Measuring human capital

ONS currently measure Human Capital Stocks (HCS) using a well-established method that cumulates the expected value of discounted future earnings for individuals in the working age population. Human capital is defined as the stock of skills, knowledge and experience of an individual or population, which can productively be applied in the economy and is widely referred to as one of the main drivers of economic growth. This project makes an important contribution to improving this measure by incorporating health status into the HCS model and developing a skills classification measure that can be used instead of qualifications. The research also considers the impact of internal migration on regional HCS and assesses the extent to which these affect regional growth and standards of living.

This project makes several contributions to improving the measurement of human capital stocks (HCS) in the UK. We review existing measures and suggest a number of extensions and alternatives.

Previous research on HCS combined data from the Understanding Society survey and the Labour Force Survey to model the impact of health on retirement decisions and wages. We carried out research to understand how health status impacts various labour market variables. Using a full accounting framework, we developed historical and current estimates of how a person being in poor health, rather than good,  impacts HCS.

The UK has well-documented regional disparities in economic growth and productivity. Variation in the level of skills and qualifications of each region is one contributing factor. If more productive individuals congregate in some local areas, often cities, this can generate a brain drain elsewhere, and have a negative impact on growth in affected regions. We investigated internal migration by gender, age and qualification and assessed the likely implication for measures of HCS.

Finally, building on other ESCoE research we used online data sources to improve understanding of the skill sets associated with particular qualifications and occupations.  Current estimates divide workers by qualification level, but this does not always provide an accurate reflection of skills.

The project looks to extend the standard method for estimating human capital stocks (HCS), in different ways and using different data.

We build on previous work that captured health status in human capital measures. This requires measurement of both the potential HCS, including all persons of working age, and the productive HCS, including only those persons currently employed. Health affects participation, employment, hours worked, pay, and more. For this project we extended this work, exploring application to different datasets and extending estimates over time.

We developed a new framework to integrate human capital and output of the education sector into National Accounts. To do this we conceptualised schooling as a social investment, treating education output as the increment to lifetime income due to the year’s schooling.

For work on regional HCS, we estimated transition probabilities – the chance an individual transitions from one region to another. We outlined how these transition probabilities might be incorporated into the standard method to better measure regional HCS.

We used online job vacancy data to explore skills, occupations, and labour demand. While these data are large and rich with information, they are also unstructured and messy, so need considerable cleaning. To do this, we use natural language processing and other data science techniques.

We found that poor health leads to a reduction in human capital stocks (HCS) by about 12% in 2018. There are significant impacts of poor health on earnings, especially for males. However, many people with poor health are not economically active.

Treating formal schooling as an investment raises the level of GDP substantially, by between 7% and 16% between 2010 and 2018 on average, depending on the assumptions. This framework also increases the rate of real output and productivity growth in the education sector, but by increasing the relative size of this (still relatively slow-growing) sector, it reduces the average growth rates for the economy as a whole.

Our analysis suggests that cross-region mobility is highest among the young, highly-educated population and that taking into account regional mobility affects the relative values of regional HCS. However, the small sample sizes in the available datasets make the precise estimates uncertain.

Our research has supported ONS in the development of their human capital statistics. This is an important area for ONS, at the interface of economic and social statistics, and they have a workplan to deliver improvements over coming years. HM Treasury has also had interest in this topic in recent years, and our work supports policy making in this area. Our work on regional human capital estimates will aid understanding of the challenges and opportunities relating to “levelling up”.

The development of a new framework to think about education as an investment in social infrastructure, and later as a “schooling knowledge inventory”, provides a proof of concept to national accounts considering a revision to the national accounting rules. This work has been presented at international conferences and influenced thinking amongst the research community.

Our work using online job vacancy data supported ONS to launch and innovate their faster indicators of labour demand during the coronavirus pandemic. Our expert advice enabled the rapid and continual improvement of these statistics. The weekly indicators have been valuable insights to decision makers and analysts in government and beyond.

Romanko, O. (2023) ‘Technology and firm performance: evidence from online data sources’ King’s College London Doctoral Thesis

de Coulon, A., Da Silva Marioni, Larissa. and O’Mahony, M “Transition Probabilities, Wages and Regional Human Capital Stocks” ESCoE Discussion Paper Series, ESCoE DP 2022-28

de Coulon, A., Da Silva Marioni, Larissa. and O’Mahony, M “Transition probabilities, wages and regional human capital stocks” 7th WorldKLEMS Conference, Parallel Session, Regional, University of Manchester, 12-13 October 2022

Romanko, O. and O’Mahony, M. (2022) “The use of online job sites for measuring skills and labour market trends: A review” ESCoE Technical Report Series, ESCoE TR-19

ESCoE Human Capital Workshop: Exploring Skills and Education in partnership with ONS and ESRC, 22 February 2022 (online)

de Coulon, A., Da Silva Marioni, Larissa. and O’Mahony, M “Regional Human Capital Stocks” ESCoE Human Capital Workshop Exploring Skills and Education, 22 February 2022

Nwosu, C. (2022) ‘Essays on Student Higher Education Participation and Attainment in the United Kingdom’ King’s College London Doctoral Thesis

O’Mahony, M., Corrado, C. and Samek, L ‘ How does education contribute to productivity? An intangible infrastructure approach applied to the UK and the US’ 36^th Annual Virtual General Conference, IARIW Plenary Session: New Developments in Economic Measurement, IARIW, 23-27 August 2021

Corrado, C., O’Mahony, M. and Samek, L. ‘How does education contribute to productivity? A social infrastructure approach applied to the UKS and UK’ The Sixth World Klems Conference 2021, Session 1: Intangibles, 16 March 2021

O’Mahony, M. and Samek, L. (2021) “Incorporating Health Status into Human Capital Stocks: an Analysis for the UK” ESCoE Discussion Paper Series, ESCoE DP 2021-03

O’Mahony, M. and Samek, L. “‘Incorporating Health Status into Human Capital Stocks: An Analysis for the UK” ESCoE Blog, 11 March 2021

O’Mahony, M. and Samek, L. ‘Health and Human Capital’ The Sixth World Klems Conference 2021, Session 2: Human Capital, 10 March 2021.

ESCoE Online Job Vacancy Data Workshop, in partnership with ONS and Department for Business, Energy and Industrial Strategy, 9 December 2020

Andrieu, E., O’Kane, L., O’Mahony, M. and Romanko, O. ‘How well were companies prepared for COVID? Merging job vacancy, company accounts and web scraped data’, ESCoE and BEIS Online Job Vacancy Data Workshop, 9 Dec 2020

O’Mahony, M. and Samek, L. “Health and Human Capital” ESCoE Conference on Economic Measurement 2020, Contributed Session A: Capital Stock I, 16-18 September 2020

Romanko, O.,  Vassilev, G., Evans, K., Schofield, R. and Wood, M. “Developing Online Vacancies Data to Inform on Human Capital and Labour Market COVID Impacts and Changes” ESCoE Conference on Economic Measurement 2020, COVID-19 Session D: Effects of COVID-19 on Business and the Labour Market, 16-18 September 2020

Corrado, C., O’Mahony, M. and Samek, L. (2020) ‘Measuring Education Services Using Lifetime Incomes’ ESCoE Discussion Paper Series, ESCoE DP 2020-02

O’Mahony, M. “Health and Human capital”  National Bureau of Economic Research conference ‘The macroeconomic valuation of health’, Boston, Friday 8 November 2019