The disputes and chaos that followed the United States’ sudden imposition of blanket tariffs in April have threatened to pick apart the tightly woven fabric of global production. Around 300 million companies worldwide connected by an estimated 13 billion supply links now face unprecedented uncertainty. But the current confusion is just the most recent example of the economic disruption that has marked the last half-decade. Since the start of the COVID pandemic in 2020, surprising bottlenecks in global supply chains have made scholars reassess how the economy works. Production slowdowns and shortages of goods as varied as hand sanitizer (which required speciality chemicals imported into the United States during the pandemic) and airplanes (Airbus faced delays in meeting demand because of shortages of important components in 2024) have exposed the vulnerabilities of a global economic system in which goods cross borders multiple times at successive stages of production and assembly. They have also challenged conventional understandings of how best to measure growth and productivity.
Trade and technology have rewired global production, but economic statistics measuring outcomes, such as GDP growth and productivity, are still collected on the basis of frameworks devised in the 1940s that focus primarily on the demand side of the economy and its current and past performance. They say little about the supply side, or an economy’s ability to respond to pressure. As a result, around four-fifths of the output of advanced economies has been designated by economists as “hard to measure.” A disproportionate focus on easier-to-measure industries such as manufacturing to evaluate performance underestimates the importance of the industries actually powering the modern economy.
In an era of unprecedented interconnectedness—and upheaval—policymakers need new tools to track the condition of global production networks and capture forms of activity unimaginable at the birth of modern economic statistics. The overhaul will require more and better data gathering to fill gaps in knowledge, including indications of supply chain vulnerabilities, and innovative techniques for measuring how qualitative and intangible aspects of an economy contribute to growth and inflation. A new framework that wrestles with what standard statistics cannot is not just an academic matter; it is the only way leaders can make sound economic policy in uncertain times.
A NECESSITY OF WAR
The current international framework for measuring economic growth and productivity is the System of National Accounts (SNA), a set of statistical standards agreed upon every ten to 20 years by a United Nations–led committee of official statisticians. It includes GDP, long considered by economists and policymakers to be the single best measure of economic progress.
The SNA was established by American and British economists, including John Maynard Keynes, during World War II. As Allied countries dedicated resources and productive capacity to the war effort, they needed to strike a balance between consumption and war production and devised the SNA to measure, among other things, domestic demand. The system took hold in the years following the war and has been revised roughly every decade in response to structural changes in the economy, such as the rising influence of finance. The revision process is intentionally slow and incremental, requiring consensus among UN member countries before any changes are adopted.
Today’s SNA is a sophisticated measure of the mid-twentieth-century economy, in which the manufacturing sector dominated, production took place within the borders of single countries, and broadband networks and data did not yet power economic growth. Recent updates to the SNA have made modest, modernizing changes. But no amount of tinkering around the edges will change the fundamental obsolescence of a system built around the drivers of growth and productivity from a bygone era. The SNA can perform the autopsy, but it cannot make a diagnosis or prescribe treatment.
GET WITH THE TIMES
The SNA has standardized the measurement of economic growth and enabled international comparisons for nearly a century. It has not, however, kept pace with changes in economic activity enabled by globalization and technological advances. Since the 1980s, manufacturing and service production, the networks of which were once national, have shifted to rely on global value chains. Where global trade once meant completed cars or refrigerators being shipped between countries, it now entails the transport of highly specialized components across borders at each stage of production and assembly.
New business models have flourished as the global economy has become more elaborate and entangled. Traditional integrated manufacturers that controlled the process from design to production and wholesaling have given way to “factory-less goods producers” that design, wholesale, and retail; contract companies that carry out the manufacturing; and downstream “servitized” manufacturers that focus on selling related services and maintenance. Contracted-out manufacturing accounts for some 15 to 20 percent of output in sectors such as electronics and pharmaceuticals in the United States and the United Kingdom, and the bundling of services with products (Rolls-Royce, for example, sells real-time monitoring services for its airplane engines) has become a popular practice among large companies.
These shifts have led to broad economic growth as specialization and the division of labor increase productivity in an ever-widening global market. But the trade-offs have become clear: goods have become cheaper, more accessible, and of higher quality, but companies have grown overly reliant on a few overseas suppliers for key components, making them more vulnerable to global economic shocks that are only becoming more common. Traditional economic statistics were not designed for this complex network of interdependencies. The lack of reliable measurements on the supply side of the economy has made it difficult to capture the patterns and condition of global production chains even in calmer times.
So, too, has the inadaptability of these statistics in the face of sweeping technological advancement. Traditional stalwart measures such as GDP are not capable of account for the effects of cost-saving and productivity-increasing infrastructure such as cloud services on growth. They also underestimate, for example, the value of data to large companies, accounting only for the cost of building data centers and not the less tangible productivity enabled by data collection and use. What is more, statisticians cannot agree on how to record the value of widely used free digital services such as open-source software.
Central banks and researchers have begun to construct broad measures of supply chain pressures, such as the Federal Reserve’s Global Supply Chain Pressure index, because of the central role that such stresses have played in driving the inflation that has racked the global economy since the pandemic. But statisticians—and therefore policymakers—still lack sufficient metrics for measuring the risks to supply chains for specific products, even high-profile ones such as the iPhone. Nor do they have the tools or expertise to understand how digitization and artificial intelligence are reshaping business models and trade. Put simply, they are attempting to craft policy without a useful statistical guide to current and future production in today’s economy.
FILL IN THE BLANKS
Developing statistics about global production networks is not impossible. Academic experts in this area, such as the University of Cambridge economist Vasco Carvalho, have explained how tax and payments data could be used to understand and track the flow of product components. But a truly effective update of economic statistics must go beyond mapping production networks and the new business models enabled by technology. It must also focus on capturing the economic resilience and future potential of an economy while better detailing its current patterns of production.
The economic concept known as comprehensive wealth is key to this reimagining. Comprehensive wealth broadly describes a country’s balance sheet, including conventional productive capacity, such as the condition of the buildings and equipment owned by businesses, and its national infrastructure, from roads to ports. Crucially, it also should factor in communications networks and their potential chokepoints and vulnerabilities, along with other intangible infrastructure, including the data available to the government and private sectors, and “digital public infrastructure,” meaning hardware such as cloud servers and telecom networks; data and identity verification systems, which enable the digitization of public services and store all the economic and official statistics that businesses and governments need); and applications, including payment systems and government services. The digital stack is peppered with its own bottlenecks and vulnerabilities, and any government concerned with national security and resilience will need to be able to quantify these risks.
A national balance sheet needs a measure of human capital to quantify how the skills and health of the workforce contribute to productivity and earnings. It should also take into account a country’s natural capital: although some of the natural resources that enable economic activity, such as the rare earths and other critical minerals needed in cutting-edge technology, are partially valued by current statistical frameworks, a fuller account must also consider undervalued resources, such as the national parks that provide leisure services and the soil quality and biodiversity that enable agricultural productivity.
Finally, a truly comprehensive statistical account should include what economists such as the Nobel Prize winners Daron Acemoglu, Simon Johnson, and James Robinson have identified as the contribution that a country’s legal and governmental institutions make to its prosperity through confidence in the rule of law, the security of contracts, and economic freedoms. Available metrics do not distinguish between productivity in countries with “extractive” institutions that concentrate wealth among elites at the expense of the broader population, discouraging private investment and individual skill acquisition, and countries in which institutions provide the stability that allows all people to thrive in their jobs or businesses.
WHEN NUMBERS GET SERIOUS
Switching from traditional economic surveys of consumers and businesses to novel data sources will be a cultural revolution for traditionally cautious statisticians. Collecting the volume of data needed to build a more complete picture of a country’s productivity and its economic vulnerabilities will take considerable planning by national statistical agencies, whose resources have been steadily reduced in most countries since the global financial crisis of 2008 tightened government budgets worldwide. It will also require an international effort to develop a consensus on definitions and data standards. But with response rates to official statistical surveys declining and businesses and individuals increasingly afflicted with “survey fatigue,” profound changes are both necessary and inevitable.
The very revolutions in data collection and technology that have necessitated a new statistical paradigm will be useful in creating it. Much of the underlying data already exists or could be collected, from the payment and tax numbers needed to track global production networks, to employers’ payroll figures, information collected from store scanner systems, and the many new sources of sensor and satellite data. Pilot data initiatives such as the Anthropic Economic Index can help statisticians track the adoption and use of artificial intelligence across society, a defining technological shift that is currently invisible in economic statistics.
Global production networks are under immense strain. Policymakers rightly expect their statistical systems to deliver useful information on the state of their economies. Just as the imperatives of war led to the development of the SNA in the 1940s, the new exigencies of a world riven by trade wars, geopolitical instability, and supply shocks require a modern statistical framework. Without it, a true grasp of the nature of the global economy will remain elusive—and the world will remain in the dark.
Loading…
