What counts as a job?


When key labour market indicators barely move while millions suffer from the deepest economic crisis in decades, it is time to revise them.

Poor indicators make for poor public policy.

The global crisis highlights the compelling need to set clear and relevant international standards to collect labour statistics and measure labour market trends.

This job falls to the International Conference of Labour Statisticians (ICLS), which is convened by the International Labour Organisation (ILO) every five years in Geneva.

Between 2 and 11 October of this year, the 19th ICLS will re-consider and re-set the guidelines for Measurement of Work and the Labour Force.

Today’s standards date from the 13th ICLS in 1982.

Thirty years ago the main objective of statistics on employment was to measure labour inputs in production, to calculate GDP and measure economic growth, and the lens for measuring employment and unemployment was constructed to meet the needs of producing the National Accounts.

To this end, the ‘economically active’ population was defined as those persons engaged (or seeking to be engaged) in producing the output that counts as GDP. This meant that:

  • Persons (notably women) doing household work were not included in the ‘economically active’ population, even though their work clearly supports the household and contributes to national well-being, because own-production of household services is not counted as output in the National Accounts;
  • Persons engaged in subsistence production were counted as ‘employed’, even though they receive little or no market income and (most) would want a paid job if there were one.

Today, policy makers want to know about labour markets, the effectiveness of labour market programs, labour underutilisation, the participation of different sub-groups of the population and the engagement with the market economy.

For countries where four-fifths of the population is engaged in subsistence agriculture, a low official unemployment rate that is stable over time is a misleading statistical artefact.

To label people keeping families afloat as ‘economically inactive’ is both wrong and offensive.


A new framework

The 19th ICLS will consider an entirely new framework, designed to measure all forms of work done by persons over 15 years of age, using a ‘main purpose’ test.

Using a one-hour criterion to capture all labour input and work done, five new categories of work are proposed:

  • Own-production work is that work done for the main purpose of providing goods or services for use by own self/household
  • Employment is that work done for the main purpose of generating an income
  • Trainee work is that work done for the main purpose of acquiring skills or workplace experience
  • Volunteer work is that work done for the main purpose of helping others
  • Other work includes state mandated work such as community service orders made by a court or compulsory work tests required to be met in order to receive a social benefit

In any week, most people will spend some time doing more than one form of work.

For example, they may be employed (i.e. work for wages/income) as well as performing household chores, volunteering with a community organisation, studying, or a combination of these.

The main purpose of this is to allocate workers across the five categories without double-counting, to enable accurate estimates of the labour force.


Employment, unemployment and the labour force in the new framework

Where the main purpose of work done is production for own/household use, the worker will no longer be counted as ‘employed’ – even where some of the output they produce is bartered or sold.

This is a major change for countries where most or many workers are engaged in subsistence agriculture with little or no engagement with the market economy.

This will identify changes in the population shares engaged in subsistence production and paid employment.

It will measure adequately the market-sector unemployment, and help quantify (rather than conceal) the true extent of underutilised labour in the country.

Excluding persons engaged in own-production work, a person will be counted as employed if they worked for at least one hour with the main purpose of generating an income. This includes employees and self-employed persons.

Regrettably, to date the ILO Department of Statistics has proposed no criteria for determining what constitutes ‘an income’ for work to count as employment.

The 18th ICLS recognised ‘low pay’ as a form of inadequate employment and called for it to be measured.

In our view, if the main purpose of work is to generate an income but the income generated is paltry, then it should not count as employment.

A definition of income appropriate to national circumstances – such as two-thirds of median hourly earnings, or national hourly minimum wage – is essential to any meaningful measure of employment, not just to the measurement of decent work.

Persons who did not work for at least one hour with the main purpose of generating an income, but did work for at least one hour for other reasons, will be classified according to the main purpose of their work as trainee workers, volunteers, or other, but will not be counted as employed.

Persons will be classified as ‘unemployed’ if they:

  • Were ‘without employment’ in the previous week, and
  • Actively sought employment within the last month; and
  • Were available to start employment within a short period

The ‘labour force’ is the total of persons employed and unemployed. The unemployment rate is the number of persons unemployed divided by the labour force.



A person is in time-related underemployment if they are working less than full-time hours and want to work more hours.

A person may also be underemployed if they have skills which are not utilised in their current job (‘skills mismatch’).

The 18th ICLS asked the ILO to develop measures of skills mismatch for consideration by the 19th ICLS with a view to adopting an international standard, but this work has not yet been much advanced by the Office.

The new framework proposes that ‘labour slack’ based only on time-related underemployment be measured and reported.

When added to the unemployment numbers it provides a broader gauge of the true need for more and better jobs, but not broad enough to encompass the full extent of underemployment.


Underutilisation and ‘marginal attachment’

There are several dimensions to labour underutilisation.

‘Discouraged workers’ are persons without employment who are available to start work within a short period, but have given up actively looking for work.

A smaller group may have sought work but not be available to start within a short period.

Historically, this group has been said to have a ‘marginal attachment’ to the labour market.

The new framework proposes that this ‘marginal attachment’ be measured and reported.

When added to the unemployed and underemployed numbers it provides a still broader gauge of the true need for more and better jobs.


Trainees and apprentices

The ILO proposal for a new framework struck some difficulty in finding an appropriate classification for apprentices and trainees.

The situation differs profoundly between countries, with paid apprenticeships and traineeships having extremely close affinity with employment (main purpose generating income) in many instances, while in others this work is unpaid (main purpose acquiring skills or workplace experience).

The ILO proposed grouping all apprentices and trainees in the ‘trainee’ category, removing them from employment and the labour force, and removing young persons seeking such apprenticeships from the unemployed.

This was strongly resisted by workers, employers, and most government experts, and received hesitant support at best from some international observers.


What next?

The 19th ICLS will set the framework for generating statistics of work and the labour force for decades to come.

The decision will be determined if necessary by votes cast.

Getting better data about the true need for more and better jobs can help to refocus policies.

Trade unions should raise the issues with their governments before the ICLS to underline the importance of a new definition of work, the inclusion of an income criterion for the definition of employment, the broad and encompassing measurement of underemployment and underutilisation, and the inclusion of apprentices and trainees in employment were the main purpose is generating an income.

This sounds rather dry and technical, but disguising or revealing the real employment situation is obviously eminently political.


This article was original published on Global Labour Column.