The Spanish housing bubble
Most commentators focus on house prices, usually in real terms, as the gauge the housing bubble and its own developments (e.g. Münchau 2012). Data claim that house prices have indeed adjusted, however, not enough. Figure 1 shows the home price index for Spain (measured as in accordance with rents) at roughly the same level as in 2003 but still much above its pre-2000 levels. We favour the price-to-rent index to the true price because it shouldn’t be suffering from immigration; any upsurge in demand for housing from that source should manifest itself in upwards pressure on both rents and prices. On the other hand, immigration should put more pressure on rents than on house prices since most immigrants will tend to be short of capital and therefore apt to be renting, instead of buying.
The chart confirms that house prices have followed the price-to-rent index since their peak within the last quarter of 2006 but remain greater than the pre-bubble period.
Figure 1 . House prices in Spain
Source : OECD
An analogous chart for Ireland would show a complete hump-shaped curve, with the existing price/rent ratio already back again to the amount of late 1990s. While Ireland’s price adjustment is near complete, you can argue that Spain continues to be no more than halfway.
Though this may be true, here we argue that for macroeconomic developments, house prices are less important compared to the amount of real resources found in the housing sector. What counts for the true economy may be the size of the construction sector (especially employment) and its own time path in accordance with the long-run equilibrium. Figure 2 shows spending in construction as a share of GDP for both countries with the largest bubbles in Europe: Ireland and Spain.
Figure 2 . Housing overhang in Spain and Ireland
Note: Investment in construction as share of GDP on the vertical axis, the red line may be the average over the time 1970-2000. Source: European Commission services (Ameco)
The figure may be used to compare the genesis of the housing overhang and its own absorption in both of these countries. In each chart, the region below the hump curve and above the long-run average of investment in construction (red line) represents our estimate of excess construction and therefore our estimate of the overhang (of housing and other fixed structures). For Ireland the cumulated more than construction between 1997 and 2008 is the same as about €99 billion, or 55 % of (2008) GDP. Regarding Spain this is a lot more than €380 billion or 37% of (2010) GDP.
It is argued that Spain differs because there is a solid demand by foreigners for vacation homes. While this may be true, that is already considered over time average as shown by an ‘equilibrium’ rate of construction spending (in accordance with GDP) 5% points higher in Spain than in Ireland and above most Europe.
What lengths along may be the adjustment process?
The region below the red line representing average long-run construction investment measures the procedure of absorption of the bubble. As houses usually do not decay rapidly, the only method to bring supply and demand back balance is because they build less.
In both countries, the procedure is actually incomplete in the sense that the region measuring the absorption is a lot smaller compared to the area measuring the overhang.
- In Ireland, the adjustment has occurred at high speed and about 1 / 3 of the bubble was already absorbed.
- In Spain, absorption has hardly started and the forecasts from the EU Commission for 2012 and 2013 (represented by the dotted line) claim that the adjustment will probably stagnate.
If construction were to keep at the still relatively higher rate of today, the procedure of absorption of the bubble would take a lot more than 30 years. Newer data, from different sources, on housing starts (and completions) indicate that the Commission forecasts for 2012 and 2013 might understate the fall in the construction activity, which is accelerating again. This may also be the primary reason (as opposed to the fiscal adjustment) why growth rates for Spain are being revised downwards.
The relative good performance of the Spanish economy in 2010-2011 might thus have already been because of the fact that of these years the adjustment in both government accounts and the housing sector slowed up. The long-term costs of the delay are actually becoming apparent.
Our very rough estimate of the construction overhang also informs the losses that the banking sector could be facing once adjustment is complete – at least from an order-of-magnitude perspective. Ultimately, the construction overhang represents the number of real resources wasted in a sector whose expenditure was financed mostly by credit. It’s understandable our estimated total of €380 billion exceeds by far the provisions and write downs accumulated by the Spanish bank operating system (and specifically the savings banks) up to now.
Figure 3 . Spain: Construction and current account deficit
The foundation of funding matters
A housing overhang by itself doesn’t have to result in an acute financial meltdown if it had been financed by domestic savings (like in Japan and Germany). Unfortunately this as false in Spain.
Figure 3 implies that investment expenditure in construction has closely tracked Spain’s accumulation of foreign debt (i.e the existing account).
- During the last decade Spain has accumulated a stock of foreign debt of near 90% of GDP.
- The Macroeconomic Imbalances Procedure scoreboard of the Commission implies that Spain’s foreign debt already equals Greece’s (EC 2012).
The foreign debt level is rising as the current account continues to be in deficit. On present trends, it could increase to about 100% of GDP by 2016 (about the amount of Portugal today). At this stage Spain would clearly maintain a threat of being take off from the marketplace.
What could possibly be done?
Figure 3 implies that the existing account adjustment, that was extremely swift in 2008-2009, slowed considerably when the adjustment in the construction sector slowed aswell. The main element short-run task for Spain is to avoid the accumulation of further foreign debt so the country no longer depends upon continuing inflows of foreign capital. This is often achieved quickly only when the construction sector (which probably wastes further resources) were permitted to shrink further.
Of course this may work only when the resources liberated by the shrinking construction sector are used elsewhere throughout the market; at this time the economy can grow only when exports grow aswell. This reallocation of labour towards the tradable sector is proceeding only very slowly and can need a fall in wages – at least in accordance with all of those other Eurozone, and thus specifically relative to Germany. This involves a labour market where wages can fall when there is a surplus supply. The Spanish record upon this account, however, is quite discouraging – as an be observed from the partnership between unemployment and wage growth (the Phillips curve) for Spain shown in Figure 4.
Figure 4 . Spain’s recent Phillips curve
It really is apparent that in Spain the Philips curve has deteriorated since 2007 (the rate of wage inflation was higher in 2008 for the same degree of unemployment as through the early 2000s). That is probably as a result of backwards wage indexation, which transmits the terms-of-trade shocks from higher oil prices to the labour market. Inspection of the chart seems to claim that an unemployment rate of over 20% is required to keep wage inflation near zero. In Germany, in comparison, wage inflation visited zero at an unemployment rate around 10%.
Overall, it would appear that Spain hasn’t yet fully adjusted to the collapse of its enormous housing bubble, which propelled its economy on an unsustainable path until 2008. House prices need to fall further, the construction sector must shrink further, and the reallocation of labour towards exportables is slowed up by a labour market that prevents wages from falling quickly enough.
European Comission (2012), “Alert Mechanism Report”, February 14.
Münchau, Wolfgang (2012), “There is absolutely no Spanish siesta for the eurozone”, Financial Times, 18 March.
 Large fiscal deficits will be the consequence of an enormous drop in tax revenues following the housing bubble burst and a substantial upsurge in spending for unemployment benefits.
 For additional information see Annex 2 in Gros (2007).
 Here we consider total construction, which include two main subsectors: dwellings and non-residential and civil engineering constructions. Between 1997 and 2006 how big is the dwelling as share of GDP has almost doubled, going from 6.7% to 12.5%, as the rest of constructions has increased from 7.3% to 9.7%.
 Our measure will probably understate how big is the overhang because through the boom GDP was above its sustainable level since it contained construction and other related activities at an unsustainable pace.
 Spain’s long-run average is greater than other European countries’ not merely for the entire construction, also for dwellings.
 A recently available study by a Spanish bank "Leaving stereotypes behind: Spain recovers competitiveness and productivity faster than any other eurozone economy” (BBVA Economic Watch 2012) implies that exports are indeed growing, however, not fast enough to keep unemployment from rising.