Japan (1)

Title of the survey

Monthly Labour Survey (national and prefectural surveys)

Organization responsible

Ministry of Labour, Policy Planning and Research Department

Periodicity of the survey

Monthly.

Objectives of the survey

To measure and monitor changes in employment, earnings and hours worked, both at the national and prefectural levels.

Main labour topics covered by the survey

Employment, earnings, hours actually worked and days worked.

Reference period

A whole month, from the day following the final pay-day of the previous month to the pay-day of a given month.

Coverage of the survey

Geographical

The whole country.

Industrial

All divisions of economic activity, except agriculture, hunting, forestry and fishing, private households with employed persons and Government services.

Establishments

Establishments with five to 29 employees and establishments with 30 and more employees, separately.

Persons

Regular employees, including working directors. Working proprietors, self-employed persons and unpaid family workers are excluded.

Occupations

Data are not collected by occupation.

Concepts and definitions

Employment

Regular employees are workers who come under the following categories: Part-time workers are either workers whose daily scheduled working hours are shorter than those of regular employees, or workers whose daily scheduled working hours are the same as those of regular employees but whose weekly scheduled working days are shorter than those of regular employees. The following categories of workers are included as employees, but not separately identified: wage earners and salaried employees, apprentices, temporary and seasonal workers (provided they meet the criteria of the definition of regular employees), persons temporarily absent from work because of paid or unpaid vacation or holiday, industrial dispute, sickness or accident, or for any other reasons, whether authorized or not, as well as persons temporarily present on payroll during notice period preceding retirement, resignation or dismissal. Excluded are trainees and employees on probation, piece workers, commission agents, home workers, casual employees, employees sub-contracted from other companies or firms, employees from temporary work agencies, and persons absent from work because of temporary or indefinite lay off or temporary military service.

Earnings

Data are collected on gross cash earnings of regular employees. Gross cash earnings represent the sum of money earned before deductions for income tax, social insurance contributions, union dues, payment for goods purchased, etc. They are classified as follows: Total cash earnings correspond to the total of contractual cash earning and special cash earnings. Earnings also include cash allowances such as cost-of-living allowances, house rent allowances, transport allowances, and family allowances paid by the employer, as well as remuneration for time not worked. Profit-sharing bonuses and earnings in kind are excluded. Data on gross cash earnings are collected according to sex. Information is also collected on employees' income tax, social insurance contributions and inhabitants tax.

Wage/salary rates

Not relevant.

Hours of work

Data are collected on hours actually worked, i.e. the total number of scheduled and non-scheduled working hours, excluding paid and unpaid rest periods. However, rest periods for mine workers in mining and waiting periods for workers in transportation are included. Scheduled working hours are the number of hours actually worked between the starting and ending hours of employment, as determined by the establishment's regulations. Included, but not separately identified are: time spent at the place of work during which no work is done but for which payment is made under a guaranteed employment contract; time spent at place of work for preparation of workplace, repairs, maintenance, preparation and cleaning of tools, preparation of receipts, time sheets and reports; inactive periods of time spent at the workplace due to mechanical or electrical breakdown, lack of supply of materials or client, bad weather, etc.; and study and training periods. Non-scheduled working hours are the number of hours actually worked when reporting to work early in the morning, when working overtime or on a day off, or when on emergency call-up. Excluded from hours actually worked are: Data on hours actually worked cover regular employees and are collected by sex. Data are also collected on the number of days worked, i.e. the total number of days during the survey period when an employee actually reports for work. Days the employee does not report for work are not counted, even though they may be paid holidays. However, if an employee works for at least one hour between midnight and noon, this is counted as a day worked.

International recommendations

The definition of earnings used in the survey conforms with the international guidelines on earnings, except that earnings in kind and profit-sharing bonuses are excluded. The definition of hours worked is slightly more restrictive than that contained in the international guidelines on hours actually worked, in that it excludes short paid rest periods.

Classifications

Industrial

The Japan Standard Industrial Classification (JSIC) is used, which distinguishes nine major groups. The JSIC is linked to the International Standard Industrial Classification of all economic activities (ISIC), Rev.2, 1968 with a few exceptions (for example, hotels are included under community, social and personal services in the JSIC).

Occupational

Not relevant.

Others

Employment data are classified by sex and type of employment (e.g. full- and part-time). Data on earnings and hours of work are classified by sex. All data are classified by size of establishment.

Sample size and design

Statistical unit

The sampling unit is the establishment, i.e. an economic unit engaged in the production of goods and/or the provision of services.

Survey universe / sample frame

The 1986 Establishment Census, carried out by the Statistics Bureau, Management and Coordination Agency. The Census covered about 1,490,000 establishments with five and more employees, and about 2,120,000 establishments with one to four employees. The frame is updated every five years, following each Establishment Census.

Sample design

The size and coverage of the sample varies with the type of survey, as follows: The sample is stratified by industry and size of establishment and the survey is designed keep the relative sampling error for contractual cash earnings within a fixed magnitude by industry and size of establishment. For the national survey, the required precision level is two per cent at the first-digit level of the JSIC, and three per cent at the two-digit level in manufacturing. For the prefectural survey, the required level of precision ranges between three and ten per cent. Establishments with 30 or more employees are selected from the Census list according to a stratified one stage sampling method. Establishments with 500 or more regular employees are included with certainty. The sampling fraction for the other strata varies by industry and size of establishment, and ranges from 1/2 to 1/576. New samples are selected at intervals of roughly three years. Establishments with 5 to 29 regular employees are selected by a stratified two-stage sampling method. In the first stage, about the 2,600 districts to be surveyed are selected. In the second stage, establishments are selected from the lists made for the selected districts. These establishments are divided into three groups and rotation sampling is used: the survey period is fixed for 18 months and the survey starts in January and July, respectively; one third of the sampled establishments is shifted every six months, and selected establishments remain in the sample for two consecutive survey rounds. The sample is completely renewed after 18 months.

Field work

Data collection

For the national survey, a questionnaire is sent by mail to the heads of selected establishments with 30 or more regular employees, who fill it in and return it to the Prefectural Office by the 10th of the following month. In establishments with 5 to 29 regular employees, enumerators interview the head of the establishments to fill in the questionnaires and submit them to the Prefectural Office by the 10th of the following month. The questionnaires are then inspected at the Prefectural Office and sent to the Ministry of Labour by the 15th of the following month. The procedure is similar for the prefectural survey, except that the questionnaires stay at the Prefectural Office. A permanent survey organisation is responsible for data collection, and if need be, additional enumerators are recruited temporarily.

Survey questionnaire

Instructions are provided along with the questionnaire and an interviewers' Manual is used by enumerators.

Substitution of sampling units

Sampling units which have gone out of business, changed address, refused to respond, etc., are not replaced.

Data processing and editing

Data are processed by computer. They are coded manually and edited by machine. In the case of missing or inconsistent data, enumerators contact the units by telephone or personal visit.

Types of estimates

Construction of indices

Wages, hours worked and regular employment indices, in establishments with 30 or more regular employees, by industry and establishment size, are computed each month, using as benchmark data the results of the latest Establishment Census (at present, the 1986 Census).

Weighting of sample results

Employment estimates are computed using the link-relative method. The current month's number of employees is estimated using the ratio of the previous month's estimates to the current month's surveyed employees. Hours of work and wages data are estimated using the sample data.

Adjustments

Non-response

None.

Other bias

None.

Use of benchmark data

Benchmark data are derived from the Establishment Census. Adjustments are made when new benchmark data are obtained.

Seasonal variations

The survey results are regularly adjusted for seasonal variations, using the Standard option of the Census Method (II, X-11).

Indicators of reliability of the estimates

Coverage of the sampling frame

The sampling frame (Establishment Census) covers about 80 per cent of the total number of regular employees.

Sampling error / sampling variance

For the national survey, the magnitude of the relative sampling error for contractual cash earnings is around two per cent at the first-digit level digit of classification, and three per cent at the second-digit level in manufacturing and at the other digit levels in other economic activities. For the prefectural survey, the relative sampling error for contractual cash earnings is five per cent in industry (three per cent in manufacturing) and seven per cent in major manufacturing groups.

Non-response rate

Not available.

Non-sampling errors

Not available.

Conformity with other sources

Not available.

Available series

The following tables and cross-tabulations are regularly prepared:

History of the survey

This survey was first conducted in July 1944 by the Statistics Bureau of the Cabinet, under the Order for a Labour Statistics Survey. With the promulgation of the Statistics Law after World War II, this survey became Designated Survey No.7 in August 1947. Responsibility for the survey was transferred to the Ministry of Labour when it was established in 1948 (although a complete transfer of the overall survey work did not take place until April 1951). Since then, the survey has undergone a number of major revisions, as follows:

Documentation

Ministry of Labour, Policy Planning and Research Department: Monthly Labour Survey (monthly, Tokyo). The survey results are published two months after the survey reference period. idem: Year Book (annual, ibid.). Detailed results which do not appear in national publications can be made available upon request, on print-outs or magnetic tapes. Methodological information on the survey can be obtained from the Ministry of Labour, Minister's Secretariat, Policy Planning and Research Department.

Confidentiality / Reliability criteria

The Monthly Labour Survey is conducted under the Statistics Law. According to Article 14 of this law, confidential data of individual corporations or other organizations which are covered by a survey approved by the Director-General of the Management and Coordination agency, cannot be released or published.

Other information

Data supplied to the ILO for publication

Annual averages of the number of paid employees in manufacturing, and of hours actually worked per week by employees (by sex) and monthly earnings of employees (by sex) in non-agricultural activities, manufacturing, mining and quarrying, construction, and transports, storage and communication, are published in the Yearbook of Labour Statistics (respectively in Tables 5A and 5B, 11 to 15 and 16 to 21). Monthly data on the number of employees, average weekly hours actually worked and monthly earnings of employees in non-agricultural activities and manufacturing (by sex) are published in the Bulletin of Labour Statistics

Other sources of data

Special Survey: In addition to the Monthly Labour Survey (national and prefectural surveys), a Special Survey is conducted each year, at the end of July or as of the last day of the payroll period, in establishments with a fixed last day. It covers all establishments with one to four regular employees. Its coverage of industries, persons, topics, etc. is similar to that of the national and prefectural surveys, and the concepts and definitions used are the same. The Special Survey covers about 77,425 establishments, selected from 4,750 districts. For data collection, the enumerators interview the head of the establishments to be surveyed, fill in the questionnaires and submit them to the Prefectural Office by the 10th of September. The questionnaires are then inspected at the Prefectural Office, and sent to the Ministry of Labour by the end of September. The results are estimates for all establishments with one to four regular employees. The relative sampling errors for contractual cash earnings are as follows: