Nordic Margins
A 1930s city at night seen from a rooftop: still factory chimneys, idle harbour cranes and new apartment blocks under construction

Origin of a Depression: how housing overbuilding fed the 1929 slump

A Swedish housing planner looked at American construction statistics and saw a surplus of homes building up for five years before the crash. His article named the housing cycle as a cause economists had underrated.

Science and society 6 min read

The United States economy of 1929 looked unstoppable. Output was well above the previous year, unemployment stood around three percent, and the financial press spoke of a new era without serious slumps. Within four years a quarter of the labor force was out of work. Explanations of what happened between those dates fill a large shelf, and most of them start with the stock market crash of October 1929, the banking panics that followed, or the policies of the Federal Reserve.

In 1967 the American journal Land Economics published a different account by Carl O. Nordling, then working as an architect and planner in Stockholm. His article, "Origin of a Depression", argued that a decisive cause had been in plain view for years: the United States had built far more homes in the mid-1920s than its population could use.

A planner's yardstick

Nordling's day job was preparing housing programs for Swedish municipalities. The method is simple in principle. Over any period of years, the number of new dwellings a country can sensibly build is roughly the number of new households plus the number of worn-out homes demolished. Build much more than that and a surplus accumulates; prices and rents fall, foreclosures rise, and sooner or later construction has to drop sharply to let demand catch up.

He applied the same yardstick to the United States using the census bureau's historical statistics. After the First World War there was a genuine shortage of housing. The nonfarm population of about 75 million was growing by around two million a year, which implied a need for about half a million new dwellings annually before replacing old stock. In 1920 only about 250,000 nonfarm dwellings were started.

The boom of 1921–1925

Construction responded strongly. Nonfarm housing starts rose to about 450,000 in 1921 and kept climbing through 720,000 and 870,000 to 890,000 in 1924, peaking at roughly 940,000 in 1925. By Nordling's estimate the shortage had been made up by around 1923. A sustainable long-term rate, allowing for population growth and demolitions, was about 700,000 units a year. Each year above that added to a hidden surplus: nearly 200,000 extra dwellings in 1924, and more than that again in 1925. Rents and prices began to slip from about 1925.

He argued that the right moment to cut back was 1924. After that, any adjustment would require a drop of a third or more in one of the economy's largest industries, which no one at the time was prepared to contemplate. Demolitions could be brought forward for a while, masking the surplus, but that only reduced the stock of old houses available to replace later.

The slow slide to 1929

The decline after 1925 was gradual: about 850,000 starts in 1926, 810,000 in 1927 and 750,000 in 1928. Meanwhile population growth slowed. Immigration quotas introduced in the early 1920s cut new arrivals, and annual growth in the nonfarm population fell toward 1.5 million. Real demand for new dwellings, by Nordling's reckoning, dropped to around 600,000 a year, so even the reduced output kept adding to the surplus.

In 1929 the adjustment could no longer be postponed. Starts fell to about 510,000, a cut of a third in a single year, and to about 330,000 in 1930. Nordling estimated that the 1929 drop alone put between 1.5 and 2 percent of the nonfarm labor force out of work, roughly the year's entire normal growth in that labor force. The effects spread to brickworks, timber, cement, plumbing, trucks and tools, and through lost incomes to consumer spending. Uncertain households postponed buying cars and clothes, which pushed the slump into other industries.

Nordling did not claim that housing explained everything. He accepted that many factors determined the depth of the Depression. His point was that an overbuilt housing stock made a severe downturn very likely from about 1926 onward, whatever the stock market did, and that the cause was visible in data anyone could have read. He mentioned that Wassily Leontief, in a private letter, had suggested the effect could be measured precisely using an input–output table of the American economy.

How it compares with later explanations

Economists have offered several main explanations for the Great Depression, and the housing cycle sits among them, not at their center.

The most influential account since the 1960s is monetary. Milton Friedman and Anna Schwartz, in A Monetary History of the United States (1963), blamed the Federal Reserve for allowing the money supply to collapse during the banking panics of 1930–33. Later work by Ben Bernanke stressed the breakdown of credit when thousands of banks failed, and Barry Eichengreen and others showed how the international gold standard spread deflation from country to country and how leaving it helped recovery.

Housing does appear in these discussions, especially as a reason the economy was already weakening before the crash. Residential construction did peak in the mid-1920s and fall for several years before 1929, and several economic historians have treated that decline as one of the forces behind the downturn of 1929–30. Nordling's article is an early and clearly argued version of that view, framed in demographic terms by someone who had built housing forecasts for a living.

What his account cannot explain alone is why the recession of 1929–30 turned into the catastrophe of 1931–33. That stage is generally attributed to financial and monetary collapse, which a surplus of houses does not by itself produce.

A warning for planners

Nordling closed with a practical lesson. Modern statistics on population and housing make it possible to see a surplus building up and to slow construction in time. He believed Sweden in the mid-1960s was itself heading for overbuilding as a result of ambitious public planning, a point he had made the year before in a Swedish trade journal. The large Swedish housing program of 1965–74 did end with many vacant flats in the early 1970s, which lent his warning some weight in hindsight.

His other contributions to the sciences include the history of the multi-mutation theory of cancer, which grew from the same familiarity with statistical tables, and a more speculative piece on the Big Bang model. A fuller account of his planning career is on the biography page.

Questions and answers

What did Nordling argue in "Origin of a Depression"?

He argued that the United States built too many homes between 1922 and 1928, well beyond what population growth and demolitions required. The resulting surplus forced housing construction down sharply in 1929 and 1930, throwing many building workers out of work and spreading the downturn to supplying industries and consumer spending.

When and where was the article published?

It appeared in Land Economics, an American journal devoted to land use, housing and urban economics, in 1967. Nordling wrote it while working as an architect and planner at the Stockholm consultancy Vattenbyggnadsbyrån, and he based his figures on the US Census Bureau's historical statistics published in 1960.

Do economists accept the housing explanation?

Most treat the fall in residential construction after 1925 as one contributing factor to the downturn of 1929–30. The dominant explanations of the Depression's depth stress monetary contraction, banking failures and the gold standard. Housing is seen as part of why the economy was vulnerable, not as the single cause.

What happened to US housing construction in those years?

According to the figures Nordling used, nonfarm housing starts rose from about 250,000 in 1920 to a peak of about 940,000 in 1925, then declined slowly to about 750,000 in 1928 before falling to about 510,000 in 1929 and 330,000 in 1930. Construction remained very low through the early 1930s.

Continue the voyage

A moonlit South Pacific beach with tall palms, a thatched Samoan fale glowing from inside and an outrigger canoe on the sand
Science and society

Margaret Mead and the Samoa debate

Read
An empty Stockholm street corner on a winter night: old stone facades, a dark cinema canopy, steps down to the subway and a single street lamp in the snow
Science and society

The murder of Olof Palme

Read
An observatory dome open on a snowy hilltop at night, its telescope aimed at a spiral galaxy and faint rings of light in the sky
Science and society

The Big Bang: a brake block?

Read