The 1990s were characterized by a fairly consistent pattern of industry growth. Manufacturers benefited from surging exports, driven by free trade, booming demand for ICT exports and robust auto exports. More trade fuelled goods-handling industries. Business services were fuelled by rapid growth for computer services as well as outsourcing. But many of our traditional resource exports faltered, while the government sector downsized and construction and real estate remained in the doldrums.

These industry trends reverberated throughout the labour market. Job growth in urban areas was favoured by the ICT sector, which is largely located in large cities. The rapid growth of the ICT sector also drove higher demand for more highly educated and white-collar workers. Conversely, the slump in resources (including agriculture) and construction depressed growth in rural Canada and for the less educated, especially blue-collar workers.

However, since 2000 the economy has been subjected to a number of shocks. The collapse of demand for ICT goods, notably telecom equipment and fiber optics, punctured the stock market bubble in 2000. The ensuing downward trend in interest rates accelerated after the September 11 attacks, sparking a surge in household spending, especially for housing.

Meanwhile, the restoration of government surpluses allowed a re-investment in public services, notably hospitals. Finally, the increasing integration of China into the world economy sparked a surge in commodity prices, which helped send the loonie soaring.

The result has been a sea-change in the pattern of industry growth. This paper documents how labour demand since 2000 has been in many ways the reverse of the 1990s, shifting from high-tech to housing and resources. This has profound implications for the urban-rural distribution of jobs as well as the education required of workers, their occupation and even the size distribution of employers.

The changes in labour demand have been met by important shifts in the supply of labour. In particular, workers over 55 years have contributed one-third of all labour force growth in the last decade, reversing several decades of declining participation. Still, the recovery in mining and construction jobs sent their unemployment rates to record lows, allowing these industries to rejuvenate their labour force. As well, job prospects for youths in rural areas are now better than for their urban counterparts.

Demand

Employment growth last year was spearheaded by mining, construction and real estate, all with gains of about 5%.¹ These three industries have led job growth so far this decade with increases of 17%, 18% and 10%² respectively since 2000. The recovery in mining was led by a 43% leap for oil and gas. (Business services³, health care and utilities were the only other major industries to post double-digit gains since 2000).

The job increases in these three industries follow a decade in which they were among the weakest performers. Construction shed 1% of its employment in the 1990s, mining 20% and real estate 6%. These industries, along with agriculture and public administration, bore the brunt of job losses in the 1990s.

Figure 1

This turnaround reflects booming commodity and housing markets. Commodity prices recovered by one-third since 2000, led by energy.⁴ Residential investment rose nearly 50% over the same period, led by house sales.The resulting surge in mortgage demand drove a recovery for finance, especially banking, as well as manufacturers of non-metallic minerals.

Two government-related sectors swung from losses in the 1990s to gains in this decade. Public administration had borne the brunt of cutbacks designed to reduce government deficits last decade. Sharply increased hospital funding led to more hiring after virtually no growth in the 1990s.

Conversely, some of the fastest-growing industries in the 1990s have lagged so far this decade.⁵ The biggest reversals have been in manufacturing. Computers and electronic products led this turnaround, giving back all of the 45,000 jobs added in the 1990s. The auto industry (including parts) has stalled after growing by 69,000 in the previous decade.⁶

Figure 2

In retrospect, the rally in factory jobs in the 1990s went against the grain of its long-run trend. Figure 3 shows manufacturing’s share of all jobs peaked at 19% in 1980 and declined steadily to 14% in 1993. It recovered to 15.3% in 2000, with nearly two-thirds of the increase in the auto and ICT sectors. Since then, it has declined slowly to near the all-time low set ten years ago.

Figure 3

The boom and bust for ICT manufacturers was echoed in higher demand for designers of computer systems. These jobs surged by 182,000 in the 1990s, but growth has since come to a halt after the high-tech train wreck and the completion of the Y2K conversion.

Several trends related to free trade in North America in the 1990s have also petered out since 2000. The FTA ushered in a decade of rapid growth of truck transport, while increased trade in high-tech goods and tourism helped boost air transport and sight-seeing. The post-2000 slump in trade and tourism with the US has hamstrung growth in these industries. Clothing and textiles shed jobs at an accelerating rate after weathering the introduction of North American free trade in the 1990s with modest losses.

Table 1: Average Annual Job Growth by Industry*

Other ‘sure-fire’ new trends of the 1990s have not flourished since 2000. Temporary help agencies grew in the 1990s, reflecting a business model that emphasized flexibility and contracting-out. They have grown little since. Publishing as well as courier and messenger jobs also slumped after gains in the 1990s, possibly because communication via the Internet dampened demand.

While many industries have seen their fortunes reversed since the millenium, a few trends were common to the 1990s and 2000s. For some goods-producing industries, this meant more losses. Most notable was the ongoing erosion of farm jobs, both for crops and livestock (the latter began to contract long before the discovery of mad cow disease in 2003). Fishing and pulp and paper jobs continue to shrink.

A few services sustained growth over both decades. Most business services outside of computer services expanded steadily, especially security, building services and architecture and engineering. Demand for education rose in both periods, as did social assistance and health care outside of hospitals. In consumer services, gambling and accommodation and food saw solid gains in both decades, although the increase in retailing was three times faster this decade. Goods-handling industries (wholesale and transport) posted comparable gains, although they had to shift from handling continental trade in manufactured goods to imports from China and exports of natural resources.

The upturn led by housing and resources so far this decade and the slump in high-tech have resonated throughout the labour market: employment growth has been more evenly balanced between those with and without a university education and between white-collar and blue-collar occupations than in the 1990s. As well, the dominance of small establishments in job growth has been surpassed by large ones in the 2000s, while rural areas have caught up to urban areas.

From 1996 (when rural/urban data were first collected) to 2001, jobs in urban areas rose one-half faster than in rural areas (12.8% versus 8.7%). This was largely because ICT industries were concentrated in urban areas: they accounted for 4 out of every 10 jobs created in big cities in the 1990s.⁷

Since then, rural jobs have grown slightly faster (6.5%) than urban ones (6.2%). As a result, unemployment in rural Canada has fallen while rising in urban areas: since the peak of the high-tech boom in 2000, the gap between rural and urban unemployment rates has shrunk from 1.6% percentage points (8.1% versus 6.5%) to 0.5 points (7.6% versus 7.1%).⁸

The upturn in rural job opportunities was led by the resource sector. Non-farm resource jobs grew 14,000 in the last two years, all in rural areas. Meanwhile, manufacturing jobs in rural areas rose by 41,000 even as they stalled in urban areas. Pulp and paper and smelting and refining are the largest source of factory jobs in rural Canada, while the implosion of the ICT sector helped dampen urban jobs. Transportation jobs needed to haul natural resources also rose faster (+10%) in rural areas.

Figure 4

All goods-producing industries as well as transportation have an above-average share of jobs in rural areas (the average was 26%; primary industries alone are 60% located in rural areas). As well, construction has grown as fast in rural as in urban areas since 2001. Rural areas would have fared even better except for the continued loss of farm jobs.

Just as important is the demographic composition of the rural revival. The recovery has been led by a 4.4 point jump in the employment rate for youths aged 15 to 24, nearly twice as fast as the increase for adults. In fact, for the first time ever job opportunities for youths were better in rural areas than in cities, with employment rates of 58.7% and 57.9% respectively. The increase was slightly greater for young men than women in rural areas, with job gains of 11% in the last three years after none in the previous five years. Conversely, the levelling off of jobs in cities had a large impact on young people in the urban core, where job growth slowed to a crawl after a 15% jump from 1996 to 2001. As a result, the unemployment rate for youths in urban areas surpassed rural areas (13.5% compared with 12.8%) for the first time on record.

Figure 5

The distribution between blue-collar and white-collar jobs was also significantly affected by changes in industry growth. Despite steady losses in agriculture, blue-collar job growth accelerated from nil in the 1990s to 1.4% in the 2000s, thanks to the rebound in construction and mining jobs. But white-collar occupations also saw growth pick up, from 1.7% to 2.3%.

Figure 6

However, the type of white-collar jobs being added did change significantly, especially in terms of the education needed. Management jobs led growth in the 1990s with an overall 18% gain. These jobs have levelled off since 2000. Instead, growth has been driven by retail sales and clerical jobs (both up 15% in just four years), reflecting the acceleration in consumer spending. Other rapidly expanding areas include security services and restaurant jobs. None of these high-growth areas require the post-secondary education of most management jobs.⁹

Employment in large establishments has risen faster than in small ones since 2000, reversing the trend of the 1990s (when large firms raised productivity by shedding workers). The gap between the two began to widen in 2000, at the height of growth in ICT manufacturing. More recently, surging commodity exports favoured growth for large resource-based multinationals.

Figure 7

Supply

The new patterns of employment were reflected in unemployment. Mining posted its lowest-ever unemployment rate of just 4.5%. The recovery of mining jobs over the last two years also allowed a rejuvenation of its labour force, with youths aged 15 to 24 the fastest-growing segment. Similarly, the boom in housing sent unemployment in construction to a 30-year low of 8.6%, even as it too renewed its labour force by hiring more young people.

Unemployment in manufacturing has not increased over the last two years, despite job cuts as the dollar rose. This was due to an exodus by its labour force, mostly young people as job prospects dried up. Meanwhile, the labour force for factories continued to age rapidly, with the number of people over 55 years up almost 60% in the past decade.

Industries that cannot transfer their excess labour force to other sectors continue to be plagued by high unemployment. Forestry and fishing stood out with unemployment rates of 16% and 22% respectively, the only double-digit rates in the economy.

The aging of the post-war boomers continued to be the dominant demographic force in the labour market. The fastest-growing segment of the population was those aged 55 to 59 years, up 4.7% just last year. Since 1998, people aged 55 and over have grown in number by 19%, driven by a 28% gain for the 55 to 64 group due to the arrival of the leading edge of the boomer generation. Conversely, the number of prime-aged people between 25 and 44 edged down in absolute terms for a seventh straight year, for a total drop of 2.3%, while youths aged 15 to 24 grew 7% since 1997.

But while boomers were moving into age groups traditionally associated with more retirement, evidence suggested that their approach to old-age would be quite different from previous generations. For example, the labour force participation rate of people 55 to 59 years rose to an all-time high of 67.6%. Their participation rate had held steady at about 60% from 1976 to 1998, before beginning a steady rise as boomers entered this age bracket.

The increased participation rates for aging boomers is part of a marked trend towards older people aged 55 to 69 remaining in the labour force.¹⁰ Figure 8 shows the participation rate for this group has shot up from a low of 36% in 1996 to a record 47% last year. The largest gain was for the 60 to 64 group (from 33% to 43.7%), while 65 to 69 year olds rose from 11.5% to 16.2%.

Figure 8

Overall, the increase in both the number of people 55 and over and their participation rate meant that they contributed over one-third of the 2.35 million increase in Canada’s labour force since 1996. Interestingly, 1996 was the year the first boomer turned 50.

No one socio-economic factor explains the rising presence of older people in the labour force. It may reflect changing attitudes to work (one no longer sees references to ‘Freedom 55’) and expected life spans of older people, and more flexibility offered by employers who otherwise would be facing a shrinking labour force.¹¹ The increase started well before the stock market crash in 2001 put a dent in many private pension plans, although it did accelerate slightly after. The boomer culture alone does not explain even faster increases in the labour force for people older than them. Neither does the higher educational attainment of boomers, as their growing attachment to the labour force is evident irrespective of educational attainment¹².

Staying in the labour force past the age of 55 reverses almost three decades of declines for older men (the participation rate for older women was quite steady at 17% until a surge in the last five years to 24%). One factor pushing older men out of the labour force in the 1990s was losses for this group in primary, construction, manufacturing and public administration jobs (which together employed nearly half of all these men in 1990). The resumption of growth in these areas was accompanied by labour force growth for older men in all these industries; after declines in the 1990s, they contributed one-third of the total increase for men over 55.

Overall, 67.6% of Canadians 15 years and over were in the labour force, just beating the record set the year before. The participation rates of all demographic groups have risen in recent years, not just older people. The long, gradual slide for prime-aged men from 96% in 1979 to below 92% has reversed slightly in recent years. And the sharp exodus of young people after the recession in 1990 has been largely recouped with an increase from 61.5% in 1997 to 67% last year.

Finally, the increasing presence of adult women aged 25 to 54 in the labour force continued unchecked as it has since the current survey began in 1976 (barring major recessions), rising to 81.6% last year. While still 10 percentage points below the rate of their male counterparts, this was the smallest gap on record (it was over 40 points in 1976). The steady increase for adult women in Canada contrasts sharply with a drop in the US since 1997, partly because of their 23% increase in the child-bearing rate for women aged 30 to 44. While older Americans are also participating more in the labour force, the drop for adult women and youths pushed their overall participation rate to a 16-year low of 65.8% early in 2005.