In the years immediately following the First World War, economists in industry, government, and academia came to perceive data as the antidote to the interrelated crises afflicting American capitalism: the spasms of the business cycle and its associated recessions, a prolonged housing shortage, and a disorganized and strike-prone construction industry. Herbert Hoover’s Department of Commerce took the lead in centralizing private data series, fostering analysis of those statistics, and disseminating the conclusions to industry players. Construction data was the cornerstone of this push. The building industry, both a gargantuan segment of the American economy and its most backward element, was widely perceived as the engine behind the ups and downs of the business cycle. To tame the construction industry would be to tame American capitalism itself.

The major indices and monthly statistical series of housing and construction widely used today by economists, housing scholars, investors, and others in the building industry are largely a product of this cathartic moment. As I argue in a new paper, How to See the Housing Sector: Herbert Hoover’s Department of Commerce and the Origins of the Nation’s Housing Data Ecosystem, these statistics were assembled for specific political purposes, which have been largely forgotten by contemporary users. These overlooked political purposes give context to data sources now widely taken for granted by housing scholars, specifically, and by the larger pool of economists who use supply and construction data as an indicator of economic health.

In the context of the 1920s, data was expected to play a palliative role, allowing economists, policymakers, and builders to understand and anticipate market movements and thereby collectively tame an industry known for chaos. Regular, widely available statistics would guide the construction sector toward a state of stabilized production, ensuring full employment and an adequate supply of high-quality housing, without the need for heavy-handed public regulation. The role of the Commerce Department in this vision was akin to that of a weather service: not regulating private activity outright, but supplying private actors with the information they needed to manage their own behavior.

Statistical information, in the end, failed to serve this stabilizing role. The Roaring Twenties ended with a catastrophic real estate bubble fueled by the proliferation of mortgage-backed securities that left millions of Americans penniless and hindered the recovery of the financial sector.

It is easy to dismiss the excesses of 1920s real estate development as an episode of “irrational exuberance,” an inevitable spasm of speculative mania in a decade of paltry federal regulation. But this explanation misses the mark: in this period, construction, real estate, and housing production were monitored and analyzed – if not regulated – to an unprecedented degree of precision. The national real estate market had come into focus as a coherent economic object capable of being forecasted and managed.

The origin story of these federal statistics serves as a cautionary tale about the dangers of a blind trust in data and in the capacity of private actors to use information for socially desirable ends. This faith in numbers, as well as the corresponding faith that the real estate industry would use the data to regulate its own behavior, would not survive the 1920s. But the indices, statistical agencies, and analytic methods are still with us, and remain the means by which policymakers, economists, and regulators understand the nation’s housing supply and its relationship to the broader economy.