Understanding ABS productivity statistics

Productivity growth has been a cornerstone of Australia’s economic growth and material wellbeing. Sustained improvements in productivity are essential for addressing demographic challenges and balancing environmental constraints with economic progress. Because productivity growth plays a pivotal role in raising living standards, it is a central focus for policymakers, economists, and the broader community.

The Australian Bureau of Statistics (ABS) provides a suite of productivity measures and data integration programs that underpin rigorous productivity analysis. This information paper outlines the key ABS productivity measures and examines the drivers of productivity growth, as well as their links to broader macroeconomic indicators.

Productivity measures how efficiently inputs - such as labour, capital, and resources – are transformed into outputs like goods and services. Productivity improves when more output is produced from the same inputs, or when the same level of output is achieved with fewer inputs. High levels of productivity and strong productivity growth are desirable because they support higher material living standards while also contributing to non-material benefits, including greater leisure time and enhanced overall quality of life.

The ABS compiles and publishes comprehensive measures of output, inputs, and productivity for individual industries and for the economy. Chapter 19 of Australian System of National Accounts: Concepts, Sources and Methods provides further information on the compilation of productivity statistics. The definitions of the key productivity measures and their associated concepts of output and inputs are outlined below.

Labour productivity measures the amount of real output produced per unit of labour input. It is typically expressed as the volume of gross value added (GVA) produced per hour worked. When labour productivity increases, the workforce is producing more output for each hour worked, which often leads to higher real wages and improved living standards. While labour productivity is easy to interpret, it is a partial productivity measure because it does not account for the contributions of other inputs beyond labour. In practice, labour productivity is strongly influenced by factors such as capital investment, technology, and the use of intermediate inputs. Further discussion on labour productivity measures is available in A primer on labour productivity.

GVA-based MFP measures the volume of GVA produced per unit of combined labour and capital inputs. It reflects how efficiently these two primary inputs are used together to generate output. By considering both labour and capital, industry-level MFP provides insight into an industry’s ability to contribute to economy-wide growth in income per unit of primary input. These measures are widely used in macro-level productivity analysis.

KLEMS MFP is defined as the volume of gross output divided by the combined volume of all inputs - capital (K), labour (L), energy (E), materials (M), and services (S). This approach offers a more detailed and comprehensive view of productivity by fully accounting for both primary (K, L) and intermediate (E, M and S) inputs in the production of gross output. Because of its input granularity, KLEMS MFP is particularly well suited to detailed industry-level analysis.

Traditionally, productivity measurement relies on highly aggregated national accounts data. However, over recent decades, the increasing availability of business‑level and employee‑level microdata has transformed both academic research and official statistics. The granularity of microdata enables productivity analysis that goes beyond what is possible with macro data. Examples include the analysis of productivity distribution within narrowly defined industries; decompositions of aggregate productivity growth into components such as within-firm improvements, between-firm reallocation and business entry and exit; and causal and descriptive links between productivity and business characteristics.

The ABS data integration combines data from two or more sources at the unit level (for example, persons and/or businesses). These sources include data that enable rigorous productivity analysis. Examples of integrated datasets include the Business Longitudinal Analysis Data Environment (BLADE), Person Level Integrated Data Asset (PLIDA), Linked Employer-Employee Dataset (LEED) and Longitudinal LEED (L-LEED).

Productivity research at the meso level sits between economy‑wide aggregates and micro-level analysis and is well suited to thematic accounts that provide deeper insight into parts of the economy. By organising data around an economic theme rather than a single industry - such as tourism, care, or digital services - these accounts can bring together production, labour inputs, capital, and intermediate use in a coherent framework. This creates meso‑level indicators that allow productivity to be examined in contexts where standard industry classifications are too broad or where economic activity spans multiple industries. In this sense, thematic accounts can offer a practical way to improve understanding of productivity, constraints, and structural change without losing the consistency of national accounting frameworks.

The Tourism Satellite Account provides a clear precedent for this approach, demonstrating how a set of productive activities that cut across industry definitions can be measured coherently and linked back to the core national accounts.

There is a strong opportunity to further develop this meso‑level productivity framework as industrial classifications are updated over the next couple of years, allowing new and emerging forms of economic activity to be more clearly identified. Aligning thematic accounts with updated classifications would strengthen both the relevance and analytical power of productivity statistics, supporting more nuanced policy analysis in areas of growing economic and social importance.

Productivity growth reflects the interaction of multiple underlying factors that influence how efficiently labour, capital and other inputs are combined in production. These factors include skills and human capital, investment in physical capital, competition, business dynamism, innovation and technology. Together, these forces interact and reinforce one another, and hence sustained productivity growth usually depends on progress across the whole system.

Sustained growth in productivity is closely linked to a wide range of positive macroeconomic outcomes. By enabling more output to be produced for each hour worked, productivity growth is a fundamental driver of higher living standards, as it expands an economy’s capacity to generate more income and consumption possibilities without requiring additional labour. It also underpins real wage growth: when workers become more productive, firms can afford to pay higher real wages while maintaining profitability. Strong productivity growth also helps alleviate inflationary pressures, as increases in demand or wages are more likely to be absorbed through efficiency gains as opposed to being passed on as higher prices.

Higher productivity reflects more efficient use of labour, capital, and technology, improving overall resource allocation within the economy. This efficiency translates into lower unit labour costs, enhancing the competitiveness of firms both domestically and internationally, supporting investment, and strengthening economic growth in a sustainable, non-inflationary manner.

Measuring productivity growth, particularly in parts of the economy such as health care and education, presents unique challenges that stem from a difficulty in capturing improvements in the quality of services provided to consumers.

The core challenge underpinning non-market productivity measurement is the absence of economically significant prices. For market activities, prices at which products are bought and sold reflect supply and demand. Volume measures of market-based output can be obtained through price deflation, which implicitly captures quality change. However, non-market services are generally provided free of charge, or at heavily subsidised prices that don't reflect economic value. Price deflation isn’t possible for these activities, because prices either don’t have sufficient economic meaning or don’t exist at all.

Growth in output volumes for non-market activity are generally measured by tracking changes in the quantities of services produced (such as the number of hospital admissions, or the number of students enrolled in the school system), but these types of indicators are unlikely to capture the outcomes that matter to consumers (such as faster recovery from surgery, or higher exam scores). When improvements in service quality cannot be measured, it is likely that output volumes are underestimated, with a consequence that productivity growth is also likely to be underestimated. A more detailed discussion on these measurement challenges is available in A primer on labour productivity. This article also outlines other measurement challenges pertaining to the non-market sector in areas such as public policy making and national defence.

The UK’s Office for National Statistics has used outcomes-based data as a proxy to measure improvements in non-market service quality in Public service productivity estimates: sources and methods. In Australia, the Productivity Commission recently published an alternative approach to measuring productivity in health care adopting an outcomes-based focus from the perspective of the patient in Advances in measuring healthcare productivity.

The ABS is advancing work to separate the health care and education industries into distinct underlying market and non‑market components. Although a significant share of activity in these sectors is market‑based - such as private schools, general practitioners and medical specialists - current national accounts and productivity statistics classify all activity in these industries as non‑market. This obscures important differences in pricing, incentives and output measurement. By delineating market and non‑market segments, productivity analysis can better capture the distinct characteristics, growth patterns and underlying drivers within each part of the industry. Experimental productivity segmented estimates for education are scheduled for release in 2026–27 and for health in 2027–28.

The ABS statistics are a key input into several international productivity databases, notably World KLEMS, OECD productivity databases, and the OECD’s Structural Analysis database (STAN). In these systems, ABS data are either incorporated directly as official national statistical inputs or are harmonised and re‑expressed to enable cross‑country comparability. While international databases may apply additional standardisation, aggregation, or modelling assumptions, the underlying Australian data on output, labour input, capital services, and MFP originate from ABS national accounts, labour accounts, and productivity releases, ensuring consistency with Australia’s official statistics framework.

Productivity is best understood as a long-term economic concept. Short-run productivity measures can be affected by modelling assumptions, measurement challenges, and limitations in the underlying data. High frequency productivity estimates - whether quarterly or annual – can reflect temporary shocks that may obscure underlying long-term trends. For this reason, productivity statistics are most informative when examined across productivity growth cycles. Further information about productivity growth cycles is available in the Feature Article: Experimental Estimates of Industry Value Added Growth Cycles.

The article Interpreting ABS productivity statistics outlines some of the key assumptions and limitations to consider when interpreting ABS productivity statistics. These include, but are not limited to:

• Influence of factors outside the production boundary: some external factors, such as weather conditions, can affect the productivity performance of the agriculture industry, even though they are not captured as inputs in the national accounts.
• Assumed proportionality between capital services and capital stock: capital services are modelled as proportional to the capital stock, an approach that does not account for varying capital utilisation throughout the business cycle or delays between investment and productive use. This is particularly relevant in the mining industry, where substantial investment occurs well before new assets become operational and production commences.
• Indirect measurement of capital services: capital services are not directly observed and must be estimated using several assumptions.
• Challenges in non‑market sectors: for non‑market industries such as health and education, the absence of meaningful market prices and difficulties in measuring changes in service quality over time can make productivity estimation more complex and less precise.