Multi-Factor Productivity Growth

What Is Multifactor Productivity Growth?

Multifactor productivity (MFP) growth — sometimes called total factor productivity (TFP) growth — measures the portion of output growth that cannot be explained by the accumulation of measurable inputs such as labour and capital. It is, in the famous phrase of economist Robert Solow, the "residual": the part of economic growth left over after accounting for how many more workers are employed and how much more machinery, equipment, and structures they have to work with.

This residual captures a wide range of forces that make an economy more efficient. Technological innovation is the most commonly cited source, but MFP growth also reflects improvements in managerial practices, gains from the reallocation of resources toward more productive firms and sectors, the benefits of economies of scale, and the diffusion of knowledge across the economy. Because it encompasses all of these channels, MFP growth is often treated as the broadest available gauge of an economy's dynamism and innovative capacity.

Economists assign MFP growth an outsized role in long-run prosperity. The accumulation of capital is subject to diminishing returns — adding more machines eventually yields less and less additional output — but improvements in the efficiency with which inputs are combined face no such inherent limit. Over the very long run, it is sustained MFP growth that separates economies that achieve rising living standards from those that stagnate.

The indicator is inherently difficult to measure, however. Because it is defined as a residual, it absorbs every form of measurement error in the output and input data. Economists sometimes quip that MFP growth is a "measure of our ignorance" — a reminder that the number reflects not only genuine efficiency gains but also whatever the statistical framework fails to capture. This feature makes MFP both the most conceptually important and the most practically elusive of the major productivity indicators.

The OECD publishes MFP estimates for its member countries, and many national statistical agencies produce their own series as part of their growth-accounting programmes. Because the estimates depend on the full apparatus of national accounts — real output, capital stocks, labour-input measures, and income shares — they are typically available only with a significant lag and are subject to substantial revision as underlying data are updated.

How It Is Calculated

MFP growth is derived from a growth accounting framework that decomposes output growth into contributions from inputs and a residual. The standard formulation assumes a constant-returns-to-scale production function and competitive factor markets:

$\Delta \ln Y = \alpha \, \Delta \ln K + (1 - \alpha) \, \Delta \ln L + \Delta \ln \text{MFP}$

Rearranging to isolate the residual:

$\Delta \ln \text{MFP} = \Delta \ln Y - \alpha \, \Delta \ln K - (1 - \alpha) \, \Delta \ln L$

where $Y$ is real output, $K$ is the capital services index, $L$ is labour input (typically quality-adjusted hours worked), and $\alpha$ is the output elasticity of capital, usually proxied by capital's share of total income. The term $(1 - \alpha)$ is labour's share.

Input Measurement

The quality of MFP estimates depends critically on how inputs are measured. Modern growth accounts go well beyond simple headcounts and aggregate capital stocks.

Labour input is typically adjusted for changes in workforce composition — shifts in educational attainment, age structure, and experience — so that "labour quality" improvements are attributed to the labour input rather than to the residual. Similarly, capital input is measured as a flow of capital services rather than the stock itself, weighted by the rental prices of different asset types. This ensures that a dollar's worth of short-lived, high-depreciation ICT equipment is treated differently from a dollar's worth of long-lived structures.

The Role of Income Shares

The parameter $\alpha$ — capital's share of national income — serves as the weight on capital input growth in the decomposition. In a perfectly competitive economy, this share equals the output elasticity of capital. In practice, statistical agencies estimate $\alpha$ from national accounts data as the ratio of gross operating surplus to total value added. The complement, $(1 - \alpha)$, is compensation of employees as a share of value added.

This share is not truly constant over time, and its measurement is complicated by self-employment income, which contains both a labour and a capital component. Statistical agencies adopt various conventions to split this mixed income, and the choice can meaningfully affect the estimated MFP series. Furthermore, the growth-accounting framework assumes perfectly competitive markets and constant returns to scale — assumptions that are approximations at best and may systematically bias the residual in economies with significant market power or increasing returns.

Extensions and Refinements

Some growth-accounting frameworks extend the basic model by distinguishing among many types of capital (ICT, non-ICT, structures, intellectual property) and many types of labour (differentiated by education, age, and gender). These richer decompositions provide more granular insight into the sources of growth but require correspondingly more detailed data and impose additional modelling assumptions.

How to Read the Numbers

MFP growth rates are typically small — often well under 1 percentage point per year — but their cumulative impact over decades is enormous. The table below offers a rough interpretive guide for annual MFP growth in a mature advanced economy.

Negative readings do not necessarily mean the economy has become less technologically capable. They can arise from mismeasurement of output (particularly in hard-to-measure service sectors), from temporary disruptions that reduce the efficiency of existing input combinations, or from large-scale resource reallocation that temporarily lowers aggregate efficiency before producing gains in later periods.

Because MFP is a residual, it is among the most heavily revised macroeconomic indicators. Revisions to GDP, capital stock estimates, or labour input data all feed through to the MFP series. Analysts should treat any single year's estimate with caution and focus on multi-year averages to identify the underlying trend. Five-year or ten-year moving averages are common in the academic literature for precisely this reason.

Economic Significance

MFP growth lies at the heart of the growth debate. The slowdown in measured MFP growth across most advanced economies since the mid-2000s — well before the global financial crisis — has prompted a vigorous academic and policy discussion. Some researchers argue that the great innovations of the twentieth century (electrification, the internal combustion engine, indoor plumbing) were uniquely transformative and that recent digital innovations, while impressive, simply do not deliver the same broad-based productivity gains. Others contend that official statistics systematically undercount the value created by the digital economy, from free online services to improvements in product quality that price indices fail to capture.

For central banks, MFP growth is a key input into estimates of potential output. A persistent decline in MFP growth means that the economy's speed limit has fallen, which implies that interest rates consistent with full employment and stable inflation — the so-called neutral or equilibrium rate — are also lower. This has profound implications for the conduct of monetary policy, particularly when the nominal neutral rate falls close to the effective lower bound on interest rates. In such an environment, monetary authorities have less room to cut rates in response to downturns, making unconventional policy tools more likely to be needed.

Governments seeking to raise long-run growth have few levers more powerful than policies that support MFP. Investment in basic research, a well-functioning intellectual-property regime, competitive product markets that reward innovation and allow inefficient firms to exit, and an education system that equips workers to adopt new technologies all contribute to the residual. Because MFP captures the combined effect of these forces, tracking it over time provides a high-level diagnostic of whether an economy's innovation ecosystem is functioning well.

For businesses, MFP trends signal the broader environment in which they operate. In periods of rapid MFP growth, early adopters of new technologies and processes gain a competitive edge, and creative destruction accelerates. In periods of slow MFP growth, competitive dynamics tend to favour incumbents and the returns to innovation investment may appear lower, potentially discouraging the very spending needed to reignite productivity gains. This feedback loop between weak MFP growth and declining innovation incentives is one of the mechanisms through which a productivity slowdown can become self-reinforcing.

At the international level, differences in MFP growth explain a significant share of the divergence in income levels among advanced economies. Countries at the technological frontier must generate MFP gains through original innovation, while those behind the frontier can draw on a larger stock of existing ideas to adopt and adapt. This distinction shapes the policy priorities: frontier economies need strong basic-research institutions and risk capital markets; lagging economies need effective technology-transfer channels and the absorptive capacity to deploy foreign innovations.

Related Indicators

• Labour Productivity Growth — output per hour worked growth, which MFP partially explains
• Real GDP Growth (QoQ) — the headline output growth rate that MFP helps decompose
• R&D Spending (% of GDP) — investment in research and development, a key driver of MFP
• Capital Per Worker — the capital-deepening component that MFP strips out