Growth accounting is a method economists use to break down a country’s economic growth into its underlying sources: how much came from adding more workers, how much from investing in machinery and equipment, and how much from getting better at using those inputs. It answers a deceptively simple question: when an economy grows by, say, 2% in a year, what actually drove that growth?
The Core Idea
At its heart, growth accounting treats an economy like a recipe. Output depends on ingredients (labor and capital) and how skillfully you combine them. If a restaurant starts producing more meals, you’d want to know whether they hired more cooks, bought more ovens, or just figured out a more efficient way to run the kitchen. Growth accounting does the same thing at a national scale.
The framework identifies three drivers of economic growth: increases in capital (machinery, buildings, infrastructure), increases in labor (more workers or more hours worked), and improvements in productivity, meaning the economy’s ability to squeeze more output from the same inputs. That third component, productivity, turns out to be the most important and the hardest to measure.
How the Math Works
Growth accounting starts with a standard model of how economies produce output. Total output is a function of three things: the amount of capital, the amount of labor, and a productivity term that captures how efficiently those inputs are combined. The key equation breaks the percentage growth in output into three pieces:
Output growth = Productivity growth + (capital’s share × capital growth) + (labor’s share × labor growth)
Capital’s share and labor’s share refer to how much of national income goes to each factor. In most developed economies, labor’s share sits around 60-70% and capital’s share around 30-40%. These shares act as weights, telling you how much a given percentage increase in each input contributes to overall growth.
The clever part is how productivity gets calculated. Since you can measure output, capital, and labor directly, productivity growth is whatever’s left over after you subtract the contributions of capital and labor from total output growth. In other words, productivity growth equals output growth minus the weighted growth of capital and labor. This leftover piece is called the Solow residual, named after economist Robert Solow, who pioneered this approach in a landmark 1957 paper.
Total Factor Productivity: The Mystery Ingredient
The residual has a formal name: total factor productivity, or TFP. It captures an economy’s ability to generate more income from the same inputs. If a country increases its total output without using more workers or more machines, that gain shows up as higher TFP. It reflects things like better technology, smarter management practices, improved institutions, and gains from trade.
TFP also turns out to be the only source of sustained long-term growth in living standards. You can only add so many workers or build so many factories before running into diminishing returns. But improvements in how efficiently you use those resources can compound indefinitely. Historical data from the United States illustrates this clearly: of the roughly 2 percentage points of annual GDP growth per person over recent decades, about 1.3 percentage points came from TFP growth. Human capital (education and skills) contributed around 0.5 percentage points, while capital deepening and changes in employment rates filled in the rest.
The catch is that TFP is, by construction, everything the model can’t explain. Economist Moses Abramovitz famously called it “a measure of our ignorance.” It bundles genuine technological progress together with measurement errors, organizational changes, regulatory shifts, and anything else that affects output but isn’t captured by simply counting workers and machines.
Measuring Labor and Capital
The usefulness of growth accounting depends on how well you measure its inputs, and both labor and capital are trickier to quantify than they first appear.
For labor, simply counting the number of workers misses a lot. An hour of work by a surgeon contributes differently to output than an hour by an entry-level retail employee. To address this, statistical agencies create what’s called compositionally adjusted labor input. This approach categorizes workers by characteristics like age (as a proxy for experience), education level, gender, and industry, then weights each group’s hours by their share of total labor income. Education is the single biggest factor in these adjustments. The UK’s Office for National Statistics, for instance, sorts workers into six educational tiers, from those with no formal qualifications up to holders of advanced degrees, across 19 industry groups, three age brackets, and two gender categories, creating 684 distinct worker types.
Capital measurement has its own complications. Capital stock includes machinery, equipment, buildings, software, and infrastructure. Some of these assets, like computers, depreciate within a few years. Others, like railroad structures, can last up to 90 years. Economists need long historical investment records to build accurate estimates of the capital stock at any point in time, and they must account for the fact that older equipment may become economically obsolete well before it physically wears out.
What Growth Accounting Is Used For
Policymakers rely on growth accounting to diagnose economic performance and design targeted interventions. If a country’s growth is coming almost entirely from adding more workers rather than from productivity gains, that signals a vulnerability: growth will stall once the labor supply stops expanding. Conversely, if TFP is the main driver, the economy is on a more sustainable path.
International comparisons are another common application. Growth accounting lets economists ask why one country grew faster than another during a given period. Was it because they invested more heavily in equipment? Because their workforce became more educated? Or because they adopted technologies and practices that made existing resources more productive? Each answer points toward different policy prescriptions, whether that’s encouraging business investment, expanding education access, or reducing barriers to innovation.
The framework also helps with economic forecasting. Understanding how much of past growth came from each source provides a baseline for projecting future growth. Research from the Federal Reserve Bank of New York has decomposed the 1.3 percentage points of U.S. TFP growth further, attributing about 0.7 percentage points to research intensity (the economy’s investment in innovation), 0.3 to reductions in misallocation of resources, and 0.3 to population growth effects. If any of those factors shifts, so does the growth outlook.
Key Limitations
Growth accounting has been a workhorse of economics for over six decades, but it comes with real limitations that shape how seriously you should take any specific result.
The biggest issue is that TFP is a residual, not a direct measurement. Every error in measuring capital or labor gets swept into the productivity number. If you undercount capital investment, TFP looks artificially high. If you fail to adjust for improvements in worker skills, the same thing happens. The residual absorbs all the noise in the data alongside the genuine signal of technological progress.
There’s also a deep conceptual problem with separating capital accumulation from technological change. New technology often arrives embedded in new capital, like faster computer chips or more efficient manufacturing equipment. When a firm buys a better machine, is the resulting output gain a capital contribution or a technology contribution? Growth accounting has to assign it to one or the other, but in reality they’re tangled together. Some economic models argue that the residual is actually an endogenous function of capital growth itself, meaning the neat separation the framework promises may be somewhat artificial.
The standard framework also assumes competitive markets where prices reflect the true cost of production. In industries dominated by a few large firms, where prices deviate from marginal cost, the standard interpretation of the numbers breaks down. And as Solow himself acknowledged, the entire exercise rests on the concept of an aggregate production function, a theoretical simplification that treats the entire economy as a single giant factory. That abstraction, while useful, requires a certain leap of faith.
Despite these caveats, growth accounting remains the standard framework for understanding where economic growth comes from. No alternative method provides the same combination of transparency, simplicity, and practical usefulness for policymakers trying to figure out what’s driving their economy and what might slow it down.