Texte de mon allocution lors de l'American dream coalition annual conference à Seattle, le 18 Avril 2009 (en fait, j'ai dû raccourcir un peu en séance, mais l'idée est là...)
-------------
LAND USE REGULATIONS,
Housing unaffordability and other socially undesirable impacts in France
First, I would like to thank the American Dream Coalition for this invitation (…)
I’m here because two years ago, I released for a French Think Tank, the Turgot Institute, a full study reviewing the housing public policies in France. As the study was deeply appreciated, it became a book.
One of the purposes of the study was to understand why we experienced such an unprecedented housing bubble. When the study was ordered, we thought that our study would mainly emphasize monetary phenomenon, or issues relative to credit practices, just like the one which exist in the USA. We expected that land use regulations would be one factor among others. It appeared to be, by far, the main factor explaining our bubble.
CAUSES OF THE REAL ESTATE BUBBLE IN FRANCE
Did we have a real estate Bubble in France, like in some parts of the USA ? The answer is obviously: Yes. The median multiple (using net disposable income) of real estate deals on the territory began to diverge from its historical mean of 2.9 (which was observed, within a 10% fluctuation range, between 1965 and 2000), at the beginning of the new millennium, and reached a whopping range of 4.6 to 5.0 in 2007, according to various sources (The following graph uses ratios relative to gross disposable income so figures are a bit lower, from 2.5 to 4.2) .
Gross median multiple ratio – 1936 to 2008
source ADEF
This bubble was especially remarkable because it was the first time it reached such intensity, but also because for the first time, it was not an “urban bubble” or a “Paris Bubble”. Prices of existing or new homes peaked in the same proportions either in central metro areas or in rural cities as far as 30 miles from these metro areas. They peaked in many metro areas with 50 000 inhabitants and more, not only in very big dense areas like Paris or Lyon or Azure Coast.
This graph shows that house prices increased 90% faster than households’ revenues between 1997 and 2007. In non inflation-adjusted prices, it corresponds to a whopping 140% gross increase.
index of median home prices / median net household revenue, 1965 = 1
Source: ADEF
Most publications explain this only by an increase in demand favoured by low credit rates. Of course, this explanation is half true, since average fixed rates to households were divided by a two fold between 1997 and 2005 (from 8.6 to 4.3%). This fall of credit rates allowed borrowers to lengthen the maturity of their loans, and, at equal revenue, a typical household could borrow about 50 to 70% more than it could before. Many analysts stop their reasoning here. “That’s demand, period”.
Every good economist should react to this by saying that if the supply was not constrained, an increase in household’s solvability would result in an increase of housing supply, since early signals of price hike would bring newcomers on the market. Especially, housing market has very low barriers to entry, isn’t prone to the formation of monopolies… This didn’t happen. We can see that, despite price increase begun in 1999, the supply began to soar only in 2004, and remained far under historical records (553 000 homes in 1973).
Supply didn't follow demand
between 1999 and 2004
Source: Bulle-immobiliere.org
How can we be sure that most part of the housing bubble comes from land use regulations?
First, if the gross price of homes and flats grew by 140% in the ten years of the bubble, construction costs only grew by 30% in the same time.
Orange and green curves: condo and house prices
purple curve: construction costs
Source: Bulle-immobiliere.org
It’s about 10% more than inflation. It can’t explain a 100% price hike on every home, new or ancient. Construction costs grew at a much slower rate than CPI between 1965 and 1997, the current trend inversion comes mainly from some regulatory strengthening in public housing, and some workforce shortages in some particular housing jobs. Not something really relevant compared to the global price increase.
But when I studied the share of land into the price of new homes, I discovered that raw developable land prices increased 6 times in the same period, though the costs of technical land development (sewage, electricity and other networks, roads and so on) only grew in the same proportions than construction costs. In fact, in most areas, land prices were the same in Euros in 2007, than they were in our good old Francs in 1997 (1€ # 6.5 Francs). On coastal areas, the prices increased from 10 to 12 times. Only in some very isolated and economically very depressed rural areas, we didn’t see such prices increases.
Agricultural land didn’t follow the same trend. Even at more than 15 miles from big metro areas’ fringes, developable land is as much as 200 times more expensive than agricultural one. At the fringe of big cities, the factor can jump up to 500 or 1000 times.
The latest point that convinced me that we were experiencing a “land oriented bubble”, was the discovery that in most housing programs with multiple lots, small lots were sold at much higher rates, in Euros per square meters, than bigger one. This shows the presence of a “regulatory penalty”, as first described by US economists Glaeser and Gyourko. This penalty corresponds to the price of a theoretical zero square foot lot on an operation.
We don’t have, in France, national databases with free access that could have allowed me to compute precisely this regulatory penalty. But some examples coming from real operations show that this penalty, in rural towns neighbouring (20-30 km) big cities, could represent about one third of the total cost of the home in 2006, much more than in 1998 or 2000. And, sometimes, it may be more.
Do we have a tough land use regulation in France?
The answer is, once again, yes.
The current frame of our regulation comes from a 1967 law, which allowed cities to promulgate zoning plans. That doesn’t say that there was nothing before: there were rules that “protected”, i.e. prevented from being built, agricultural land. But the 67 law marked the beginning of the application of zoning laws in France.
You will say: how a regulation promoted in 1967 could produce some real estate price frenzy only since 199...