CBC Radio tonight reported on a statscan study on the redistribution of
overtime hours published in Perspectives on Labour and Income and featured
in the Dec. 10 The Daily. I had a call from the Edmonton Journal asking for
comment on the statscan article and, although I had seen the article at that
time, I expressed doubts about the estimates. Having now seen the article,
I'm appalled at the cavalier shallowness of the analysis.
I'll just point out the most flagrant distortion. The authors estimate the
number of hours that workers would voluntarily give up as "5%". They claim
to base this on the 1995 Survey of Work Arrangements finding that "only 5%
of employees who regularly work overtime would agree to have their hours
reduced for less pay." However "hours" and "employees" are two very
different kettles of fish.
Going back to the 1995 Survey, there were substantial variations in the
willingness to have hours reduced between provinces, sexes and job
characteristics. Workers in provinces with high unemployment tended to be
less willing to have their hours reduced. Women were more eager than men to
work fewer hours. And interest in reducing hours rose significantly as the
number of usual hours in a worker's main job increased.
Given all these variations between employee preferences, the authors' glib
extrapolation of "hours" from "employees" is dubious. The results of a
recent Southam News COMPAS poll throws more doubt on the authors' calculations.
Forty-two per cent of respondents to that poll said they were unwilling to
take on extra work at a marginal tax rate of 50 cents on every extra dollar
of income, 15% said they were "definitely willing" to take on extra work and
another 26% were "probably willing." It just so happens that the marginal
tax burden for someone in the middle tax bracket comes pretty close to 50%,
including provincial income and sales tax, GST, and surtaxes.
The 1995 survey of work arrangements asked people whether they would like to
work less or more hours at the same hourly rate. Given a marginal tax rate
of 50% and substantial fringe benefits the "same hourly rate" stipulation is
misleading. Even at a time and a half rate, overtime yields many workers
_less_ than their regular hourly rate.
The 1995 survey also made no mention that the reduction in hours might be
used to reduce unemployment. A similar survey done in 1985 mentioned work
sharing as the reason for reducing hours of work. Responding to that
context, a substantially higher percentage of workers indicated a preference
for reduced hours. In other words, the 1995 survey didn't want respondents
answers to be "skewed" by suggestions of social conscience.
Here is the Statscan article from The Daily:
The redistribution of overtime hours
Up to 169,000 jobs might have been created in November 1995 had it been
possible to convert all the paid overtime accumulated that month into
full-time jobs. However, if overtime had been cut only for those workers who
agreed to such a measure, the potential for job creation would have been
less than 10,000.
Results based on the Survey of Work Arrangements show that in November 1995
paid workers regularly worked an average 6.8 million hours of paid overtime
weekly. Assuming this overtime could be converted into jobs of 40.5 hours a
week, which is the average number of weekly hours reported by full-time
workers, the total hours could have been converted into 169,000 new jobs. It
would have resulted in a 13.2% reduction in the number of unemployed
workers, and the unemployment rate in November 1995 would have dropped from
8.7% to 7.5%.
The creation of these 169,000 hypothetical jobs rests on several
assumptions. For example, a perfect match would be necessary between the
qualifications of the unemployed and those required in the newly created
jobs, as would full mobility of the unemployed between provinces.
Many of these new hypothetical jobs call for a well-qualified, often highly
educated labour force. Depending on the region, managerial, administrative
and professional positions represent nearly a quarter of the overall
potential, and those in construction, a tenth. One-third of potential new
jobs would also come from processing, machining and product fabricating
occupations, which may also require fairly specialized workers.
To examine the effect of these issues, a match was made between the skills
required for the new positions and those offered by the unemployed. The
province in which unemployed workers lived and the location of the
hypothetical jobs were also taken into consideration. With these constraints
taken into account, the potential 169,000 jobs would then drop to 93,000.
The unemployment rate would shift from 8.7% to 8.0%, instead of the 7.5%
suggested by the most optimistic scenario.
However, the Survey of Work Arrangements also found that only 5% of
employees who regularly work overtime would agree to have their hours
reduced for less pay. If overtime had been cut only for those workers who
agreed to such a measure, fewer than 10,000 full-time jobs would be created,
even with no matching by occupation or location. Consequently, the number of
unemployed workers would have been reduced by only a fraction of a percent
and the unemployment rate would have remained unchanged.
Over 1.9 million out of a total 11.4 million paid workers (17%) worked
overtime in a typical week from January to September 1997. Over half (53%)
were not paid or otherwise compensated for any of their overtime, while 45%
were paid for all their extra hours. The average number of overtime hours
worked per week was 8.8 for those who were paid and 9.5 for those who were
not compensated.
Extra hours are more common in some industries and occupations than others.
Overtimers are most prevalent in the communication industries (26% of paid
workers), educational services (24%), consumer durables manufacturing (24%)
and mining (23%). Among the occupational groups studied, the incidence of
overtime was highest in teaching (28%), management and administration (28%)
and among professionals in the natural sciences, engineering and mathematics
(27%).
Full-time workers exhibit much higher overtime rates than part-timers. In
1997, only 8% of all part-time employees put in excess hours, compared with
19% of their full-time counterparts. Although the average usual weekly hours
of part-time workers were much lower than those of full-timers (17 hours
versus 40), overtime hours were similar: part-timers averaged 8.7 hours
while full-timers logged an average 9.3 hours.
Weekly earnings (excluding overtime earnings) were greater, on average,
among overtimers: $734 versus $530 for non-overtime workers. And they were
significantly higher among those whose extra hours went unpaid: $840. This
disparity likely re ects the fact that unpaid overtime is more common among
full-time workers and employees in managerial, administrative and
professional occupations, which are associated with higher pay.
Only 12% of those employed on a temporary or casual basis reported overtime,
compared with 17% of those in permanent jobs. Workers in non-permanent jobs
were somewhat more likely to be paid for extra hours, however (55% versus 44%).
Overtime was somewhat more prevalent in Western Canada than in the East:
rates were highest in Alberta (21%) and lowest in Prince Edward Island
(13%). Relatively high concentrations of overtimers (both paid and unpaid)
were found in the following metropolitan areas: Calgary, Kitchener-Waterloo,
Ottawa-Hull, Saskatoon, Edmonton, Windsor, Victoria, Winnipeg and Hamilton,
where about one-fifth of employees worked overtime. Paid overtime hours were
longest in Windsor, Edmonton and Halifax (ranging from 9.7 to 10.1 hours)
and shortest in Victoria (6.3) and Vancouver (7.1).
The Winter 1997 issue of Perspectives on labour and income (75-001-XPE,
$18/$58) is now available. See How to order publications. For further
information on "The redistribution of overtime hours," contact Diane
Galarneau (613-951-4626), and on "Working overtime in today's labour
market," contact Doreen Duchesne (613-951-6379), Labour and Household
Surveys Analysis Division.
Regards,
Tom Walker
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
knoW Ware Communications
Vancouver, B.C., CANADA
[EMAIL PROTECTED]
(604) 688-8296
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
The TimeWork Web: http://www.vcn.bc.ca/timework/
