Hours of work
Decent working time is a crucial part of decent work. Statistics on hours of work are essential to assess working conditions of employed persons.
ILOSTAT presents information on various indicators pertaining to hours of work, obtained from national sources. It features statistics on the share of employees working more than 48 hours per week (considered excessive working time) and the mean weekly hours actually worked per employed person. For users interested in more indicators, ILOSTAT also includes statistics on mean weekly hours actually worked per employed person and per employee (separately) by sex, economic activity and occupation, as well as employment by hours worked by sex.
This paper reviews trends and developments in both hours of work and the organization of working time (working time arrangements) and considers their implications for the future of work.
Concepts and definitions
The resolution adopted by the 19th ICLS promotes the collection of information on both hours usually worked and hours actually worked to allow for the proper identification of all groups defined in the resolution. ILOSTAT prioritizes the concept of hours actually worked, as this was promoted in the previous standards and is thus the one with more data availability.
The concept of hours usually worked relates to the typical value of hours actually worked in a job per a short reference period such as one week, over a long observation period of a month, quarter, season or year that comprises the short reference measurement period used. The typical value may be the modal value of the distribution of hours actually worked per short period over the long observation period, where meaningful. The short reference period for measuring hours usually worked should be the same as the reference period used to measure employment.
The concept of hours actually worked within the System of National Accounts (SNA) production boundary relates to the time that persons in employment spend directly on, and in relation to, productive activities; down time; and resting time during a specified time reference period. It thus includes (a) “direct hours” or the time spent carrying out the tasks and duties of a job; (b) “related hours”, or the time spent maintaining, facilitating or enhancing productive activities; (c) “down time”, or time when a person in a job cannot work due to machinery or process breakdown, accident, lack of supplies or power or Internet access; and (d) “resting time”, or time spent in short periods of rest, relief or refreshment, including tea, coffee or prayer breaks, generally practiced by custom or contract according to established norms and/or national circumstances.
Hours actually worked excludes time not worked during activities such as: (a) Annual leave, public holidays, sick leave, parental leave or maternity/paternity leave, other leave for personal or family reasons or civic duty; (b) Commuting time between work and home when no productive activity for the job is performed; (c) Time spent in certain educational activities; (d) Longer breaks distinguished from short resting time when no productive activity is performed (such as meal breaks or natural repose during long trips).
For a paid-employment job, hours paid for refers to the time for which employees have received payment from their employer (at normal or premium rates, in cash or in kind) during a specified reference period, regardless of whether the hours were actually worked or not. It includes time paid but not worked such as paid annual leave, paid public holidays and certain absences such as paid sick leave, and excludes time worked but not paid by the employer, such as unpaid overtime, and absences that are not paid by the employer, such as unpaid educational leave or maternity leave.
Data on working time presented in ILOSTAT reflects, to the extent possible, the hours worked in different types of working time arrangements (e.g. full-time and part-time) and include the hours worked in all jobs of employed persons (if the data are derived from a labour force survey).
Mean hours actually worked per week are calculated by dividing the total number of hours actually worked per week by: a) the total number of employee-held jobs during the same period, if the estimates are derived from an establishment survey; or b) the total number of persons in employment (or employees) if the statistics are derived from a labour force survey.1Resolution concerning the measurement of working time, adopted by the 18th International Conference of Labour Statisticians (December 2008)
Employment comprises all persons of working age who during a specified brief period, such as one week or one day, were in the following categories: a) paid employment (whether at work or with a job but not at work); or b) self-employment (whether at work or with an enterprise but not at work).2Resolution concerning statistics of work, employment and labour underutilization, adopted by the 19th International Conference of Labour Statisticians, Geneva, 2013
Labour force surveys are typically the preferred source of information on hours of work. Such surveys can be designed to cover virtually the entire non-institutional population of a given country, all branches of economic activity, all sectors of the economy and all categories of workers, including the self-employed, contributing family workers, casual workers and multiple jobholders.
Other types of household surveys could also be used as sources of data on hours of work, if they have an appropriate module on the topic.
In the absence of a labour force survey or other types of household surveys with a module on working time, an establishment survey can be used as a source of statistics on hours of work. However, the statistics derived from establishments surveys would typically not refer to the whole employed population but only to employees (and often only to formal sector employees or non-agricultural formal sector employees).
Interpretation and use of the indicator
The number of hours worked has an impact on the health and well-being of workers. Some persons in developed and developing economies working full-time have expressed concern about their long working hours and its effects on their family and community life.3Messenger, J.C. (ed.): Working time and workers’ preferences in industrialized countries: Finding the balance (Routledge, 2004). Additionally, the number of hours worked has an impact on workers’ productivity and on the labour costs of establishments. Measuring the level and trends in working time in a society, for different groups of persons and for individuals, is therefore important when monitoring working and living conditions as well as for analysing economic and broader social developments.4ILO: Report II: Measurement of working time, 18th International Conference of Labour Statisticians, Geneva, November-December 2008
Employers have also shown interest in enhancing the flexibility of working arrangements. They are increasingly negotiating non-standard working arrangements with their workers.5Policy suggestions that preserve health and safety, are family friendly, promote gender equality, enhance productivity and facilitate workers’ choice and influence their working hours are provided in: Lee, S., McCann, D. and Messenger, J.: Working time around the world (Geneva, ILO, 2007). Employees may work only part of the year or part of the week, work at night or on weekends, or enter or leave the workplace at different times of the day. They may have variable daily or weekly schedules, perhaps as part of a scheme that fixes their total working time over a longer period, such as one month or one year. Consequently, employed persons’ daily or weekly working time may show large variations, and a simple count of the number of people in employment or the weekly hours of work is insufficient to indicate the level and trend in the volume of work.
“Excessive” working time may be a concern when individuals work more than a “normal” workweek due to inadequate wages earned from the job or jobs they hold. In ILOSTAT, statistics are provided on persons who work more than 48 hours a week. Long hours can be voluntary or involuntary (when imposed by employers). “Inadequate employment related to excessive hours”, also called “over-employment” has been referred to as “a situation where persons in employment wanted or sought to work fewer hours than they did during the reference period, either in the same job or in another job, with a corresponding reduction of income”.6ILO: Final Report, 16th International Conference of Labour Statisticians, Geneva, October 1998
Few countries have actually measured “over-employment” so the measure of persons in employment for more than 48 hours a week could be used as a proxy for persons in employment who usually work beyond what is considered “normal hours” in many countries. However, whether or not this situation is actually desired cannot be assessed, so nothing can be assumed about how many hours people might wish to work. Clearly, the number of hours worked will vary across countries and depends on, other than personal choice, such important aspects as cultural norms, real wages and levels of development.
Statistics based on hours actually worked are not strictly comparable to statistics based on hours usually worked. A criterion using hours actually worked will generally yield a higher weekly average than usual hours, particularly if there are temporary reductions in working time as a result of holiday, illness, etc. that will have an impact on the measure of average weekly hours. Seasonal effects will also play an important role in fluctuations in hours actually worked. In addition, the specification of main job or all jobs may be an important one. In some countries, the time cut-off is based on hours spent in the main job; in others, on total hours spent in all jobs. Measures may therefore reflect hours actually or usually worked in the main job or in all jobs. Because of these and other differences that may be specific to a particular country, cross-country comparison of working time statistics should be undertaken with great care.
The various data collection methods also represent an important source of variation in the working time estimates. Household-based surveys that obtain data from working persons or from other household members can and often do cover the whole population, thus including the self-employed. As they use the information respondents provide, their response may contain errors. On the other hand, the data obtained from establishment surveys depend on the type, range and quality of their records on attendance and payment. Importantly, worker coverage in establishment surveys is limited, as they tend to only cover medium-to-large establishments in the formal sector with regular employees, and exclude managerial and peripheral staff as well as self-employed persons.