Comparability — Canada [top]

The 1981 through 2011 samples report actual hours worked last week, including in unpaid productive labor. The 1991 sample only includes employed persons in the universe; the other samples have persons who worked zero hours.

The Canada samples 1981-2001 are top coded at 100+ hours and Canada 2011 is top coded at 80+.

The 1971 sample reports information on usual hours worked and is integrated in HRSUSUAL2 and HRSWORK2.

Comparability — Dominican Republic [top]

The 1981 records the actual number of hours worked at all jobs. If the person did not work due to illness or other reason, but still held a job, the usual number of hours worked per week was recorded. The 1981 sample does not report the number of hours under 10 and is top coded at 60 hours.

Comparability — Ecuador [top]

For the 1990, 2001 and 2010 samples, the question asked for the number of hours the person actually worked the week before the census at their main or principal job. If they were absent from work they were to report the number of hours for the last week that they worked.

Comparability — El Salvador [top]

Both samples record the number of hours the person worked last week at all jobs. If the person was temporarily absent (worked no hours) during the week prior, the number of usual hours worked per week was recorded. For 2007, the experienced unemployed population is included with responses indicating usual hours worked per week in the previous job.

Comparability — Ghana [top]

The 2000 sample reports actual hours worked for pay, profit, or family gain 7 days before the census night. The 2000 sample is top coded at 84 hours.

Comparability — Greece [top]

The 1971, 1981, 1991, 2001, and 2011 samples ask about the actual number of hours worked the previous week at any job. There are a large number of cases with zero hours worked.

Additionally, the 2001 and 2011 samples have a different variable that indicates usual number of hours worked in a typical week which is integrated into HRSUSUAL2.

The data for Greece 1971 and 1981 are top-coded at 90+ hours. The 1991 sample is top-coded at 97+ hours, the 2001 sample at 98+ and the 2011 at 70+ hours.

Comparability — Honduras [top]

In 1988, the question asked for the number of actual number of hours the person worked during the week before the census. The data for Honduras 1988 is top-coded at 69+ hours.

Comparability — Indonesia [top]

The 1976 sample reports the hours worked at a primary occupation in the past week.

The 1980, 1985, 1990, and 1995 samples report the number of hours worked during the previous week. The samples report actual hours during the past week. Persons who had a job but were temporarily absent the week preceding the census are coded as "0" hours worked. All four samples included here, in addition to the 1976 sample, contain data on hours worked in the primary occupation, found in HRSMAIN.

Comparability — Italy [top]

The 2001 census indicate the actual number of hours worked in a primary job the previous week. Responses were top-coded at 71 hours. Furthermore, if the respondent was absent from work the previous week due to vacation, illness, or some other form of leave, they were instructed to refer to their usual hours worked. This is in contrast to the 2011 census which asked about usual hours worked and is integrated in HRSUSUAL2.

Comparability — Jamaica [top]

All samples record the hours of employment completed in the week prior to the census. The universes vary slightly across samples with respect to whether or not group quarter residents were asked to provide work-related information.

Comparability — Jordan [top]

The 2004 sample reports the number of hours worked in the previous week at all jobs.

Comparability — Malaysia [top]

The 1991 sample reports the actual number of hours work during the last 7 days.

Comparability — Mauritius [top]

The 1990-2011 samples record the actual number of hours worked by persons age 12+ at all jobs during the week before the census. In the 1990 and 2000 samples, responses indicating less than 1 hour of work are coded as "none".

Comparability — Mexico [top]

All samples report actual hours worked in the previous week. The 1990, 2000, and 2010 Mexican samples all have numerous cases with implausibly high numbers of work hours. They are included in the top-code of 140 or more hours.

The 1990, 2000 and 2010 samples code persons that had a job but were temporarily away as "0" hours worked. The 1995 sample asks for the usual number of hours worked for these persons who were temporarily away.

Comparability — Nicaragua [top]

The samples report actual hours worked in the last week that the person worked at all jobs. For persons temporarily absent from work or the experienced unemployed population, a usual number of hours worked per week is reported.

Comparability — Pakistan [top]

The sample reports the number of actual hours worked last week. If the person had a job but did not work last week, the enumerator will report "0" hours worked.

Comparability — Palestine [top]

The variable is only available for the 2017 sample, in which the data are derived from a question about employment status. The sample records the number of hours worked at all jobs during the week before the census.

The intervals provided in the 2017 sample are not entirely compatible with those listed in HRSWORK2: the sample provides intervals of 15 to 34 hours (coded to 15 to 29) and 35 hours or more (coded to 30 to 39). The original intervals are available in the corresponding unharmonized source variable.

Comparability — Panama [top]

The 1970 sample excludes self-employed persons from the universe. The samples report actual hours worked last week.

Comparability — Puerto Rico [top]

The data for 1970 report actual hours worked in the previous week at all jobs.

The 1980-2005 samples record usual hours worked per week during the previous year, and these sample are integrated instead in HRSUSUAL2.

Comparability — Romania [top]

The 2002 sample refers to the number of hours done during the census reference week at a main job. Persons who had a job but were absent from it were considered to have worked zero hours during the reference week.

Comparability — Saint Lucia [top]

Both samples report hours worked in the week preceding the census.

Comparability — South Africa [top]

The question asked for the number of hours the person actually worked the week before the census. If they were absent from work they were to report the usual number of hours they worked.

Comparability — Trinidad and Tobago [top]

The census question asks about hours of work in the week prior to the census.

The 1980, 1990, and 2000 hours of work responses are provided in intervals, which are roughly approximate to those in HRSWORK2. See details in the general comparability discussion. In all samples, responses indicating less than 1 hour of work were coded to "0 hours" per week.

Comparability — United States [top]

In 1960 and 1970 the data are actual hours worked in the week prior to the census. For 1980-2010, the data record usual hours worked per week the previous year; these samples are recorded in HRSUSUAL2.

Comparability — Venezuela [top]

The meaning of the question differs among the samples. In 1971 and 2001 the question asked for the number of hours the person actually worked last week across all jobs. In 1981 the question asked for the hours the person worked last week; if the person did not work last week, the number of usual hours was reported.

The Venezuela 1971 sample has some minor differences in how it categorizes hours worked: 15 to 30 and 31 to 39 hours.