Contemporary critiques of GDP’s role in policymaking see it as an ideological abstraction, emblematic of neoliberalism, that misrepresents “real” economic conditions. What these critiques often ignore, however, is the unique circumstances that led to its creation.
Editor’s note: The current debate in economics seems to lack a historical perspective. To try to address this deficiency, we decided to launch a Sunday column on ProMarket focusing on the historical dimension of economic ideas. You can read all of the pieces in the series here.
Indicators are vital to modern macroeconomic governance. Among these, Gross Domestic Product (GDP) is the most well-known. Once an obscure acronym, today it is ubiquitous in newspapers and primetime news shows, often touted as the ultimate measure of national economic wellbeing. A low GDP figure can trigger a market selloff and determine the fate of political leaders. It even affects whether nations are considered “developed” and investment worthy. As historian Philipp Lepenies put it, it is “the most powerful statistical figure in human history.”
In recent decades, however, GDP has become controversial. Critical social scientists, historians, and journalists argue it is an ideological abstraction, emblematic of neoliberalism, that misrepresents “real” economic conditions. Ironically, Simon Kuznets, instrumental in the invention of GDP’s predecessor, the Gross National Product (GNP), was the first to warn in his 1934 report to the US Senate that national income statistics measured not socio-economic welfare, but the productive and consumption capacity of a nation. As feminist economists, most notably Marilyn Waring and most recently Caroline Saunders and Paul Dalziel, recognize, they also exclude non-market activities such as housework. Most recently, Joseph Stiglitz, Amartya Sen, and Jean-Paul Fitoussi launched a devastating attack on environmental grounds, highlighting GDP’s indifference to the social and environmental harms of economic activity.
These critiques assume that GDP’s primary function is to measure economic growth. According to this perspective, GDP was invented by Keynesian economists in response to war mobilization needs during World War II and was operationalized during the postwar period to manage growth. This narrative, however, is functionalist and substitutes GDP’s eventual utility for the reasons for its creation.
In a piece entitled “Institutionalism in Action,” I problematize this narrative and show the origins of GNP lie in the efforts of institutionalist economists, Kuznets and his mentor Wesley Clair Mitchell, and their New Dealer allies Gerhald Colm, Gardiner Means, and Lauchlin Currie, to stabilize the business cycle in the 1920s and the 1930s. Thus, I argue that GNP’s invention was part of the effort to balance inter-sectoral imbalances to prevent economic depressions. This project did not just co-produce the concept of “the economy” and the macroeconomic state as a domain for governing this new entity. It also beget the idea of growth.
Grasping the persistence of indicators requires us to think of them not as isolated social constructions but as socio-technical artifacts intrinsically woven to the knowledge infrastructures that produce them. In the case of GDP in the US, this infrastructure is the National Income and Products Accounts (NIPA) System. Housed in the Commerce Department’s Bureau of Economic Analysis (BEA), NIPA is an ever-growing matrix of index numbers. As such, it allows the state to monitor the flow of goods and services, all the way from nature to consumers, at the level of monetary flows generated during production and consumption.
NIPA acts as an interface between the state and the market. It constructs “the economy,” to put it in political theorist Tim Mitchell’s words, as a composite statistical object, composed of many disparate, non-fungible parts that could not be otherwise patched into a coherent whole in the form of GDP. As a representational technology of economic reality, it forms what science studies scholar Paul Edwards calls a “closed world” around the policymaker. Consequently, the state sees reality in the form of formally defined aggregate magnitudes such as “demand” and “supply,” as opposed to the substantive qualities of everyday life. This is a critical accomplishment. Not only does it give a statistical soul to macroeconomics, rendering it more than mere theoretical speculation, but it also constitutes the position of the macroeconomic policymaker, occupied by a new type of actor responsible for managing NIPA’s statistical artifacts, most notably GDP.
The Invention of GNP
NIPA was not the first economic knowledge infrastructure. In the 1920s, various government offices were busy building their own knowledge infrastructures to stabilize the sectors under their jurisdictions. What set apart the Commerce Department’s Bureau of Foreign and Domestic Commerce (BFDC), BEA’s predecessor, was its ambitions to produce a totalizing statistical image of multiple sectors, containing a vast number of industries.
BFDC’s ambitions stemmed from its close ties to the National Bureau of Economic Research (NBER) and its business cycle stabilization project. A bastion of institutionalism, the NBER was established by Columbia University economist Wesley Clair Mitchell and Harvard Business School’s founding dean Edwin Gay. Mitchell theorized “the business cycle” in his 1913 magnum opus Business Cycles as a sui generis phenomenon, manifesting itself as cyclical booms and busts. Because economic units competed under liberalism, the capitalist system had to be coordinated by the price system.
This mechanism, however, resulted in gradual over-accumulation of inventories and capital. As misallocations reached unsustainable levels, companies had to liquidate inventories and capital stock. Since firms are interdependent for payments and interconnected through the price system, such disruptions triggered “deflationary spirals” that caused depressions. The NBER and its allies believed that while the business cycle was endemic to capitalism, rationalizing business management and the provision of information to companies could tame cyclical fluctuations.
The business cycle project was, paradoxically, a successful failure. The Great Depression was simply humiliating; as BFDC’s chief statistician admitted during the 1931 hearings on the depression, subscription rate to the agency’s publication disseminating business cycle statistics was only one-tenth of its expected level. And BFDC’s statistics were far from comprehensive: as one moved from raw materials to end products, serious statistical errors and gaps plagued the data.
Nonetheless, the business cycle project resulted in the conceptual innovations necessary for the invention of GNP. First and foremost, institutionalists began to conceive of “the economy” as an interdependent system that needed to be monitored by experts from a vantage point within the state. In their 1928 report to President Herbert Hoover, Mitchell and Gay underscored that “our economy” was a “shifting, dynamic complex,” a “living organism,” consisting of “chief component elements” such as consumption, productive sectors, labor, and national income. The growth of this object depended on the pace of flows between producing and consuming units.
Yet, growth, Mitchell and Gay wrote, also resulted in imbalances: “the rapidity and vigor of growth of some elements [was] so great as seriously to unbalance the whole organism.” Since “all parts of our economic structure . . . were . . . interdependent and easily affected,” a “technique of balance” for “the maintenance of equilibrium” was necessary. This required the creation of a state agency staffed with experts with “a general knowledge of the relations of the parts to each other.” To prevent depressions, these experts had to “apply the principle of equilibrium . . . in every economic relation” through “incessant observation and adjustment of our economy.”
National income statistics were developed in the 1930s for this purpose. In the 1920s, national income was marginal to the business cycle project. Mitchell’s theory conceptualized money as an overlay on material flows and therefore emphasized the over-accumulation of productive capacity and inventories over income. Faced with insurmountable measurement problems at the BFDC in the mid-1930s, Mitchell’s student Kuznets and his assistant, Robert Nathan, turned to this long-neglected component in their search for a reliable way to monitor imbalance.
Instead of chasing endless material flows, they simply calculated the income companies paid out to workers and owners. Kuznets called this “national income paid out” and then generated “national income produced” by adding profits. While the former concept (later renamed “Gross National Income”) measured economic welfare, the latter (now “Gross National Product”) gauged annual economic output. With the balance sheets used for calculating these indicators, policymakers could monitor macroeconomic constructs such as “consumption” and “investment.” Rather than pursuing messy and often controversial sectoral policies, they could correct imbalances by intervening in these constructs with fiscal tools that targeted monetary flows.
“GRASPING THE PERSISTENCE OF INDICATORS REQUIRES US TO THINK OF THEM NOT AS ISOLATED SOCIAL CONSTRUCTIONS BUT AS SOCIO-TECHNICAL ARTIFACTS INTRINSICALLY WOVEN TO THE KNOWLEDGE INFRASTRUCTURES THAT PRODUCE THEM.”
Operationalization of national income statistics as a macroeconomic interface required a final and yet controversial modification in Kuznets’s GNP: the inclusion of the state as an economic sector in national income estimates. Kuznets excluded the state in these calculations because he believed public services were an input for private production. Since businesses factored in the taxes they paid for these services in pricing their products, their inclusion would cause double counting. Yet, an emerging group of economists, calling themselves “Keynesians,” argued against Kuznets, pointing to the inconvenient fact that his justification did not account for a state that ran large deficits. Because they believed deficit spending was the only effective policy for economic recovery, policymakers needed a measure of the state’s contribution to economic activity to fine tune economic stimulus.
Led by future White House economic adviser Lauchlin Currie, then the most influential policymaker, structuralist economists Gerhald Colm and Gardiner Means convinced Nathan, who had replaced Kuznets at the BFDC, that including the state in GNP was necessary. This seeming victory, however, allowed Keynesianism’s critics to tie deficit spending to the infamous postwar inflation and thereby characterize the state as a bull in the china shop of the economy.
If GNP was invented for economic stabilization, then why did growth become the dominant policy paradigm? The answer lies in war mobilization. As the BFDC was building NIPA in the early 1940s, Kuznets and Nathan, now at the War Production Board, sought to determine how much the US could spend on the war. This task required calculating what Kuznets called “feasible maximum” production. The duo used the macroeconomic balancing approach Means developed in the late 1930s. This entailed first determining what level of national income would result in full employment and then reverse engineering how much production in each sector would produce such a level of income. The task of mobilization planning demanded projecting feasible maximum production into the future, and this required Kuznets and Nathan to estimate how much each economic component would grow over time and how much the state had to spend accordingly.
This new approach of stabilizing future GNP at its feasible maximum level was what Colm and others would call growth management in the postwar period. Indeed, after becoming the macroeconomist of Truman’s Council of Economic Advisors (CEA)—the agency established under the 1946 Employment Act as the institutional interface of the macroeconomic closed world built around NIPA—Colm would boast to his foreign counterparts that the US no longer needed to worry about imbalances, because it could simply grow its way out of depressions.
Given this twist, it is unsurprising many associate GDP with growth today. After all, Keynesians themselves actively promoted this association as they embraced the promises of unbalanced growth in the 1950s.
Since the global financial crisis of 2008, however, we are once again faced with the problem of inter-sectoral imbalances. As economist Thomas Philippon also recognized, the share of finance within US GDP reached a historic level before the crisis. Despite a temporary correction, NIPA estimates suggest that this imbalance has only grown since then. Back in 2010, Christina Romer, then Obama’s CEA chair, made a similar observation about health care spending: Testifying on the Affordable Care Act (ACA) in 2009, Romer warned that health care had reached one fifth of US GDP and was destined to reach 34 percent by 2040. Such an imbalance would hinder growth and result in an economy 15 percent below the potential GDP. Avoiding this, she argued, required intersectoral balancing.
The ACA, however, is seldom recognized as macroeconomic policy, because we have forgotten the origins of GDP in economic stabilization. The recent policy experiments conducted under the Coronavirus Aid, Relief, and Economic Security (CARES) Act during the Covid-19 pandemic—most notably, enhanced unemployment insurance schemes, direct cash payments to households, and grants and loans to businesses—already demonstrated the importance of policymakers monitoring emerging imbalances in real time. As economic historian Morten Jerven also underscores in the developmental context of Africa, the uses of economic knowledge infrastructures such as NIPA beyond growth management will be critical going forward. Policymakers will be confronted with new imbalances, especially as they face the unprecedented challenge of transitioning into a “green” economy to mitigate the looming climate crisis.
Learn more about our disclosure policy here.
