and thereby render the determination of inventories difficult. Nevertheless, we may explore a few approaches.

Spain: Housing stock

article image
Source: INE Housing Census 1991 and 2001.

30. Since 2001, house completions have exceeded household formation by one million units. Callau and Pac (2007) use this figure as their inventory estimate. They argue that possible upward bias from speculative demand is counterbalanced by potential downward bias from some new households having likely been formed in pre-existing homes.29

31. Also, home completions have substantially outstripped new home sales. Since the inception of sales data in 2004, a cumulative difference of 1.3 million unsold units has built up. However, this figure constitutes an upper bound for inventories owing to two upward biases. First, completions also include homes not intended for sale, such as those build by individuals for their own use or by the public sector for rental purposes. Second, it is not taken into account that some homes are demolished or become uninhabitable. Appropriate adjustments for these biases put inventory buildup between 1997 and 2008 in a 0.8–1.4 million range (BBVA, 2008b).

uA01fig1

Difference between housing completions and household formation

(thou. units)

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: INE; Ministry of Finance BDSICE database; and IMF staff calculations.

Spain: New home sales and completions

(Thousands)

article image
Sources: Ministry of Finance BDSICE database.

Data through third quarter.

32. Taken together, analysts tend to estimate inventory at around one million homes. For instance, Tinsa, a property appraiser with a market share of 20 percent, sees inventory reaching 930,000 by end-2008 (Tinsa, 2008a). Garcia-Montalvo’s (2007) estimate is 1.3 million. A study by the Ministry of Housing put inventory at 500,000 in June 2008 with a rise to 650,000–930,000 units expected by end-2008.30 For our staff scenarios, we use a “consensus” inventory estimate of 1 million. Inventory estimates are crucial to the analysis because inventories predominantly drive short-run housing dynamics, while demographics play a minor role (Klyuev, 2008).

House Prices

33. Real house prices started to decline at end-2007 and have been falling steadily since then (Table 4). The various available house price series differ in timing and extent of the house price downturn (Box 1). Assessed prices tend to lag the market: For instance, Ministry of Housing statistics show only a 5½ percent annual decline in 2008–Q4. Tinsa, however, already reported a decrease of 10 percent with a sharper drop after the financial turmoil started in August 2007. A new transaction price series developed by INE appears closer to the Tinsa results. Looking forward, further declines are likely. Asking prices have dropped more strongly throughout 2008, foreshadowing future declines as these properties get sold. Overall, present developments confirm that prices are playing a crucial role in the adjustment process alongside lower sales volumes.

Table 4.

Spain: Recent House Price Developments

(Year-on-year percent change)

article image
Sources: Ministry of Housing; INE; Tinsa; and Fotocasa.

34. The house price correction in the current downturn is likely deeper than in 1991–96, when interest rates declined significantly. In the past, the average European housing downturn lasted almost six years and resulted in an average real house price correction of 29 percent—varying between 16–49 percent (Figure 8). In comparison, Spain’s last house price correction of 20 percent between 1991–96 was mild. This was partly due to rapidly falling real interest rates in 1993 sustaining housing affordability and prices. Moreover, data suggest that Spain’s last adjustment was not as strongly inventory driven as the current one; house finishes in 1992–94 were low compared to household formation.

Data on House Prices in Spain

House price series for Spain can be subdivided into assessed, market, and asking prices:

A. Assessed Prices

Ministry of Housing

Free market house prices:

  • Quarterly data of prices per square meter submitted by the association of property appraisers

  • For properties valued at less than 1.05 million euros

  • Two types of weights have been used to aggregate prices across geographic areas with surprising differences for the period 1999–2002 (see Box Figure)

    • ○ Population (1987–2004)

    • ○ Number of assessments (1995-present)

  • Subseries are available that distinguish between houses above and below two years of age

  • Shorter time series are available for prices of subsidized housing

General index of house prices (IGP):

  • Weighted average of free and subsidized house prices available since 2005

  • Otherwise same characteristics as price of free housing

Tinsa

  • Developed by property appraiser TINSA based on its own appraisals

  • Monthly chained LaSpeyres index of housing prices in the free market

B. Market Prices

INE

Index of Housing prices (IPV)

  • Chain-linked LaSpeyers index based on actual sales prices in the free housing market submitted by the National Association of Notaries.

  • Available from 2007Q1 with breakdown into new and used housing.

  • Prices are per dwelling. Every dwelling is assigned a category. The overall index is a weighted average of the categories.

C. Asking prices

Fotocasa

  • Monthly asking prices for used properties starting in 2005, constructed based on advertisements on the internet real estate portal with the most visitors and largest home database in Spain

uA01bxfigI1
Sources: Ministry of Housing, Ministry of Finance BDSICE database, Tinsa and Fotocasa.
Figure 8.
Figure 8.

Housing Slumps in International and Historical Perspective

Citation: IMF Staff Country Reports 2009, 129; 10.5089/9781451812299.002.A001

Sources: Bank of International Settlements; IMF, International Financial Statistics; Bank of Spain; Hilberset al. (2008); and IMF staff calculations and projections.1/ Interest payment on a new typical mortgage, defined as an 80 percent loan-to-value mortgage on a 149,000 euro home.2/ Staff projections assume a front-loaded 30 percent decrease in real house prices over the next fouryears.

Spain: House Price Booms and Busts

article image
Sources: Bank for International Settlements; and Hilbers et al. (2008).

35. In the current scenario, affordability is likely to be restored via real price declines—end-user interest costs are seen as playing less of a role. While existing mortgagees are profiting from recent decreases in the 12-month Euribor, evidence suggests that Spanish banks are tightening both availability and risk spreads for new mortgages (250 bp at end-2008 versus 110 bp on average since 2000). This leaves house prices to carry out adjustment.31 Experts indicate that a 20–30 percent correction of nominal prices will be needed. A scenario calculated by staff suggest that a real house price correction of 30 percent—in line with past European experiences—would improve mortgage affordability for new buyers roughly to levels of 2005. Regional data confirm the role of prices in reestablishing affordability: declines are most pronounced on the coasts and large cities, i.e., in locales that experienced the highest appreciation.

uA01fig2
Sources: Bank of Spain and IMF staff calculations.1/ The spread is calculated using the average Euribor over the previous two months (see Bank of Spain, 2007a).

36. If the recent boom reflects largely speculative demand, then the correction may be deeper. Many authors (Garcia-Montalvo 2007, Callau and Pac 2008) point to the importance of speculative demand in fuelling the housing boom given abundant financing, low yields on alternative assets, and expectations of future price increases. Surveys show that during the peak boom years more than 90 percent of homebuyers thought homes to be overvalued, yet also expected house prices to keep increasing at rates of 20 percent per year.32 Sluggishness of supply made it commonplace for buyers to buy an apartment in a building yet to be finished, and often yet to be started.33 Capital gains on a deposit in such a transaction could reach 800 percent, thus dwarfing transaction costs.34 Econometric studies largely agree on overvaluation of Spanish house prices relative to fundamentals in a range of 15–30 percent.35 With increased carry costs and tightening credit conditions, the price adjustment could thus still be deeper than implied by deteriorating fundamentals alone.

37. Given high outstanding inventory, the correction is likely front-loaded. The staff scenario assumes an 18 percent real price decline by end-2009. High inventory and tight credit are expected to make house prices the main adjustment variable, especially as medium-term housing fundamentals are deteriorating with slowing immigration and household formation. Further, rising unemployment tends to affect crucial first time buyers the most.

D. Policy Conclusions

38. Reducing inventory is key to limit undershooting of house prices. Klyuev (2008) finds that inventory-to-sales ratios and foreclosure starts tend to be the main drivers of housing market corrections. The gap between actual and equilibrium prices, however, does not exert a powerful influence over price dynamics in the short run. Thus, even if current Spanish house prices were not far above their equilibrium values, large inventory implies substantial risk for undershooting. Attempts to artificially sustain house prices above fundamental levels can be costly (Glaeser and Gyourko, 2008). At the same time, house price undershooting can also be a problem, given harmful effects of forced sales on the financial system and wider economy. To this end, rapid absorption of inventories is paramount.

39. Expansion of subsidized housing construction may be counterproductive. The authorities are planning to double subsidized housing (VPO) construction, partly as a means to support construction activity (Box 2). This seems unlikely, given that VPOs are a small share of residential construction (15 percent). Instead, more VPOs may prolong the adjustment. At worst, resulting inventory increases may depress market prices and/or construction further. Moreover, as social housing is allocated below market prices, excess demand is resolved through queues with detrimental effects on labor mobility, where Spain scores poorly (ECB 2003; Bank of Spain 2007b).36 The authorities are partly trying to address the issue by orienting 40 percent of social housing development towards rentals. However, social housing for purchase remains high—with undesired redistributive effects.37 In present circumstances, low income households’ increased difficulty in securing home financing further limits the effectiveness of social housing sales. Housing needs of disadvantaged groups are an urgent problem that may be better addressed through the private rental market, e.g., through means-tested vouchers (below).38 Laid-off construction workers could be aided directly in a targeted and timely manner (e.g., in public projects, re-training) to facilitate reorientation of the economy.

Recent Housing Market Policies

A. Measures related to subsidized housing (VPO):

Reducing inventory

  • Property developers are now allowed to convert free market housing into highest tier VPOs (vivienda concertada) at any time until end-2009, while before a home had to be on the market for more than one year.

  • Official Credit Institute (ICO) 3 bn euro/year credit line for developers which turn finished homes into rentals. Loans can be rolled over up to seven years, while units remain rentals.

  • 8 percent increase in income ceiling to reduce inventories converted into highest tier VPOs

  • Increase of maximum sale price for highest tier VPOs

  • New rent with option to buy scheme for highest tier VPOs for rental periods up to 10 years

  • Eligibility expanded for direct subsidies of rental payments and down payments

  • Autonomous communities will establish registries of potential VPO buyers to assist banks in assessing credit worthiness

Supporting construction activity

  • Double construction of subsidized housing units to 150,000/year over the next 10 years, of which at least 40 percent will be rented

  • Buy land from property developers for up to 300 million euros with aim of constructing 20,000 VPOs (until April 2009)

  • New requirement that 25 percent of new land developments must be used for VPOs

  • New subsidies for construction or refurbishing of VPOs

  • ICO guarantees for securitizing mortgage loans for VPOs (5 bn euros in each 2009 and 2010)

Non-VPO measures to support construction activity include a euro 2 bn ICO credit line for residential energy upgrading (“Plan RENOVE”) starting in 2009, and a VAT reduction for home renovations to 7 percent—the same as that applied to new construction. Also now only a qualified majority (instead of unanimity) is needed in home owner associations to implement energy efficient improvements.

B. Measures aimed at homeowners:

Forestalling foreclosures

  • Fee-free extension of mortgage terms

  • Allowing unemployed workers with dependents and mortgages of less than 170,000 euros to capitalize 50 percent of monthly mortgage payments during 2009–10 to the period 2011–20. Up to 500,000 persons are expected to qualify. ICO will guarantee delayed payments with a credit line of up to 6 bn euros, of which 3 bn are expected to be used.

  • Income tax deduction for mortgage interest to be considered in the calculation of income tax withholding upon request for persons earning less than 33,000 euros/year. Expected to result in 2 million euros of additional liquidity.

Other measures

  • Incentive of 1,500 euros/year for firms hiring unemployed persons with families to support

  • Income tax relief on deposits in dedicated savings accounts earmarked for home purchases extended from 4 to 6 years

  • Exempting capital gains from home sales if proceeds are reinvested in another residential property through end-2010 for those who have purchased a property but have been unable to sell their previous home.

C. Measures aimed at developing the rental market:

Supply measures

  • Subsidies for construction, acquisition and remodeling of rental units

  • Subsidies for purchase of insurance against damages to property and risk of non-payment

  • 50 percent of revenue from rental property exempted from income tax; exemption is 100 percent if tenant is young and meets certain income requirements

  • Since 2005, autonomous communities have created rental agencies to absorb risks typically borne by landlords in return for a share of the rent payment. The agencies’ market share is small.

  • Real Estate Investment Trusts (SCIMI in Spanish) to be introduced as vehicles to transform unused inventories into rental properties.

  • Evictions for non-payment to be eased

  • Arbitrage system for speedy rental conflict resolution to be implemented

  • Owner’s right to reclaim the property extended to when a first degree relative is in need

Demand measures

  • Means-tested cash benefit of 210 euros/month for 22–30 year olds moving into rental housing

  • Tax deductibility of rent payments reintroduced (after abolition in 1999 tax reform) in order to offset tax advantages of home purchase. This benefit is subject to a 28,000 euro annual income ceiling and received by 700,000 households.

  • Direct grants to needy tenants of up to 2880 euros per year

40. Limiting foreclosures will slow inventory expansion. Both the authorities and the financial sector have moved to limit foreclosures. Some mortgagees were allowed to renegotiate their mortgage terms at no fee. A voluntary program allows unemployed heads of families to defer half of mortgage payments through end-2010 with official guarantees backing deferred amounts.39 Banks are repossessing and then renting back property to former homebuyers as well as accepting properties from developers to manage loan impairments.40 The authorities are now permitting the setting up of real estate investment trusts (SCIMIs), which should absorb properties from developers’ books and turning them into rentals.

41. A private rental market could prove to be the most important tool to absorbing inventory. Spain’s private rental market only accounts for 6 percent of housing, much less than in peer countries.41 On the supply side, long leases with initial durations of 5 years and landlord-unfriendly legislation are largely to blame, although some progress is being made.42 Demand is discouraged by fiscal incentives to ownership (below). Expansion of the small rental market holds the most potential for inventory absorption, because potential buyers expect further price drops and are putting off purchases.

42. Rentals could tap into unrealized demand of the young and immigrants. A substantial share (60 percent) of persons aged 18–35 still live with their parents. Relative to other countries, this share has further increased as housing affordability deteriorated.43 Together with immigrants living in above-average size households, the young hold the most potential for housing demand. However, these groups are also the most immediately affected by credit constraints and unemployment because most have temporary job contracts.44 Home appraisals for immigrants have fallen by ¾ between 2007H1 and 2008Q3 (Tinsa, 2008b). Harmonization of labor contracts could increase household formation as income streams become more certain than under temporary contracts (Bank of Spain, 2008b).

43. In the medium run, measures to foster the rental market should include a gradual fading out of fiscal incentives to homeownership. Housing policy in Spain has mainly aimed at ownership, making Spain the industrialized country with the highest (85 percent)—and continually rising—home ownership rate.45 Numerous taxes, deductions and subsidies are aimed at the housing market, resulting in a high fiscal cost (approximately 1 percent of GDP).46 Income tax liabilities are reduced by 15 percent of mortgage expenses.47 As the deduction is calculated on nominal payments, it becomes more powerful in a low real interest rate environment with above average inflation, inducing more volatility into prices (van den Noord, 2005). Dominguez-Martinez (2004) estimates that due to the income tax deduction, persons can afford to pay 15–20 percent more,48 most of which inelastic supply transfers into higher prices. Higher prices imply a redistribution from younger to older people, in addition to the regressive effect of the non-means tested deduction itself. Reduced labor mobility and diversion of resources away from productive investment are negative implications for Spain’s reorientation towards a new growth model. Policies to foster the rental market have aimed at offsetting the impact of income tax relief, e.g., by also exempting rental property from income tax and inducing further regressiveness. Distortions could be reduced by gradually fading out fiscal incentives to home ownership as the housing market stabilizes.

44. A more elastic supply response will reduce price fluctuations. Large housing demand shocks get amplified into prices due to inelastic supply. A more agile supply response can be reached by reevaluating municipalities’ incentives to supply buildable land and by reducing the length and cost of planning and building permit issuance processes. This should keep house prices close to construction and land costs, even if interest rate fluctuations induce sizable demand shifts.

References

  • Asociacion Hipotecaria Espanola (2007), “Estadisticas del Mercado Hipotecario, Ano 2007”, http://www.ahe.es, accessed on October 28, 2008.

    • Search Google Scholar
    • Export Citation
  • Ayuso and Restoy (2003), “House Prices and Rents: An Equilibrium Asset Pricing Approach”, Bank of Spain Working Paper, number 0304, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • Ayuso and Restoy (2003), “House Prices and Rents in Spain: Does the Discount Factor matter?”, Bank of Spain Working Paper, number 0609, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • Bank of Spain (2006), “Informe annual 2006”, Madrid, Spain.

  • Bank of Spain (2007a), “Informe annual 2007”, Madrid, Spain.

  • Bank of Spain (2007b), “Boletin economico May 2007”, Madrid, Spain.

  • Bank of Spain (2008a), “Boletin economico March 2008”, Madrid, Spain.

  • Bank of Spain (2008b), “Boletin economico April 2008”, Madrid, Spain.

  • BBVA Research (2008a), “Real Estate Watch Spain”, BBVA Economic Research Department, July 2008.

  • BBVA Research (2008b), “Real Estate Watch Spain”, BBVA Economic Research Department, December 2008.

  • Bover, O. and J.F. Jimeno (2007), “House Prices and Employment Reallocation: International Experience”, Bank of Spain Working Paper, number 0705, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • Brueckner, J. K., (2007), “Government Land-Use Interventions: An Economic Analysis,” 4th Urban Research Symposium, World Bank, Washington, D.C.

    • Search Google Scholar
    • Export Citation
  • Callau, J.B. and R.D. Pac (2008), “Sobre el peso del sector de la construccion en la economia espanola”, Boletin Economico de Informacion Comercial Espanola, 2944, 21–31 de Julio de 2008, pp. 3145.

    • Search Google Scholar
    • Export Citation
  • Cemex (2008), “Global Economic and Construction Outlook”, CEMEX Economic Analysis, July 2008.

  • Credito-Vivienda.com (2008), “SIMA 2008 - El mercado español de segunda vivienda”, http://www.credito-vivienda.com/sima-2008-el-mercado-espanol-de-segunda-vivienda.html, accessed November 10, 2008.

    • Search Google Scholar
    • Export Citation
  • De los Llanos, M. (2006), “Las Medidas de la Politica de Vivienda en Materia de Alquiler y Vivienda Protegida”, Boletin Economico, Bank of Spain, Madrid, Julio/Agosto 2006.

    • Search Google Scholar
    • Export Citation
  • DiPasquale, D. and Wheaton, W. (1994) “Housing Market Dynamics and the Future of Housing Prices”, Journal of Urban Economics, 35(1), pp. 127.

    • Search Google Scholar
    • Export Citation
  • Dominguez Martinez, J.M. (2004), “Fiscalidad y coste de endeudamiento de la inversion en vivienda en Espana”, Cuadernos de Informacion Economica” No. 180, May-June 2004, Fundacion de las Cajas de Ahorros Confederadas, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • ECB (2003), “Structural Factors in the EU Housing Markets”, European Central Bank, Frankfurt, Germany.

  • Eicher, T. (2008), “House Prices and Land Use Regulations: A Study of 250 major U.S. cities”, Northwest Journal of Business and Economics, forthcoming.

    • Search Google Scholar
    • Export Citation
  • El Pais (2008), “El numero de casas nuevas sin vender sube mas del 20% en solo medio ano”, El Pais November 12, 2008, p. 20.

  • El Mundo (2008), “Bancos y cajas encarecen las nuevas hipotecas”, El Mundo November 6, 2008, http://www.elmundo.es/mundodinero/2008/11/06/economia/1225961479.html, accessed November 19, 2008.

    • Search Google Scholar
    • Export Citation
  • Expansion (2008), “Corredor: El ‘stock’ de viviendas debe absorberse antes de dos anos”, published October 25, 2008, http://www.expansion.com, accessed October 25, 2008.

    • Search Google Scholar
    • Export Citation
  • Garcia-Montalvo, J. (2006), “Deconstruyendo la burbuja inmobiliaria”, Papeles de Economia Espanola, 110.

  • Garcia-Montalvo, J. (2007), “Algunas consideraciones sobre el problema de la vivienda en Espana”, Papeles de Economia Espanola, forthcoming.

    • Search Google Scholar
    • Export Citation
  • Garcia-Vaquero, V. and J. Martinez (2005), “Fiscalidad de la Vivienda en Espana”, Bank of Spain Working Paper No. 0506, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • Gimeno, R. and C. Martinez-Carrascal (2006), “The Interaction between House Prices and Loans for House Purchase. The Spanish Case”, Bank of Spain Working Paper, number 0605, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • Girouard, N., Kennedy, M., van den Noord, P. and C. Andre (2006), “Recent House Price Developments: The Role of Fundamentals”, OECD Economics Department Working Paper, No. 475.

    • Search Google Scholar
    • Export Citation
  • Glaeser, E. L. and J. Gyourko, (2002). “Zoning’s Steep Price,” Regulation. 25(3) pp: 2431.

  • Glaeser, Edward L. and Gyourko, Joseph (2008), “The Case against Housing Price Supports,” The Economists’ Voice, Vol. 5, Iss. 6, Article 3. http://www.bepress.com/ev/vol5/iss6/art3

    • Search Google Scholar
    • Export Citation
  • Glaeser, E. L., J. Gyourko and R. Saks, (2005). “Why Is Manhattan So Expensive? Regulation and the Rise in Housing Prices,” Journal of Law and Economics 48, (2), pp. 33169.

    • Search Google Scholar
    • Export Citation
  • Hilbers, P., Hoffmaister, A., Banerji, A. and H. Shi, (2008). “House Price Developments in Europe: A Comparison”, IMF Working Paper, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Klyuev, V. (2008), “What Goes Up Must Come Down? House Price Dynamics in the United States”, IMF Working Paper, number 08/187, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Lopez Garcia, M. A. (2003), “Politicas de vivienda: eficiencia y equidad”, Sector Publico y Eficiencia”, Papeles de Economia Espanola, No. 95, April 2003, Fundacion de las Cajas de Ahorros, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • Martinez Pages, J. and L. Angel Maza (2003), “Analisis del Precio de Vivienda en Espana”, Bank of Spain Working Paper, number 0307, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • Martinez Pages, J. and M. de los Llanos Matea Rosa (2003), “Precios de la vivienda en Espana: evolucion y factores explicativos”, Perspectivas del Sistema Financiero, 78, Fundacion de las Cajas de Ahorros, Madrid, September 2003.

    • Search Google Scholar
    • Export Citation
  • Ministerio del Medio Ambiente (2008), “Boletin Mensual de Estadistica, Junio 2008”.

  • Ministerio de Vivienda (2008), “Encuesta sobre la compra de vivienda principal en el periodo 2002–2006”, Madrid.

  • Nieto, F. (2007), “The determinants of household credit in Spain”, Bank of Spain Working Paper, number 0716, Madrid, Spain.

  • OECD (2005), “Economic Surveys: Spain”, Volume 2005/6, April 2005.

  • OECD (2007), “Economic Surveys: Spain”, Volume 2007/1, January 2007.

  • OECD (2008), “2008 Economic Review - Spain”, unpublished manuscript.

  • Oficina Económica del Presidente (2006), “Inmigración y economía española: 1996–2006”, La Moncloa, Madrid, Spain.

  • Oficina Economica del Presidente del Gobierno (2008), “Informe Economico del Presidente del Gobierno”, La Moncloa, Madrid.

  • Pellicer, L. (2008), “Caixa Catalunya arrienda casas impagadas a sus antiguos proprietarios”, El Pais, October 27, 2008, Madrid, Spain.

    • Search Google Scholar
    • Export Citation
  • Sosvilla, S. (2008), “Immigration and Housing Prices in Spain”, Series Economia Regional, Catedra FEDEA—Caja Madrid, October.

  • Tinsa (2008a), “Rueda de Prensa 14 de Octubre de 2008”, www.tinsa.es, accessed October 17, 2008.

  • Tinsa (2008b), “Notas de Coyuntura Inmobiliaria – Mercado Residencial, Tercero Trimestre 2008”, www.tinsa.es, accessed October 17, 2008.

    • Search Google Scholar
    • Export Citation
  • Tribunal de la Defensa de la Competencia (1993), “Remedios politicos que pueden favorecer la libre competencia en los servicios y atajar el dano causado por los monopolios”, Madrid.

    • Search Google Scholar
    • Export Citation
  • Tribunal de la Defensa de la Competencia (1995), “La Competencia en Espana: Balance y Nuevas Propuestas”, Madrid.

  • Van den Noord, P. (2005), “Tax Incentives and House Price Volatility in the Euro Area: Theory and Evidence”, Economie Internationale, CEPII Research Center, issue 1Q, pp. 2945.

    • Search Google Scholar
    • Export Citation
1

Prepared by Christian Henn.

2

See OECD (2007, p. 79).

3

According to the 2001 census, 58 percent of properties were constructed before 1980.

4

Household size in the mid-1990s was well above the EU average of 2.5 persons. See Figure 1

5

Ayuso and Restoy (2003) judge that house price increases in the second half of the 1990s mainly constituted a correction of previous undershooting.

9

In particular, planning of electricity and water infrastructure is complex and lengthy at 7-10 years (OECD, 2007).

10

From 2007 on legislative changes obligated municipalities to use this percentage of land exclusively for utility provision.

11

Tribunal de Defensa de la Competencia (1993, p. 149 and 1995, p. 37).

12

In 2005. See OECD (2007).

13

Landlords are obligated to renew leases annually during the first five years. Rent payments are adjusted by the CPI.

14

Income tax relief is available for both principal and interest payments as well as other items such as taxes, and other permit and licensing costs. The general deduction rate is 15 percent and up to around 9000 euros may be applied to the deduction annually. The deduction also applies for deposits into dedicated savings accounts for home purchase. See e.g., OECD (2007).

16

Indebtedness also became more widespread. As of 2006, more than 40 percent of persons had pending debt in Spain, compared to just 10 percent in 1990 (Bank of Spain, 2006).

17

Nieto (2007) identifies higher household wealth—largely due to higher house prices—to be the single most important determinant of credit expansion to households in Spain. There has thus been a strong reverse causality through collateral effects facilitating twin booms in credit and housing prices; this financial accelerator effect is now expected to reverse and exacerbate the downwards adjustment. Gimeno and Martinez-Carrascal (2006) also find that the housing and credit booms in Spain were strongly interrelated. In a VAR model, they estimate that a one percent increase in credit growth is associated with a 0.15 percent rise in house prices. Likewise, a one percent house price increase translates into a 0.1 percentage point higher growth rate in mortgage credit.

18

This income effect of interest rates was still positive in the early 1990s. Then, households on average profited from higher rates, because they owned less real estate and more interest-bearing assets. In the current environment, competition among banks for customer deposit may lower the mentioned 0.7pp average effect on households in the short run. It is likely, however, that these benefits accrue mainly to those with higher financial wealth and thus less at risk of losing their homes.

20

Mortgages are collateralized by the property and income of the mortgagee.

21

In 2006, more housing units were under construction in Spain than in Germany, France, Italy and the UK combined. Bover and Jimeno (2007) explain the distinct reactions of construction activity to house price increases amongst countries with remaining building possibilities, using population density and persons living in free-standing houses as proxies. Their result is that in countries with few spatial building constraints, including Spain, relative employment in construction increases by roughly 0.5 percent for every 1 percentage point increase in real house prices. In dense countries, such as the UK, construction activity hardly reacts.

23

INE’s long run population projections from 2002 initially underestimated population growth. The staff scenarios construct a transition from this higher population growth to the INE long-term projections (Table 1).

24

This is also confirmed by anecdotal evidence; see e.g., Credito-Vivienda.com (2008).

25

We choose the average of housing starts and finishes in 1983 as the starting point for the perpetual inventory method. Between 1980-83 the amount of housing starts and finishes in each year were very similar and exhibited little fluctuation. In addition, the long time frame between 1983 and the present will minimize the impact of the starting value on our results.

27

As housing sector employment we define employment in the residential construction and real estate services sectors. Notes under Table 2 explain the construction of this series. Employment in real estate services is only 10-12 percent ofthat in residential construction, but has more than doubled in the last 10 years.

28

Little change since the 1991 census hints at unfavorable structural factors -such as landlord-unfriendly rental laws – forestalling a more efficient use of the housing stock.

29

Callau and Pac’s calculations implicitly assume that all newly formed households moved into new homes with none moving to homes that existed before 2001.

31

See e.g., El Mundo (2008). Furthermore, Garcia-Montalvo (2007) finds that financial variables were almost exclusive drivers of the most recent housing cycle, while demographic variables were paramount in previous cycles.

33

More than half of new homes were bought before they were finished during 2002-06 (Ministerio de Vivienda, 2008).

35

Specified levels of overvaluation were estimated for different years. See Ayuso and Restoy (2006), Bank of Spain (2004), Girourard et al. (2006), IMF (2008) and Sosvilla (2008).

36

Subsidies make interest rates on loans (up to 80 percent of the purchase price) attractive. Additionally, there is a subsidy of up to 11,000 euros to cover the down payment.

37

OECD (2005) points out that approvals for purchase of a social housing unit are based on current rather than permanent income. Thus, households remain in social housing despite increases in income, thereby reducing availability for families in need.

38

OECD (2005) reports that rented subsidized housing only made up 6 percent of the housing stock and only covered 35 percent of poor households (European averages are of 14 percent and a coverage of 73 percent).

39

Deferred payments are to be repaid over 10 years.

41

The total rental stock is around 12 percent of housing, but half of this is accounted for by social housing (OECD, 2001 and 2005).

42

See Box 2. Also, OECD (2008) confirms that plans to speed up conflict resolution have been partially implemented with 6 out of 10 planned new courts in high eviction areas already in service.

44

High incidence of temporary contracts causes the young and immigrants to be laid off first in downturns. In the year to 2008-Q2 employees aged 35 and under lost 246,000 jobs, while older workers gained 320,000 (Tinsa, 2008b).

45

In 2001 (OECD, 2005).

46

In 2004 (OECD, 2007).

47

Other important measures making homeownership fiscally attractive are no taxation of imputed rents for owner-occupiers, a reduced VAT rate for residential construction and favorable treatment of capital gains. See OECD (2007) for more details on fiscal treatment of housing in Spain.

48

Lopez-Garcia (2003) obtains a similar result in the range of 15–30 percent.

Spain: Selected Issues
Author: International Monetary Fund