In Spain, private sector credit as a proportion of GDP practically doubled between 2000 and 2007. This increase occurred in parallel with a boom in house prices, which doubled in real terms during the same period. The economy as a whole also grew at an unprecedented rate.

But in 2008, the credit bubble in Spain burst, and the shock wave left a trail of bad loans, bank failures and a prolonged economic slowdown.

Something that attracted less attention in Spain was the evolution of the construction sector, where employment grew an astonishing 47%, compared to the 27% observed in the economy in general.


New data on credit booms

The episodes of rapid credit growth – known as “credit booms or booms” – present a crossroads between immediate economic dynamism and the danger of future crises. The risk of a “bad boom” —that is, when an episode of rapid credit growth is followed by a financial crisis or mediocre economic growth – is greater when it coincides with a boom in house prices.

According to our research, the experience of a dangerous combination of credit booms with rapid expansions in the construction sector has not been limited to the case of Spain, and is also observed in periods not related to the global financial crisis.

Our conclusion is that signals from the construction sector can help distinguish between dangerous booms, which must be controlled, and episodes of dynamic but healthy credit growth (“good booms”).


Credit Booms Don’t Help Everyone Equally

The predictive capacity of construction activity

During booms, output and employment expand faster. But not all sectors behave the same way. Most of the surplus growth is concentrated in a few sectors, specifically in construction and, with much less intensity, in finance.

But it is precisely the sectors that benefit the most during booms that experience the most drastic slowdowns during dips. This means that credit booms tend to leave few lasting traces on a country industrial makeup.


The special case of construction

credit loans

Construction is the only sector that always behaves differently if the credit boom is good or bad. On average, output and employment in the construction sector grow 2 to 3 percentage points more in bad booms than in good ones. In all other sectors, the difference is smaller and not significant (except in trade, but only in terms of output growth).

What is special about construction? One aspect is that construction does not contain the same growth potential that many other sectors present. In other words, excessive investment in construction can divert resources that could be used for more productive activities, which reduces output. Furthermore, the temporary stimulus that employment in the construction sector receives and the relatively low level of job skills that it requires may discourage some workers from investing in their training and skills development. This can leave lasting effects on the product once the boom is over.

Finally, construction projects have significant initial financing needs, and the final consumers of the product (for example, houses or hotels) also tend to borrow to finance their purchases. Thus, debt can increase significantly more during booms that are driven by construction.


The predictive capacity of construction activity

construction loans

An unusually rapid expansion of the construction sector can be a sign that the credit boom is bad. A 1 percentage point increase in output and employment growth in the construction sector during a boom increases the probability that such a boom will be problematic by 2 and 5 percentage points, respectively.

Construction growth is also more reliable than other variables in predicting the economic costs of bad booms. A 1 percentage point increase in output growth in the construction sector during a bad boom coincides with a decrease of almost 0.1 percentage points in aggregate output growth during the decline.



If they detect rapid expansion in the construction sector during a credit boom, authorities should consider adopting more restrictive macroeconomic policies and using macroprudential tools (such as higher down payments on mortgages).

In certain cases, policy measures will be determined by other indicators, such as house prices or household mortgages. However, these other indicators may not always sound the alarm (for example, if the construction boom is financed by the business sector or by foreigners), and the risks accumulate. So unusually rapid growth in construction could be the red flag for, for example, limiting banks’ exposure to real estate developers and other construction companies.

Finally, since output and employment data in the construction sector are often available with a few months lag, high-frequency indicators, such as applications for construction permits, could provide useful signals. Construction indicators should also be included in the models that are used to assess future economic activity.