Measuring labour input: is it about quantity, quality, or both?

Human capital, the stock of knowledge and skills embodied in people, is a key input in economic production. Changes in both the “quantity” and the “quality” of a country’s human capital stock influence economic growth and productivity performance (Égert et al., 2022). Traditional measures of labour input in economic growth and productivity analyses, such as total hours worked, focus solely on changes in the quantity of labour input, ignoring changes in the skill composition of the workforce. For example, these measures equate an hour worked by a highly experienced surgeon and an hour worked by a junior retail salesperson, disregarding their vastly different experience and skills.

Firms recognise that workers with different skills and experience are not perfect substitutes by paying them different wages. It is therefore possible to account for differences between workers by weighting their hours worked by their respective shares in total wages. Such measures are often referred to as Composition Adjusted Labour Input (CALI), Labour Services, or Quality Adjusted Labour Input (QALI). CALI measures provide an improved understanding of whether the average “quality” of labour is increasing or decreasing over time. In addition, they can play a crucial role in productivity analysis by more closely explaining the sources of economic growth. Economists often break down output growth into that explained by changes in labour input, capital input, and multifactor productivity (MFP) growth. In this framework, referred to as growth accounting, MFP growth is estimated as a residual, capturing all growth left unexplained by growth in labour and capital inputs. When a traditional measure of labour input is replaced with a CALI measure, a larger share of output growth is attributed to labour, reducing the “unexplained” share attributed to MFP and improving the explanation of the sources of economic growth. Nonetheless, the estimation of CALI comes at the cost of timeliness and resources, as it necessitates access to microdata sources.

The practice of using wages to reflect varying skills among workers assumes that hourly wages equate to hourly productivity. However, numerous factors can lead to disparities between wages and actual productivity. These factors include wage-setting methods, seniority within the workforce, workplace discrimination, and gender pay gaps, among others. Given the current lack of more precise measures of productivity, the existing literature employs wages as a proxy for productivity in constructing CALI measures.

Ward and Zinni (2024) reviews the most relevant literature on CALI estimates and follows a generic approach to produce CALI estimates for 21 countries. The study tests the sensitivity of CALI estimates to the selection of workers’ characteristics, classifying workers according to their educational levels, age (a proxy for years of working experience), gender, industry of work, and occupation. It then analyses the evolution of CALI over time and establishes the impact of accounting for CALI on MFP growth.

Growth in the composition of labour contributed positively to CALI growth between 2004 and 2018 in all countries covered in the study (Figure 1). The greatest contributions from changes in labour composition, and hence the largest increases in the average growth rate of CALI, are observed in Portugal, Poland, and Spain, where the labour composition component grew by 1.9%, 1.2% and 1.1% per year between 2004 and 2018, respectively, well above the average annual growth rate of all countries included in the study (0.6%). In a few countries, including Estonia, Latvia, Portugal, and Romania, accounting for the composition of the workforce changes the sign of growth in labour input from negative to positive, as compared with that suggested by the volume of hours worked.

During a recession and often in the years that follow, the compositional effect of CALI tends to be higher, pointing to an increase in the average skill level of those in employment (i.e. an increase in labour quality). Indeed, firms tend to shed labour and/or reduce hours worked among lower-skilled workers during a recession, while hoarding higher-skilled individuals. During the 2008-2009 recession, most countries experienced an increase in the composition component, which counterbalanced the decline in total hours worked and cushioned the fall in CALI (Figure 2).

MFP growth is revised downwards for all countries when using a CALI measure, suggesting that labour plays a larger role as a source of output growth than previously understood. While the revision to average annual MFP growth over 2004-2018 remains small in most countries, the impact on MFP growth may be relevant in countries that have experienced larger improvements in labour quality (Figure 3). A significant downward revision in average MFP growth ranging between 0.8% and 1.6% per year is found in Greece, Portugal and Spain, which is equivalent to a cumulative downward revision ranging between 10 and 20 percentage points in the MFP index over the whole period of analysis.

The OECD study leads to a few key takeaways:

• The integration of CALI into the growth accounting framework is essential in countries undergoing significant shifts in the composition of their workforce. In these countries, accounting for changes in both total hours worked (quantity) and the composition (quality) of labour is crucial for improving the understanding of the sources of economic growth.
• Educational attainment and age (a proxy for years of working experience) emerge as the two essential workers’ characteristics to consider when building CALI estimates and investigating the contribution of labour input to output growth. Occupation holds some explanatory power of changes in labour quality, possibly accounting for skills mismatches when considered alongside educational levels. Industry of work is found to be largely irrelevant.
• The inclusion of gender as a dimension to classify workers into different categories has little explanatory power and can be avoided. The use of wages in the estimation of CALI relies on the assumption that hourly wages equal hourly productivity. Gender pay gaps bring important limitations to this assumption, as they often reflect discrimination between women and men in the workplace, alongside differences in self-selection, propensity to compete, negotiation behaviour and risk aversion, rather than differences in workers’ actual productivity.
• National statistics institutes are typically best placed to produce CALI measures, as they benefit from access to a much wider range of data sources, sometimes confidential, and have the expertise to address representative biases to maximise the quality of their estimates. Decisions regarding its calculation will depend on the expected changes in the composition of the workforce overtime and/or the cost of producing CALI, including considerations of timeliness and resources.

• Égert, B., C. de la Maisonneuve and D. Turner (2022), “A new macroeconomic measure of human capital exploiting PISA and PIAAC: Linking education policies to productivity”, OECD Economics Department Working Papers, No. 1709, OECD Publishing, Paris.
• Ward, A. and B. Zinni (2024), “The composition of labour input: sensitivity testing and results for productivity analysis”, OECD Publishing, Paris.