No Shortage of Opportunity featured image
Skills and Labour Market Policy

No Shortage of Opportunity

Policy Ideas to Strengthen Canada's Labour Market in the Coming Decade

Cliff Halliwell November 8, 2013

Canada is in the midst of two important transitions, which could affect the functioning of the labour market for years to come. The first is the slow but steady return to full employment after the recession of 2008 and 2009, and the second is the long-expected and unprecedented transition of baby boomers from work to retirement. In this study, Cliff Halliwell (former director general of policy research at Human Resources and Skills Development Canada) draws a broad and somewhat contrarian portrait of the labour market issues Canada is likely to encounter over the next decade and provides an updated framework to improve policy effectiveness in this new context.

Despite widespread concerns from employers about impending labour and skill shortages as baby boomers retire, Halliwell finds that the long-term prospects for Canada’s labour supply and demand are balanced, and that ongoing shortages are unlikely. The key question, however, is: how do we respond to an environment of much slower, but still rising, labour force growth?

For Halliwell, this is an opportunity to achieve the virtuous circle of rising wages and productivity that has long eluded Canada. Rather than simply looking for ways to increase labour supply, employers and policy-makers need to focus on improving the way we develop, match and use the skills of Canadians – those who are unemployed as well as those who are employed.

Halliwell’s analysis calls into question recent conventional wisdom, according to which all future labour force growth will have to come from immigration. This perspective, he argues, ignores the large pool of young Canadians who will continue to be available to the labour market between now and 2030. Over this period the number of young people ages 25-29 will average approximately 2.5million, far greater than the annual intake of immigrants (approximately 250,000).

While slowing labour force growth will likely put upward pressure on wages and intensify efforts to increase productivity, stronger policy is also needed. He recommends these measures:

  • strengthening and formalizing the “second-chance” system to help Canadians upgrade their education or skills, or get better jobs over the course of their careers;
  • limiting temporary foreign worker programs to jobs that are truly “temporary”;
  • reducing immigrant intake during recessions, and increasing investments in settlement supports to enhance the language and occupational skills of newcomers;
  • modernizing labour market measures to discourage the frequent use of employment insurance, better support the needs of long-tenured workers, and more effectively target the skills development and training needs of the unemployed; and
  • improving labour market information and increasing the flexibility of post-secondary education so that students can more effectively select fields of study and have better access to trades, apprenticeships and occupations in demand.

What will Canada’s labour market look like over the next decade? Now is an important time to be asking this question. Currently, the Canadian economy is still in the midst of two important transitions, the return to full employment after the recession of 2008 and 2009, and the more profound transition into retirement of baby boomers, the first wave of whom turned 65 in 2011. This demographic tipping point represents the beginning of a historic change in the composition of the Canadian workforce. Other longer-run trends also herald important new opportunities and challenges over the next decade as developments in technology and ongoing globalization continue to work their way through the industrial and occupational structure. How Canada’s labour market evolves over this period will have major implications for all Canadians and, in turn, for the demands that will be placed on public policy in the years ahead.

Will Canada face labour shortages? What will be the return on investment in higher education? Will workers enjoy a period of real wage growth supported by rising labour productivity? For these and other questions at the forefront of public discussion, a key issue is whether the labour market dynamics of the next decade will be significantly different than in the past.

Of course, any exercise of forecasting economic outcomes far into the future is fraught with risk and potentially great error. No analysts (those in government included) can be expected to have precise answers about what the world will look like in a decade. But adequate planning requires thinking now — in a coherent and structured fashion — about the medium- and longer-term implications of the broader structural factors at play. As the federal government notes in its recent analysis of Canada’s aging population: “Canada must plan and act now to prepare for the challenge of an ageing population by putting in place policies to further strengthen the Canadian economy and public finances. Early intervention will avoid the need to take drastic or inequitable actions in the future” (Finance Canada 2012).

With that in mind, the goal of this study is to paint a broad portrait of the labour market issues and the challenges that policy-makers will likely encounter as Canada approaches the third decade of the twenty-first century. I aim to examine as comprehensively as possible the environment that lies ahead and to provide a framework to inform labour market policy in Canada over the medium- and long-term horizon.

I begin by discussing the labour market policy issues that were top of mind prior to the onset of the 2008-09 recession and then briefly highlighting the new issues that are the product of that recession. I then provide an overview of the demographic and labour market indicators that can help us best assess future labour market prospects. I finish by considering a range of policy issues that might arise over the coming decade as labour force growth slows and we return to full employment. In this environment, how and to what extent we use the broader tool box of labour market policy levers — education, immigration, the retirement income system and the social safety net — will have a substantial effect on labour market outcomes. For these policies to be most effective, however, we must be careful not to succumb to the many myths recurring in the public discourse today.

The Labour Market Issues We Expected to Face Prerecession

Six years ago, Canada had a strongly performing labour market, the product of over a decade of economic expansion. Two thousand and eight was the fourth consecutive year of unemployment below 7 percent. Between 1976 (when the current Labour Force Survey began) and 2004, unemployment had dropped below 7 percent only once. The conventional wisdom had been that it could not drop below a structural rate of around that rate. While unemployment rates by province were also (largely) quite low by historical standards, the challenge was to address continuing disparities across the country, with rates ranging from 3.4 percent in Alberta to 13.2 percent in Newfoundland and Labrador. This raised, once again, the question of why jobs in some regions could not be easily filled while unemployment remained high, even in double digits, in others.

In response to these dynamics policy attention turned to further reducing frictions in the labour market and structural impediments to adjustment, by means of better information (Advisory Panel on Labour Market Information 2009), help with job matching and a greater emphasis on active labour market measures. With the 2007 round of negotiations on labour market agreements, the federal government completed the devolution of these active measures to the provinces, based on the rationale that they were closer to the local problems. The only avenue that was not seriously contemplated was changing employment insurance (EI), given the minority status of the federal government of the day.

Tightening labour markets

By 2006 and 2007, attention was also turning to labour market “shortages,” widely reported by employers in many industries. Even in regions where unemployment rates were high, employers were complaining about their difficulties in filling job vacancies, which suggested these high unemployment rates were structural. Moreover, many were predicting “shortages” counting in the millions by the middle of the next decade. Clearly, labour markets at that time had much less slack than had been the case for decades, but was this just a case of full employment after years of excess labour supply or one of true shortages?

In fact, it was both. For the economy as a whole, 2006 through 2008 were years of operating above potential. Indeed, for 2008, the Bank of Canada now estimates that real gross domestic product (GDP) was, on average, 1.8 percent above potential (2012). This would have been as true of labour markets as of aggregate GDP. This pressure, however, was not going to last. It was the Bank’s stated goal to slow growth so that demand would be in line with supply. From the point of view of employers, even an economy operating at its potential was nonetheless unusual. Over the course of several decades they had become accustomed to having at their disposal a “reserve army of the unemployed” that made it easy for them to find workers when needed, even for less attractive jobs.

Tempering excess demand would have alleviated the aggregate, national labour constraint. However, even when the economy is at full employment, individual submarkets (differentiated by geography, education, skills and occupation) might not be. This can result in widespread shortages in some areas, offset by surpluses elsewhere. Indeed, this is what the Canadian Occupational Projection System (COPS) — which Human Resources and Skills Development Canada (HRSDC)1 uses to track and simulate labour demand and supply — showed at the time. There was excess demand in several occupations nationally, especially in health occupations, the oil and gas sector, and managers in public administration; and excess supply in others (primarily in lower-skilled occupations in shrinking sectors) (Lapointe et al. 2006).

To help address these concerns about labour shortages, the federal government embarked on a substantial expansion and streamlining of the Temporary Foreign Worker Program. Immigrant inflows were increased sharply. For 2008, temporary foreign workers (TFWs) added nearly 0.4 percentage points to employment growth, and represented more than 2 percent of total employment, almost double their share in 2003.

Attention also turned to skills deficiencies. Skills are developed in three ways. The first and most important is the formal education people get early in their life. The second is work experience: the informal learning that happens while working. Evidence suggests that the rate of return on a year of work experience is considerably less than that on a year of formal education; however, the total return accumulates over decades (Lin, Chen and Qiu 2009; Lin and Chen 2010), as earnings in many occupations increase with experience. The third is formal on-the-job training. The evidence on the value of training is, at best, mixed, possibly because the amount people receive is actually quite low — often equal to only a few days a year — and paltry relative to other sources of skills acquisition (Percival, Cozzarin and Formaneck 2013).

On the whole, the human capital of Canada’s workforce during the mid-2000s ought never to have been better. Yet as employers reached down deeper into the labour pool, they started reporting concerns about fundamental skills deficiencies. Some pertained to a lack of technical skills, often in STEM (science, technology, engineering and math) disciplines or the trades. Many pertained to what employers call “soft skills,” such as communications, the ability to work in teams, commitment to the job and so on. Concern was also expressed, however, about basic literacy and numeracy skills. These are important because, beyond formal educational attainment, literacy and numeracy are among the most significant factors explaining economic growth (Green and Riddell 2001). By the middle of the past decade, provincial education ministries began to focus more on these skills in light of new results from PISA (Program for International Student Achievement) surveys. These showed two key things: first, that young Canadians (PISA captures data for 15-year-olds) had very good skills by international standards and, second, that there was considerable variation among provinces (Bussiere, Cartwright and Knighton 2004).

International data comparing adult literacy results across countries also began to emerge, providing important new and more disconcerting insights about the skills of Canada’s workforce (Statistics Canada 2005). The data, based on the International Adult Literacy and Life Skills surveys, spearheaded by the Organisation for Economic Co-operation and Development (OECD), revealed that, despite the fact that Canadian students at age 15 were doing well by international standards (as PISA results suggested), results for adult Canadians fell well behind those in many other countries (Sloat and Willms 2000). Somehow Canada’s rating slipped from a good to a not-so-good performance in the working age population. The results also showed that those with lesser skills were losing them faster over time (Willms and Murray 2007; Green and Riddell 2013). As we know, deficient literacy skills are increasingly an impediment to individuals’ performance in the labour market. Moreover, because formal training requires a basic level of literacy, poor literacy skills make it more difficult and costly to improve other skills. Here one should underscore a key aspect of reading comprehension: the ability to learn without being taught.

Another growing labour market concern was that compared with earlier cohorts of immigrants, the earnings profiles of new waves of immigrants were not converging toward those of native-born Canadians, even as the effects of the recession during the early 1990s receded (Worswick 2004; Picot and Sweetman 2005). This was especially true for the most educated immigrants.

Spurred by these poor outcomes, and also by evidence from other countries, such as Australia, attention turned to tightening selection criteria — in particular, official-language skills — engaging the provinces more in the selection process and increasing settlement budgets.

Labour force growth slowdown

By 2007 and 2008, labour force growth was still relatively healthy: the average annual growth rate from 2000 through 2008 had been 1.7 percent. This was actually above expectations, reflecting both higher immigration (more permanent immigrants and TFWs) and increased participation rates among older workers after decades of decline. At the same time, though, macroeconomic projections uniformly anticipated a slowing of growth due to population aging. Both the working-age population and the active labour force were projected to grow more slowly. This in turn was expected to lead to lower employment rates and slower growth of aggregate real GDP and real GDP per person (see Denton and Spencer 2009). These projections highlighted clearly why it was more important than ever to improve Canada’s sluggish labour productivity.2

With shortages already perceived as widespread by many employer and stakeholder groups, attention turned to where to find additional labour. The obvious starting points were underemployed and disadvantaged groups, such as persons with disabilities (PWDs), Aboriginal people, recent immigrants and older workers. With respect to the first two groups, some unique factors should be emphasized. Thanks to a healthier labour market in the early and mid-2000s, both PWDs and Aboriginal people had already experienced significant improvements in employment and earnings outcomes. Between 1999 and 2006, for example, the employment rate of PWDs rose by nine percentage points to 53.4 percent (Prince 2013); for Aboriginal people, the increase was nearly five percentage points (Statistics Canada 2007). In both groups, however, the remaining unemployed continue to face significant challenges to labour market integration (such as lack of education and long-term disability), which make it unlikely that they could attain employment ratios comparable to the ratio of the labour force as a whole in the short or medium term.

This left older workers, who, by and large, already had a track record of employability and productivity. Simulations using HRSDC’s Computable General Equilibrium (CGE) model showed that getting older workers to work even just a few additional years added considerably to GDP per capita (Fougère et al. 2009).3 These simulations, however, were “what if” scenarios, assuming older workers worked longer; the simulations offered no road map to what policy instruments would entice them to do so. Subsequent HRSDC analysis of potential instruments found only modest effects in aggregate,4 since such incentives increase the relative “price” of work, with sometimes-offsetting effects. Although the “substitution” effect makes work more financially rewarding and encourages more effort, the “income” or “wealth” effect means less work effort is needed to attain the level of wealth required to retire. The fact that the substitution and income effects operate in opposite directions dampens the net impact. Even simply lowering the income taxation of older workers produced only modest gains in labour supply in the simulations (Fougère, Harvey and Rainville 2012).

Another option that was examined is facilitating partial retirement, which would enable people to work longer by not having to choose between too much work and none at all. Here again, microsimulations (and simple reflection) tend to suggest partial retirement would produce only modest gains in aggregate labour supply: some would work longer part-time, but others who start to work part-time would retire earlier than they would have otherwise (Carr 2008). One factor that seemed to matter more, and where evidence was available from the collapse of the high-tech equities boom, was the wealth effect resulting from changes in financial assets. In other words, people are less likely to retire early if their assets intended for retirement decline in value or are less secure, as happened to many during the recent recession.

In the end, this body of research shows that the greatest disincentives to working longer are the structures of many defined-benefit (DB) pension plans (Fougère et al. 2009). DB plans, unlike alternative defined-contribution (DC) plans or personal investments, provide relative certainty about retirement income. They also make working beyond the plan’s normal retirement age not very lucrative — although nothing stops a worker in a DB plan from retiring and then taking another job with another employer. As these are private plans, however, this would not be easy to change.

That said, efforts were made to remove disincentives to working longer. Prominent examples include changes in the actuarial adjustment factors in the Canada and Quebec Pension Plans in 2010, the repeal of mandatory retirement legislation in both federally and provincially regulated employment, and new income tax measures to facilitate partial retirement. Moreover, the participation rates of older workers, which had been declining for several decades, actually began to increase during the 2000s, even before these changes were implemented (Halliwell 2009; Carrière and Galarneau 2011).

Current Challenges Postrecession

With the 2008-09 recession, the short-term policy agenda changed completely as the focus shifted toward dampening the effects of substantial job losses, although initial fears of an unprecedented recession turned out to be pessimistic. Real GDP and employment levels returned to prerecession peaks within just a few quarters, as they had after previous recessions, for two reasons. First, the output recession was somewhat less severe in Canada than in the United States, as Canada did not have the imbalances that triggered and compounded the US recession.5 Second, the impact on the labour market was also different in the two countries, as shown in figure 1. In the United States, employment plunged dramatically with the decline in real GDP, so much so that, along with increases in capital stock per employee, GDP per hour worked (labour productivity) actually increased quite substantially during the recession and subsequent recovery. In Canada, the drop in hours worked was far more significant than the decline in employment, while GDP per hour worked remained relatively flat. Since the recession, the difference in employment growth between Canada and the United States might also be attributable to Canada’s generally weak productivity performance.

Halliwell figure 1

What takes much longer to recover — typically years — is unemployment. In the recessions of the early 1980s and early 1990s, the unemployment rate was still well above pre-recession levels a full eight years after their onset. Canada might now be seeing an encore performance. As of September 2013, the seasonally adjusted unemployment rate stood at 6.9 percent, one full percentage point above where it was at its lowest point (October 2007) just before the recession of 2008 and 2009. More than one-third of the increase in the unemployment rate between October 2007 and its peak in August 2009 (8.7 percent) has yet to be reversed. The reason for this is simple: although employment is up by 858,000 since October 2007, the labour force has increased by approximately 1.1 million, fuelled by a gain of 2.1 million in the population over the age of 15 in the same period. This has simply overwhelmed the modest decline in labour force participation. (For an overview of Canada’s labour market as of 2012, see box 1.)

Halliwell box1

A poor labour market for new entrants

A slow recovery in labour market conditions has adverse consequences, especially for new entrants, who are predominantly the young and recent immigrants. Those who enter the job market in good times do much better than those who enter in bad times, not just in the short term but also in the long term. Recessions are a time of low hiring and high layoffs, and the adverse effects fall particularly on younger workers and the less skilled (Oreopoulos, von Wachter and Heisz 2008). Many of those affected have poorer employment and earnings outcomes, even years afterwards, as a result of these “scarring” effects.

Youth have one advantage in a weak labour market: they can choose to further their education more easily than can older people with family and debt obligations. In essence, post-secondary education (PSE) is cheaper as the opportunity cost of forgone labour income falls. As a result, recessions stimulate PSE enrolment.

The bigger issue is those ages 25 to 29 who enter the labour force after completing their post-secondary education. This group grew at a 1.6 percent annual rate between 2008 and 2012. Meanwhile, their employment numbers grew at a rate of only 1.2 percent annually, with full-time employment up at half that pace. As of the end of 2012, the unemployment rate for 25-to-29-year-olds remained 1.4 percentage points above the prerecession peak, somewhat higher than the rate increase for the total labour force. This suggests that the fallout of the recession has been significant for these young potential workers, effectively making it harder for them to embark on their envisaged careers.

But has the aftermath of the most recent recession been as bad as that of previous recessions for the 25-to-29 age group? The answer is no: four years out, the fall in the employment ratio and the rise in the unemployment rate are smaller than occurred after previous recessions.

Immigrants are also subject to such “scarring” effects. Employment growth among immigrants aged 25 to 54, at 2.6 percent annually between 2008 and 2012, was strong but short of that needed to absorb the 3.1 percent annual growth in the labour force for this immigrant cohort, which swelled the total number of unemployed immigrants by 7 percent annually.6

In the end, there is no surprise about these results. The inflow of immigrants remained strong through the downturn and initial recovery and expansion, but immigrants ages 25 to 54 have had worse labour market outcomes than their Canadian-born counterparts. The unemployment rate for immigrants was 2.2 percentage points higher than that for the Canadian-born in 2008, and 2.5 percentage points higher in 2012;7 thus, the unemployment gap between the two groups has widened somewhat since the recession.

Selecting and integrating immigrants

There is a large body of literature analyzing which characteristics of prospective immigrants are associated with better labour market outcomes over time. These are, by order of importance: arranged employment on arrival, official-language skills and work experience in Canada. Age (that is, youth) and education also have positive effects on earnings (Begin, Goyette and Riddell 2010; Ferrer, Picot and Riddell 2012; Picot and Sweetman 2012), although in the case of the latter the interaction with language skills is a key factor. In particular, Bonikoska, Green and Riddell (2008) show that differences in official-language literacy and in numeracy skills explain a good part of the earnings gap between immigrants and their Canadian-born counterparts. It is noteworthy that although immigrants’ educational attainment has risen over the past several decades, their economic outcomes have progressively worsened (Picot and Sweetman 2012).8

halliwell figure 2

halliwell figure 3

For several years now, there have been changes in immigration policy in response to this research. Language requirements have been enhanced, and greater weight has been given to applicants who are younger, and who have work experience in Canada and skills in high-demand occupations. In particular, those with prior Canadian work experience (such as returning temporary foreign workers) or who have attended Canadian post-secondary institutions have been given preferential treatment. The expansion of provincial nominee programs and the creation of the Canadian Experience Class for highly skilled professionals now provide accelerated pathways to permanent residence for applicants who have specific job offers.

That being said, Canada still selects only a minority of its immigrants based on economic criteria. Although approximately two-thirds of immigrants admitted under the economic class  category (figure 2), only the principal applicants (PAs) are selected based on the point system, and they represent just over a quarter of total immigrants (figure 3). Each economic class PA brings with him or her approximately one and a half spouses and dependants, whose education and skills are not assessed, but who are nevertheless included in this immigrant class.

Determining how many foreign workers Canada “needs” right now

How to manage the inflow of immigrants relative to the overall business cycle is an important consideration for labour market policy. Adjusting this inflow in the event of a recession might allow for a faster reduction in unemployment during recovery. Before the recession of the early 1990s, immigrant inflows were reduced during downturns, but during that and the most recent recessions they were not. In case of the most recent recession, the federal government likely was constrained because of the need to deal with what was already an enormous backlog of applications. This policy, we now know, exacerbates the poor outcomes for immigrants who arrive in the midst of weak labour market conditions (Abbott and Beach 2011; Picot and Sweetman 2012). It likely also compounds the difficulties young Canadian labour market entrants have in finding suitable work.

What are the alternatives? A return to the old “tap on, tap off” policy would mean less immigration during recessions, but would it mean more afterwards — a case of cycling below and then above the longer-run track envisaged in policy? Cycling above might prove challenging, in terms of both processing and absorptive capacity. In that case, fewer immigrants during recessions could mean fewer immigrants over the long term. For the most recent recession, this issue is now largely moot, as the onset of the recession was almost five years ago. That said, now is not yet the time for increasing immigration.

halliwell figure 4

Also relevant is policy on TFWs. As illustrated in figure 4, Canada has seen a significant increase in the number of temporary workers admitted each year. Since 2004, the share of employment represented by TFWs in a given year (the stock of those newly admitted and those still here) has more than doubled, to 2.8 percent of employment in 2012. In other words, TFW employment has risen faster than employment for permanent residents.

When the next recession occurs, Canada should be more careful with TFW inflows and should also reduce its immigration intake (Picot and Sweetman 2012). Of course, the next recession will come as a surprise, as they all do. Once it comes, however, one should assume it will take years to return to full employment. Reducing the intake of immigrants, which cannot be done instantaneously, would help hasten the eventual recovery for the unemployed.

Longer-Term Prospects

The standard definition of the labour force is based on the notion of a working-age population, defined as the civilian, noninstitutionalized population ages 15 and older (First Nations people living on reserve are excluded). Implicit in this definition is the notion that being 14 or younger is too young to work. It also implies that no age is too old to work, which is problematic since most people hope and expect to reach an age when they are still healthy but do not have to work. Most do. Labour market metrics, which use the total working-age population as the denominator, therefore tend to be distorted when there are demographic bulges, such as the baby-boom generation. For example, both the employment and participation rates will inevitably fall quite considerably as baby boomers retire.

halliwell fig5

halliwell fig6

A better way to look at labour market dynamics is to focus on specific “age groups” and “birth cohorts.” Age groups consist of people who are at a specific stage of life, although there is still considerable heterogeneity within age groups. Birth cohorts consist of those born during a specific time period, who have thus faced similar economic circumstances that can influence their labour market outcomes. Analysts often focus on certain age groups, such as youth (conventionally defined as ages 15 to 24) or older workers (often considered those ages 55 and older), but these groupings are far too broad. This becomes obvious when one looks at changes in labour force participation through the life cycle, as shown in figures 5 and 6.There is little in common between a 15-year-old and a 24-year-old (figure 5), a fact that seems unacknowledged in a lot of recent discussion about “youth unemployment.” Similarly, there is little in common between a 55-year-old and even a 64-year-old, let alone someone 75. As figure 6 shows, the age groups that are most similar in their labour market experience and life-course realities are, in fact, those in between these two oft-cited groups: those ages 25 through 55, when labour force participation is at its highest point over the life cycle.

Analysts also focus to a lesser degree on birth cohorts, perhaps without realizing it, such as when they speak of baby boomers or “generation X” or “generation Y.” Here, too, however, conventional analysis needs to go beyond such broad groupings. The baby boom, for example, comprises everyone born between 1946 and 1965; at any given time, the labour market reality across these two decades of birth cohorts is likely highly divergent.

The labour market metrics on which to focus

Given the significant demographic changes that are already happening and that are going to happen, aggregate labour market metrics, especially for the population ages 15 and older, are a poor measure of labour force dynamics. To look at what these changes will mean for Canada over the next decade, I consider age groups and cohort projections based on Statistics Canada’s long-term population projection model (Statistics Canada 2010), adjusted for the latest population estimates (as of 2012).

I begin by dividing the potential workforce into four groupings. The first two are the “starters” and the “enders”: those in their early stages of participation in the labour market and those at the latter stages of their careers. The former are those ages 25 to 29, who for the most part are past their formal schooling stage. The latter are those ages 55 to 59, who have accumulated considerable work experience and are at least contemplating when they will retire. These two groupings are where standard labour market metrics, such as participation and employment rates, can be affected significantly by key life-course events, such as the decision to go (back) to school, to enter the labour market or to retire. To extend this analysis further, I also add two more groups: those ages 18 to 24, whom I refer to as the “PSEers,” because they are of the age when most young people are pursuing post-secondary education and their labour market participation is often limited to part-time or summer employment; and those ages 60 to 69, whom I refer to as the “retirers,” because this age range is when most workers retire.

It is important to note that these groupings are all age-based cohorts, without regard to actual labour market or educational attachment. Rather, they reflect the demographic flow of potential workers relative to important milestones over the life course.

“Starters” are important. As the labour market has become more knowledge based — in particular, more rooted in specialized knowledge acquired through years of formal schooling and the obtaining of formal credentials — existing workers tend to become locked into specific occupational trajectories. As the demand for different occupations changes or the specific skills related to an occupation change, much of the adjustment in supply has to come from the new entrants within this demographic group.

Canada’s demographics are now at a transition point where the baby-boom generation will soon begin a long-run decline both as a share of the total population and in absolute levels. As of 2011, those born between 1946 through 1965 totalled around 9.6 million. Small reductions caused by deaths and emigration are starting to shrink the total baby-boomer population slightly, while population growth continues to reduce their demographic weight, from a high of nearly 40 percent of the total population in 1971 to around half that by 2030. Thus, the impact of baby boomers leaving the labour force will be considerably less than that of them entering it in the 1970s and 1980s.

halliwell figure 7

Figure 7 shows both the historical and projected demographic flows for each of the four age groups described above. It is instructive to compare how the PSEers (those ages 18 to 24) and the starters (those ages 25 to 29) evolve over time. The picture that emerges is one of broad stability, with some increase as a result of the “baby-boom echo” in the early part of this decade. The PSEers are at a peak about now; this is projected to diminish somewhat over the next several years and then to cycle up again around the end of the next decade. The starters hover at about 2.5 million per year, albeit with a peak around the end of this decade.10 Two obvious conclusions jump out from these projections. First, the starters are not a source of long-term-trend growth, with the upcoming peak little different from that in the late 1980s. Second, the starters are still numerous.

We also see a significant increase in the number of enders, an increase that started in the late 1990s and that will continue until about the end of this decade, when levels will dip somewhat. Indeed, the number of enders has grown from a level well below that of starters to one equal in size as of 2012. Enders are projected to exceed starters within 2014. The older group — the retirers (a 10-year grouping) — has been growing very rapidly and will continue to do so, peaking around the middle of the next decade.

halliwell figure 8

halliwell figure 9

This changing net balance between labour market inflows and outflows has led many to imply, misleadingly, that the only future source of labour force growth is immigration. In fact, the largest source by far is still the inflow of starters — that is, youth. That the enders are now so numerous does, however, have beneficial implications for the starters. Previously, Canada depended more on the creation of new jobs to generate employment for labour market entrants; now, the growing number of retirements will open up many existing jobs. It will also broaden where job opportunities are to be found, beyond growing industries and regions, to where older workers are concentrated.

Halliwell box2

Although it is true that without a major shift in the retirement age the ratio of retirees to workers (often referred to as the dependency ratio) will increase, nevertheless the total labour force will continue to grow. HRSDC’s latest projection (figure 8) shows a decline in annual labour force growth from approximately 1.1 percent in 2012 to 0.6 percent by the end of the decade (HRSDC 2011). For the most part, these results are no surprise: the large increase in the number of retiring older workers was long expected, and is now upon us. People delaying retirement might attenuate the impact on the labour force, as I discuss later, but it cannot change these trends, only delay them somewhat. The steady decline in the rate of growth of the working-age population, which started more than two decades ago, is expected to continue (see figure 9). Growth in the prime working-age population, those ages 25 to 59, has already fallen below 1 percent a year. The broader age group of 20-to-65-year-olds will get there early in the next decade.

Policy Issues in the Coming Decade

Labour-constrained growth

Once Canada reaches full employment, growth in the economy will become labour constrained. Indeed, the Bank of Canada’s Monetary Policy Reports project Canada’s potential real GDP growth to be just over 2 percent (Bank of Canada 2013), with potential labour supply eking out 0.6 percent growth by 2015 and labour productivity growth rising to around 1.5 percent (Bank of Canada 2012). The critical question is: will this result in aggregate “shortages” of millions of workers, as some observers claimed (see, for example, Macdonald 2006; Browarski 2007) prior to the recent recession? In a word, no. Such projections would have to assume some level (or growth rate) of potential real GDP that cannot be attained without additional workers.11 Since there is no basis for saying what that notional GDP might be, there is no basis for estimating such a shortfall. With slower potential growth will come slower growth in demand, moderated (largely) by the setting of monetary conditions by the Bank of Canada. The Bank could get it wrong, however, and there could be isolated shortages like those seen in 2006 through 2008. But such a situation would be short-lived.

Of course, some regions or sectors might face labour shortages even if the economy as a whole were not — arguably, parts of Alberta are again experiencing some. In general, this phenomenon is a “good thing”: it means we could be getting closer to full employment. As noted, an economy at full employment in aggregate can still have both excess demand and excess supply in specific markets. Statistics Canada’s job vacancies survey shows around a quarter of a million jobs remained unfilled as of early 2013, a “shortage” that amounts to just one-fifth of the number of unemployed (Statistics Canada 2013). The most recent COPS projections suggest both actual and prospective shortages in several areas, such as the health sector, coupled with surpluses in lower-skilled occupations, mostly in processing and manufacturing (HRSDC 2012).

Ensuring older workers who need to work longer can do so

As noted earlier, if one wishes to attenuate the effects of slowing growth in the labour force and GDP per capita, having older workers work longer is likely the most effective approach (Hicks 2012). Here, it is important to understand what “older” means: the issue largely concerns those between the ages of 55 and 64, which is when most Canadians leave the world of work, not those who are aged 65 and over (Halliwell 2009).

Do we now wish that older workers worked longer? Most commentators say yes. This raises the question of why. But “why” is not always clear. Sometimes it reflects simplistic GDP-centric thinking: the need to grow GDP. More often, it is motivated by the need to temper the fiscal implications of population aging in terms of prospective health and pension costs. At a macro level this issue is very real, but it ought to be cast as what it is: a fiscal failure to prepare for these highly predictable future costs when those who are now near or in retirement were in their prime working years of life. At the level of the family or individual, working longer can be viewed as a way to ensure the future adequacy of a household’s retirement income. Recent studies on preparation for retirement conclude that many middle-income baby boomers will not have adequate financial resources to sustain their standard of living in retirement (see Horner 2011; Wolfson 2011, 2013). This could be even more true for the next generation.

As they come to realize that their retirement income may not be adequate, many people who retire subsequently decide to return to work. The full extent of this phenomenon is unknown, but it is believed to be increasing in Canada (Schellenberg and Ostrovsky 2008; Lefebvre, Merrigan and Michaud 2012). Although retirement and return-to-work decisions are dependent on many factors, such as health issues and activity preferences, financial security is likely an important consideration.12 For the majority of Canadians who do not have a workplace pension plan of some kind — a group that already retires later than those who do have a pension plan — this is an acute risk. Those who have a pension plan that is not indexed are also not entirely secure. Two percent inflation still erodes nearly one-fifth of a pension plan’s purchasing power every decade. These issues are at the heart of the current debate on pension reform.

Related to retirement income adequacy is the issue of health costs. Canada’s health care system is effectively a seniors’ benefit comparable to or even larger than the public pension system, but entirely funded on a pay-as-you-go basis. There is no doubt that health services for older Canadians will cost considerably more in the decades ahead; the only major questions are about who will pay and how. Inevitably, most will have to pay more, either directly (through private insurance, co-payment or delayed coverage of new treatments) or indirectly (through higher taxes or lower nonhealth public services).

One answer to inadequate saving is simply to save more every year. Another is to save the same but for more years, by working more years; this could add several additional years of labour income to support lifetime consumption and contribute considerably to net savings and wealth.

It is important to look at the issue of later retirement in the context of changes in the average working life span and life expectancy over the past several decades. Figure 10, taken from the report of the recent expert panel on Quebec’s retirement income system, illustrates clearly how these factors interact. The trends it shows are generally true of Canada as a whole. Thanks to higher educational attainment and improvements in life expectancy, Canadians today are spending a smaller portion of their life in the labour market and a much greater portion in retirement than did earlier generations. This, in turn, is increasing the amount of income required to fund retirement. Working longer is therefore a fairly reasonable option.

This is already happening. There has been a major increase in the participation and employment (and unemployment) of older workers in the past several years. Some of this increase might reflect structural factors: better-educated workers, less physically demanding jobs, later starts to the beginning of careers, better health at older ages13 and so on. More important might be greater concerns about the adequacy of retirement income.14 Studies analyzing the effect of changes in wealth (such as stock market crashes or booms) on the labour market participation of older workers indicate that these fluctuations can affect individuals’ expectations of how long they will work by one to three years.15 This finding is consistent with various qualitative surveys of individual retirement planning. Since the 2008 recession, for example, the Sun Life Unretirement Index has seen the proportion of respondents saying they plan to work past age 65 increase from 16 percent to 27 percent in 2012. In the most recent survey, two-thirds of this group attributed their expectation of working longer to financial need (Sun Life Financial 2013).

halliwell figure 10

Providing clearer signals, at the very least, that people need to think about working longer is important on a psychological level. Many people have inordinately optimistic expectations of when they will retire, particularly when they are young (Schellenberg and Ostrovsky 2008). To the extent that people envisage early retirement they might not undertake the efforts needed to maintain their human capital and productivity through skills upgrading. They might consider that the remaining number of years they expect to work do not warrant the investment.

In the end, there is still a need to identify and reduce remaining disincentives to working longer. Yet the most powerful incentive to working longer lies in a better understanding on the part of Canadians of their retirement income prospects at any envisaged retirement age. Given the prospect that retirement income will fall short of what many middle-income earners will find they need, it is important that they understand what is at stake. In particular, people need to recognize that greater adequacy of retirement income tomorrow inevitably entails greater savings and work effort today.

As a growing list of analysts (such as Mintz and Wilson 2013; Morneau and Vettese 2013; Wolfson 2013) have argued, an optimal outcome of pension reform would be to enhance retirement benefits — and the requisite savings — as part of a combination of measures that include an increase in the normal retirement age. Focusing reform efforts on a mandatory, universal system such as the Canada/Quebec Pension Plan (CPP/QPP) is likely the best way to encourage longer labour-force attachment across the board.16

Having the right support for increasing numbers of older workers

In 2012 people aged 55 and older represented 18 percent of the labour force. Working longer will add to the soaring share of older workers in the labour market in coming years. This will have many implications for the labour market and the workplace, as well as for policy and programs. Greater attention to these issues is required (Expert Panel on Older Workers 2008).

Some think that the job market for older workers is now better, misinterpreting claims about the economy “creating” jobs for those 55 and older. This is wrong. First, much of the increasing share of total employment among that group just reflects population aging: many of these jobs are the same jobs filled by the same people, but who have just had their 55th birthday! Second, the labour market is still a difficult environment for older workers who lose their jobs. They are unemployed longer, and when reemployed are more likely to have lower earnings (Finnie and Gray 2011; Riddell 2011). The number of unemployed older workers (those ages 55 and older) has risen significantly in recent years, up 45 percent between 2008 and 2012 alone. This deserves more attention in public debate.

Many of our labour market programs were designed at a time when the labour force was much younger. As a result, a lot of focus then, and since, has been on retraining, because investments in human capital formation had potentially long payoff times. For older workers, training is problematic, for two reasons. First, the available assessments show that such programs result in only small, if any, improvements in earnings, even for the young (Jones 2012). Second, older workers are older: they do not have many years of work left in which to recoup the investment. This is why many suggest temporary wage subsidies rather than retraining as the best way to assist displaced older workers; wage subsidies are a way of easing the transition from a lost job to a lower-paying job. The objective and mechanisms of such subsidies, however, must be made clearer. Is the goal labour market participation or income support? Wage subsidies should also be piloted and rigorously evaluated before widespread adoption.

It might be premature, however, to write off retraining for the next waves of older displaced workers. First, they are likely to have much higher educational attainment than the older workers who lost their jobs in the 1990s and whose experience forms the basis of much of the literature. Upcoming cohorts of older workers will have stronger foundational skills of literacy and numeracy as well as a demonstrated capacity to learn. Second, given expectations about future labour market dynamics, those workers will be retraining for a much stronger job market. Third, the number of years they expect to work before they retire will also likely increase.

Investing in (the right) formal education for the young

In thinking about younger workers, we must address two myths. The first holds that an older society has few, or at least fewer, young people. This view also suggests that, with larger contingents of soon-to-be retirees, immigration becomes the only source of labour force growth. This is wrong. The first place to look for new labour force entrants and new and updated skills remains the people just entering or just completing their formal education in Canada. As a group, the pool of “starters” — young Canadians ages 25 to 29 — will be in the 2.5 million range well into the next decade. This is approximately 10 times the total annual intake of immigrants (on average, a quarter of a million per year) and 40 times the intake of economic class principal applicants.

The second myth relates to the often-repeated claim that “two-thirds of new jobs will require post-secondary education,” too often implying that Canada’s education system will not meet this need (Miner 2010; Commission on the Reform of Ontario’s Public Services 2012). Although the projection of two-thirds is likely true, the implication is unfounded. Why? Simply estimating the educational requirements for future employment does not convey anything about supply — whether it will fall short of, be equal to or even exceed those requirements. On the demand side, HRSDC’s analysis of occupational skills in the Canadian labour market confirms that about two-thirds of job openings over the next decade will require post-secondary education, with  many of the openings consisting of existing jobs vacated by retirees.17 The share of new entrants — “starters”18 (see figure 11) — with completed PSE is already over two-thirds, however. This is a consequence of youth staying in school longer and immigrants increasingly having PSE.19 Thus, the total supply of new entrants into the job market with PSE will suffice. This is consistent with the evidence that Canada seems to have had this balance right in recent decades (Boothby and Drewes 2010; Drewes 2010).20

halliwell figure 11

halliwell figure 12

It is possible that the skills intensity of jobs, in terms of the requirement for formal credentials, will grow even faster than envisaged (see, for example, Miner 2010). This is, however, just speculation; indeed, the reverse might be happening. Although this question has yet to be tested in a Canadian context, emerging evidence from the US labour market points to a growing oversupply of highly skilled workers. It suggest that since the early 2000s there has been a shift in demand away from highly skilled labour (Autor 2007; Beaudry, Green and Sand 2013), reversing a consistent trend of rising demand through the 1980s and 1990s. A similar pattern might well hold true in Canada.

It is noteworthy that, at the same time that the number of working-age Canadians without any education beyond high school has declined, the number of jobs that ordinarily require only secondary education or less has continued to rise, albeit at a slower pace than for other occupations. How is it, then, that there is not a shortfall of low-skilled workers? The answer, according to analysis by HRSDC and years of labour market data, is that a relatively constant share of those with completed PSE end up in jobs that do not normally require such education. This is true in good times as well as bad (see figure 12). This implies that a high and growing share of the jobs not normally requiring PSE are being filled by workers with PSE — the often-cited “underemployment” problem for recent graduates. The problem is even worse for those with few skills, who risk being displaced even from low-skill jobs, especially when unemployment is high.

The immediate policy implication is that we ought to continue to monitor these outcomes, through instruments such as the National Graduates Survey,21 to see if excess demand or excess supply starts to develop. The latter is as problematic as the former; after all, while young people are in formal learning they earn and work less, reducing aggregate labour supply and potential GDP during this time (Fougère et al. 2006).

Improving field-of-study choices

The problem of underemployment among many PSE graduates might relate not so much to the amount of education young Canadians get as to what they actually study. In general, the data show the expected payoff hierarchy: on average, college pays better than secondary, and university better than college — although the returns seem to be falling for the newest cohorts of young. But averages hide enormous heterogeneity, among and even within myriad fields of study (see, for example, Stark 2007; Finnie and Frenette 2003).

Diverse learning outcomes are inevitable, and they occur for various reasons. Some people learn less than others.22 People make mistakes in choosing what to study (and there is considerable field switching in PSE). It is also true that some people cannot get into their desired field of study because of enrolment constraints. And, of course, not everything always goes according to plan. Witness the high-tech bust of the early 2000s, which came just as the sector’s dynamism had successfully enticed many young Canadian students (and immigrants) into related disciplines. There will always be degrees of mismatch ex post. Finally, it is also important to remember that earnings outcomes are not the only factor that matters. Many young people choose their field of study out of interest, even if that pays less well than do others (Finnie 2002), although earnings are certainly not irrelevant to choices made (Gunderson and Krashinsky 2009).

The extent of mismatch HRSDC identifies in its detailed COPS projections (excess demand plus excess supply) has never been huge: typically tens of thousands in the context of a national job market of nearly 17 million people (although regional imbalances would be greater). Actual mismatch could be much larger, though. In COPS, the projections of labour supply by occupation start with enrolments/graduation by educational level and field of study. These are then mapped into occupational supply. Traditionally, COPS has used the historical average of where graduates end up in the labour market in order to establish this occupational supply profile. From this perspective, it is obvious how some professions relate directly to a graduate’s field of study: most nurses end up in nursing occupations, and nursing occupations require graduates who come from a nursing program. In other disciplines, however, such as English literature, graduates can end up scattered across a wide range of occupations, many unrelated to the field of study (Boudarbat and Chernoff 2010). This can be the case even in high-skill employment. In Australia, for example, half of graduates with information and communications technology (ICT) qualifications end up in occupations other than as “ICT professionals,” and only 30 percent of those employed as “ICT professionals” have such qualifications (Mavromaras 2013). Although comparable data are not easily available for Canada, this speaks to the wide array of skill sets across the occupational structure.

A few years ago, the COPS team looked at occupational supply using a different method: they assigned the labour inflows from more fields of study to occupations directly related to the respective field of study. In essence, they assumed PSE graduates would supply labour for jobs normally requiring particular kinds of PSE. This is likely more consistent with students’ expectations about where their education should take them in the labour market. This approach was referred to as the ex ante rather than the ex post distribution of outcomes. The result of this sensitivity analysis was that mismatches were much higher than previously estimated, with both more occupations showing a shortage of workers and more showing a surplus (HRSDC 2010).

Understanding these dynamics is essential to ensure the future performance of Canada’s labour market and must be the subject of more analysis.

Improving students’ effectiveness in choosing their fields of study likely will require a number of different policy instruments. The need for better labour market information, including the kinds of projections performed by COPS, must be acknowledged (Advisory Panel on Labour Market Information 2009). Many claim that projections like those of COPS are useless on the grounds that they are “always wrong.” That is a facile claim; the question is whether such projections provide better information than ignorance or guessing. Clearly, they do. But we will need to go further: our national labour market information system is deteriorating, not improving — although, at the micro level of individual job seekers and employers, labour market information has never been more plentiful thanks to the Internet.

Information informs decisions, but decisions need to be made and put into effect. This requires changes across the education system:

  • To help students better understand PSE and prepare for it. There is evidence that some students do not understand the general benefits of PSE in terms of earnings and nonearnings outcomes (Oreopoulos and Dunn 2012) as well as how particular fields of study relate to their own specific abilities (Lehrer 2013). If students do not recognize what they are good at and how education will assist them in the future, they could end up making bad choices. More important, Canadian culture seems to convey to students that university education in any field is still more desirable than college and the skilled trades; this perception needs to be corrected and anchored, one hopes, in timely and accessible labour market information.
  • To give students more flexibility to choose and switch fields of study. The ability to switch fields of study — aided by better prior learning assessment and recognition initiatives that help identify and recognize skills acquired outside of formal education — has been shown to increase a student’s chance of successful integration into the labour market. Related to this is also the issue of how soon students entering PSE must choose a major field of study. Evidence from Scotland, which allows students to select fields of study later relative to the system in England and Wales, shows that choosing later reduces the probability of students switching occupational fields when they enter the labour market (Malamud 2011).
  • To give students better access to the fields in demand. Colleges and universities need to admit more students to fields of study that are in high demand, rather than just charging higher fees. This likely would help students sort themselves better across field-of-study/career pathways.

Investing in the skills of immigrants, even long after arrival

There is a tendency to refer to immigration as “imported” human capital, in contrast to workers who were born and educated in Canada. Implicit in this description is the notion that “imported” human capital is less expensive than that produced here through formal education. This is likely true, but what about the quality and quantity of that human capital?

Not surprisingly, when attempting to explain why in the past two decades it has been taking longer, on average, for the employment outcomes of new immigrants to converge with those of Canadian-born workers, most studies largely attribute the problem to deficiencies in human capital.23 As mentioned earlier, this relates primarily to official-language literacy and numeracy skills, as well as to difficulties with the recognition of foreign credentials (Picot and Sweetman 2012).

Several factors are in play. First, only about a quarter of immigrants are actually selected through the points system as economic class principal applicants (PAs). For PAs, until changes were undertaken several years ago, selection policy inadvertently might have worsened the problem. In focusing on general educational attainment for the so-called knowledge economy, rather than on specific skills or work experience in demand in Canada, immigration policy widened the gap in employment outcomes between immigrants and Canadian-born workers. In particular, the failure to set the official-language-skills bar higher for immigrants hoping to work in the knowledge economy might have exacerbated their difficulties in integrating. Finally, the equivalence and recognition of foreign credentials are also challenges.

These issues can be resolved, either before or after arrival. An attempt is now being made to resolve them by applying tighter and more job-relevant selection criteria to determine who is admitted. This will improve the situation, but only slowly, as the new criteria operate on the flow of around 60,000 new economic PAs who immigrate each year, not on the stock of around 7-million-plus immigrants already here (Statistics Canada 2013). The relatively poor labour market outcomes of immigrants can also be addressed after landing through skills development. In recent years funding for immigrant settlement has been increased, although the amounts invested remain modest — in 2012, spending on settlement supports totalled $3,919 per newly arrived immigrant.24 This amount is quite modest when compared with what is spent (per person) on the formal education of those born in Canada. In any event, the moment landed immigrants acquire citizenship, they are fully embedded in the Canadian labour force, and any other help they get comes from other budgets.

Without greater support, it is likely that the immigrant skills gap will persist. Adult literacy data show only slow progress in the official-language skills of many of those born abroad, even long after arrival. Data from the latest Programme for International Assessment of Adult Competencies survey, for example, indicate that across Canada as a whole, established immigrants (those living in Canada for 10 years or more) had an average literacy score of 257, compared  with 254 among recent arrivals, and 280 among the Canadian-born population (Statistics Canada 2013). A much greater investment is needed to help immigrants further develop these skills, for instance, through adult language training and, especially, adequate official-language training for young immigrants who enter partway through the primary or secondary school system and struggle to catch up. This is likely an example of low-hanging fruit, which could have high returns, both economic and social. It would also go a long way toward meeting immigrants’ own expectations for economic and social integration.

Ensuring workers have skills “ladders”

In Canada, human capital is formed primarily through three channels: school, or the formal education system, which I refer to as the “first-chance” system; work, or the skills learned by doing the job, day in and day out, for years, as well as those acquired through workplace training; and unemployment supports, delivered primarily through EI and welfare programs, to assist with reemployment.

Canada has a very good “first-chance” system. Primary and secondary school education are free and of relatively uniform quality (with the exception of some Aboriginal education). PSE is heavily subsidized, through both taxpayer-funded contributions to PSE institutions and a plethora of student financial assistance programs. For those who follow the conventional educational path through primary school, secondary school, then PSE, the public subsidy is substantial, likely in the six figures.

Not everyone, however, follows this path. Many do not because they cannot: evidence suggests that lack of access to PSE for many, including many from poor socio-economic backgrounds, is primarily a consequence of a lack of preparation, low educational attainment on the part of parents or a lack of innate skills (Finnie and Mueller 2008; Restuccia and Urrutia 2004). Many people also select different paths or just make mistakes — for instance, launching into a job at the earliest opportunity — or, because of various life events, fail to complete their education. The state tends not to subsidize alternative pathways much, if at all. Someone envisaging a career as a hairdresser will not get much public support, while someone pursuing a PhD receives a huge investment from the state.

Having built a system that provides some people with a very large educational subsidy early in life while others get nothing beyond basic education, we should not be entirely surprised that this contributes to an early and persistent divergence in incomes. But can these differences be overcome by skills upgrading?

Canada has a plethora of publicly funded programs to help the unemployed improve their skills. Many were designed when Canada had prolonged periods of slack labour markets and when the labour force was younger. Yet these programs do little to help those whose challenge is not unemployment but, rather, the need to acquire new skills or improve existing skills. Why? Such people already have work, which makes them ineligible even if their employment is low-skill/low-wage.

For those who are employed, workplace training is at best a limited opportunity. In most cases, it relates specifically to tasks and skill sets required for their current jobs. Because workers might leave the firm, employers often have little incentive to assist with extensive broad-based skills upgrading. The amount and availability of training also depends on the job one occupies in the first place. Jobs requiring high-level cognitive skills intrinsically promote the acquisition of human capital as part of working; “routine” jobs such as maintenance or janitorial work do not offer much opportunity in that regard. This is why workers in low-skill occupations do not receive much training. Indeed, recent analysis concludes that employers are much more likely to direct their financing for training or education at highly skilled workers in high-skill jobs, rather than at workers with low-level skills in low-skill employment (see figure 13). High-skilled workers are also somewhat more likely than low-skilled workers to receive government financing for their training and education, as well as to finance it themselves (Desjardins 2011).

halliwell figure 13

Nonetheless, it is important to note that many low-skilled workers do manage to upgrade their skills. They take evening or even weekend PSE on a part-time basis. They adopt strategies whereby some family members work while others upgrade their skills. Indeed, Canada’s support for PSE has expanded to recognize this, providing student loans for older students and part-time studies (with course loads as low as 20 percent of the full-time norm), and now waiving interest on loans while part-time studies are under way. Individuals can even draw down part of their registered retirement savings plans, tax free, to support a full-time return to education, albeit with repayment obligations (or tax will be due). That said, the Registered Educational Savings Plan and many educational grant programs still assume that PSE is something young Canadians do more or less immediately after high school.

Of course, these options are not so easily available to many, such as older singles and, especially, single parents. The opportunity cost of their forgone earnings combined with household and family financial commitments limits their ability to take extensive formal skills upgrading. Like many recent immigrants, they might find themselves in “survival jobs.” For many, there is no second chance unless they become unemployed, and even then the extent of the retraining opportunity does not come close to that provided through formal education.

Overall, our system is still primarily focused on employment insurance, which as such offers numerous modest repeat chances, but where the odds of attaining an effective pathway out of low income are not high. While, for most, this system helps prevent prolonged periods of low income, it does not prevent frequent bouts of low income. Indeed, once someone ends up on a lower-skill life path, the record shows that major skills upgrading is unlikely. Those with low levels of education when young are least apt to acquire more later, while those with higher education often obtain more throughout life (Drewes 2008).

In the context of a labour market future where supply will be more constrained and the expectations for labour and general productivity growth much greater, it is imperative that Canadians have effective pathways to upgrade their skills throughout life. To do this, we need to develop a comprehensive “second-chance” system to provide more structured and, one hopes, successful alternatives for those whose “first chance” did not work out, whether through no fault of their own or due to poor decisions.

The payoff for these individuals could be high, especially for those still relatively young (for instance, those in their 30s). For example, evidence shows that the return for those completing PSE following a delay after high school is even higher than the return for those who completed it without delay (Ferrer and Menendez 2009). The benefits for society as a whole depend very much on how many young Canadians have missed their “first chance” and have not yet managed a “second chance” but could benefit from one.

Estimating the size of this particular group is difficult. One indicator is the percentage of Canadians who report unmet skills-development needs. According to data from the Access and Support to Education and Training Survey, in 2008 this was 38 percent of Canadians ages 35 to 44, declining to 20 percent among those ages 55 to 65 (Knighton et al. 2009). This figure is likely inflated for our purposes because it includes everyone interested in both training and education, where clearly a “second-chance” system would focus on full courses intended to raise general education levels or to acquire a credential. Another, perhaps more reliable indicator is those who currently have skills below the level typically required for their occupation, which has been estimated at about 16 percent of the Canadian workforce (Desjardins 2011).

Many of the components of a “second-chance” system are likely already in place, but not so structured as to constitute a comprehensive, coherent, permanent and hence navigable system. In particular, such a system might need to aim beyond its current goal of helping people quickly return to a job that is similar to the last one they held, which is hardly a recipe for providing a career “ladder.”

The system I envision would, effectively, constitute an extension of the risk prevention or insurance function of the existing social safety net. By focusing on the educational needs of those already employed, it could be a more efficient and effective means of helping the working poor. In essence it would support lifelong learning. Rather than waiting for someone to become unemployed and then providing them with support to help them return to their prior earning potential, a “second-chance” system would help improve the earning potential of those who are already employed but do not have the skills or resources to advance in the labour market. This would fill an important gap in the current suite of labour market programs in Canada, and could also help reduce the long-term risk of unemployment for some. A “second-chance” system also would be subject to moral hazard by potentially encouraging suboptimal “first-chance” choices; but ensuring it did not become a “third-“ or “fourth-chance” system or subsidize learning that would have taken place in any event would reduce that risk.

Ensuring that training works for the unemployed

The problem

“Active” labour market interventions encompass job search assistance, career counselling, wage subsidies and, especially, skills development. The latter is premised on the observation that those with the fewest skills face the highest risk of unemployment, in both frequency and duration, through bad times and good. Thus, if the unemployed had better skills, they would have better labour market outcomes.

It is not that simple. A significant challenge for labour market policy is that evaluations of training programs have often found them to be fairly ineffective at improving employment or earnings outcomes (Warburton and Warburton 2002; Gunderson 2003). With such disappointing results, some might view training as largely an exercise in homeopathic public policy — that is, something that has no active ingredients. Certainly, the argument that Canada should increase its spending on active labour market measures (which was 0.3 percent of GDP in 2011) to bring it more in line with the average spending of other OECD member countries (at 0.6 percent of GDP) holds little water if the returns are as low as commonly measured. In particular, in a world of scarce and prospectively scarcer fiscal resources, such low payoffs make it hard to justify additional funding.

The perception of failure might be too pessimistic: training and active measures do work under certain conditions. Paradoxically, they likely work least well when they are most in need — that is, when the labour market is weak. During these periods, there simply are not enough jobs, and many people with good skills cannot find work. The most skilled are hired first and the least skilled last. If training does not significantly change trainees’ skills, they might remain unhired, and the new skills could quickly atrophy. If trained sufficiently, they might find work but merely displace someone else further down the hiring queue.

What is needed is a serious rethink and redesign of training measures, including what we want to accomplish and for whom, how the programs work and how to measure outcomes. In thinking about each of these, it is important to ask some basic questions.

What we want to accomplish

In terms of what we want to accomplish, we have to choose what skills to offer and to whom. Better results likely come out of more job-oriented training, covering a fuller range of skills deficiencies and coupled with follow-up support to help trainees transition into work. Too often, training has been premised on a “Field of Dreams” approach: if you build (train) them (the unemployed), they (employers) will come. Turning to whom we wish to train, many argue that training should focus on those facing the greatest barriers to employment. This might not be the best approach. For some, the gap in essential skills or formal education might be too large for specific training programs to be useful. For this group, the cost-benefit ratio might appear less attractive than for those with fewer barriers, who would benefit most readily from retraining. From this perspective, “skimming the cream” actually might have the best payoff — provided, of course, that we do not help only those who would have done well in any event. The challenge for policy, therefore, is to articulate very clearly for whom and to what end training resources are best spent.

How the programs work

Key to greater effectiveness is the quantity of training. A few weeks of training will not move someone up many rungs on the skills ladder. After all, human capital is acquired over years of formal education and decades of work experience. In essence, this is the socio-economic equivalent of the “dose-response” problem in medication: antibiotics cure infections only when the patient is administered a full course. A good “second-chance” system ought to test higher doses of training treatment to see if they make a difference. Certainly, during the 2008-09 recession, some provinces, such as Ontario, went in the direction of supporting longer and more costly investments in formal training. It is important for all provinces to consider the lessons learned from this experience.

How to measure if training is working

Longer and more intensive training programs usually fall prey to being seen as “expensive.” This stems from defining “expensive” only in terms of dollars spent. In public policy, the “expensive” programs are those that, regardless of cost, deliver few or no benefits. To determine what training accomplishes, better evaluation methods need to be put in place.

Few programs are claimed to have little or no benefit. The problem is that, although evaluations are widespread (often obligatory), good evaluations are not. The lack of consistency and rigour in evaluating active labour measures is a widely noted problem (Hum and Simpson 2002). Many evaluations fail even to measure outcomes or to take into account “cream skimming” — where the beneficiaries would have done well with or without the training. Most simply consist of a postintervention analysis of descriptive statistics — for example, the change in participants’ wages relative to those they earned in their occupation before unemployment. Few evaluations include any form of randomization to ensure robust comparison groups (Warburton and Warburton 2002).

Evaluation of Canadian labour market programs is also hampered by a high degree of inconsistency. Even though federally funded training programs are relatively similar across the country, evaluation reports vary from province to province because they are tied to bilateral labour market agreements between the federal government and the provinces, and they are not subject to common outcome and reporting measures (Wood 2013). The protection of personal information in administrative data can also prevent access to and use of valuable data for evaluation and research.

Active labour market measures should include a rich and ever-changing portfolio of activities, with ongoing evaluation used to reassess and redesign programs based on evidence of what works. Investments in active measures might turn out to be expensive in terms of direct costs but cheap in terms of costs-benefits. Indeed, they could generate rates of return as high as those associated with PSE. There might soon be an increased need for better training and a payoff from it. Once the Canadian economy returns to full employment, slower labour force growth ultimately will place a greater premium on better matching of skills between workers and jobs. In this context, a revamped approach to training could enable employers to reach further down the skills ladder.

The proposed Canada Job Grant

The proposed Canada Job Grant (Canada 2013) might be a step toward such a new approach. It would be funded by reallocating $300 million per year from federal transfers under current labour market agreements (LMAs), with matching funds from both provinces/territories and employers for a total potential investment of $900 million.25 Employers would be able to access the grant (a maximum of $15,000 per trainee if fully cost shared by all three partners) to support short-term training needed to hire an unemployed worker or help someone move up within the firm. This approach, it is argued, is likely to be more responsive to actual job requirements than are current LMA programs.

The one formal critique of the proposal so far (Mendelsohn and Zon 2013) correctly highlights some potential issues:

  • the unilateralism of the initial federal proposal;
  • the significant transitional difficulties for program managers and delivery agents in winding down existing LMA programs and starting up new programs;
  • the difficulties some provinces and territories might have in raising their matching funds; and
  • the significant risk that this training subsidy would only substitute for existing employer investments, and hence not be very effective in leveraging additional training efforts and investments.

The critique might be off the mark, however, in stating that the reduction in LMA resources would adversely affect the “most vulnerable” (Mendelsohn and Zon 2013, 1). In fact, the most vulnerable might benefit the least from current training, as their labour market skills deficiencies could be considerable. Working with employers and focusing on those who are closer to being job ready could provide a higher payoff in imminent employability.

Improving the social safety net

Every labour market decision — to pursue education in a specific field, to acquire new skills, to move to a new labour market, to accept a job or to change jobs — entails a risk that the decision might not pan out but might instead result in loss of income. The broad social safety net (including income support programs and progressive taxation) helps to insure against that risk. It actually supports greater risk taking, which supports more rapid labour market adjustment, not just at the level of the individual but also at the societal level: the social safety net makes a polity more likely to support risk, such as structural change.

Of course, insurance also subsidizes the realization of the risk, which is an issue when the risk is entirely predictable, such as for seasonal employment. Risk taking also might increase because the risk taker does not bear all of the costs: this is “moral hazard.” True insurance tends to reduce moral hazard by raising insurance costs for those who experience bad outcomes more often than the average. For instance, in the case of EI, this could be done by adjusting contribution and benefit rates based on some risk factor for unemployment. Another example would be imposing time limits on benefits such as social assistance. Instead, we try to reduce moral hazard by effectively reducing the insurance coverage via restrictions on EI eligibility and qualifications, capping the amount of earnings covered under EI or setting inadequate benefit rates and rigorous means tests for general social assistance.

The corollary of relying on a social safety net primarily delivered through taxes and transfers is that, for it to be sustainable, it must be very narrowly constructed. The only way to offer adequate support for the poorest is to withdraw that support as income rises at such a rate that marginal effective tax rates (METRs) become exceedingly high and the net income gained is very low (see Duclos 2007; Millway, Chan, and Stapleton 2009).26 This distorts work incentives and likely reduces labour supply. Adding a plethora of tax-and-transfer instruments and means-tested benefits to the mix just complicates the calculations (see figure 14); it does not change the inherent problem. It is important to continue to try to lower METRs in order to reduce the “welfare trap,” as noted by the recent Ontario report on social assistance reform (Commission for the Review of Social Assistance in Ontario 2012). In the end, though, likely the most effective solution would be to push wages upward at the low end of the income distribution by reducing the number of least-skilled workers.

halliwell figure 14

Adopting a longer-term social insurance perspective would be helpful in changing these dynamics, for instance, by making benefits more generous initially but less generous with lengthy or frequent use. In the case of EI, this could amount to penalizing frequent users (as recent changes to EI have started to do) or rewarding infrequent users; both options would create similar incentives, although the latter option could prove easier politically. Social assistance programs could also implement a lifetime limit on benefits for those recipients who are considered employable, similar to changes brought about as part of welfare reform in the United States during the 1990s.27 These kinds of changes would create stronger incentives for frequent recipients to return to the labour force. They would also provide greater benefits to those experiencing their first spell of unemployment, thereby reducing the pressure on them to take the first job available, even at the risk of being underemployed.

“Fixing” EI: more than just horizontal equity

Employment insurance is a policy instrument with many objectives, not all of which directly support the labour market. Some aspects of the program facilitate the efficient functioning of the labour market — in particular, helping workers through adjustment periods in firms and industries — while others hinder it. EI supports flexible labour markets by enabling better job search for more suitable employment, at the cost of a longer term of unemployment. It supports risk taking by workers by facilitating greater labour mobility, but it can also inhibit labour mobility out of regions that cannot offer adequate full-time, full-year work. It might even inhibit workers from taking full-time employment when it is available in high-EI-benefit regions, where for some workers combining part-year earnings and EI benefits can be more attractive than full-year, low-wage employment. In general, EI is oriented toward the problems of regular unemployment, whether cyclical or, especially, seasonal. In this respect, it is helpful in reducing the cost to employers of some aspects of attracting and retraining workers. EI is not so helpful in dealing with irregular unemployment, such as that facing long-tenured older workers who lose their jobs following major structural changes in the labour market.

Since the mid-1990s, there have been numerous calls to “fix” EI, although not all these calls are clear about the specific problem that needs solving. Actual changes have been less numerous, and as a result have tended to continue to subsidize part-year employment rather than reward long-career, full-year employment (Gray 2004).

Many of the calls to fix EI are couched in terms of simply providing greater horizontal equity in a system that currently sets eligibility thresholds on a highly regional basis. Some suggest, for example, that it is wrong that someone unemployed in Toronto faces tougher qualifications and shorter benefit periods than someone in New Brunswick’s Miramichi Valley simply because the unemployment rate is lower in Toronto (Busby and Gray 2011). Others argue that such a differentiated approach results in inadequate support in some areas, such as Ontario, where the recent recession left unemployment high but the total number of EI beneficiaries remained relatively low (Mendelsohn and Medow 2010). In response, there have been calls for a common national standard for the number of hours worked required to be eligible for EI. It is also noteworthy that, to address the needs of a number of different groups, the federal government in recent years has had to adopt new training programs outside the EI umbrella (for example, the Targeted Initiative for Older Workers).

The challenge here is to make EI more automatically sensitive to major changes in the labour market arising because of business cycles or structural shifts in the economy, and at the same time less sensitive to changes in employment from season to season. It also needs to be better designed to serve the needs of large urban centres, where most Canadians work and the largest number of unemployed are found.

I argue that EI should be easier to get, and last longer, when unemployment is higher as a result of a recession or a regional or sectoral shock. This is what makes EI an automatic stabilizer. In the last cycle, EI was responsive to the recession, but not entirely automatic; the federal government had to make many discrete, ad hoc changes, including on the contribution side. Some changes were specifically targeted at workers with longer-term labour market attachment.

EI should also be more responsive to the needs of long-tenured workers who have lost their jobs due to structural change. For this group, the typical “dose” of EI and active measures might not be enough to produce the needed “response.” These workers could benefit from an “infrequent user supplement,” such as more extensive training programs and associated financial support.

It is not so clear, however, that we want EI to be more generous in chronically high-unemployment regions, since chronically high unemployment makes for structural unemployment. It is widely accepted that EI contributes to this situation. The most striking evidence comes from comparing differences in unemployment between similar Canadian and US regions, such as New Brunswick and Maine. Kuhn and Riddell (2006) find that the percentage of men working only part of the year (less than 40 weeks) was substantially higher in New Brunswick than in Maine in every decade between 1940 and 1991. In this sense, EI deviates from an insurance system.

Instead, EI needs to provide a better social safety net and better support for job search in large urban labour markets. These are the areas with the most population, the largest labour force, the most jobs and the most unemployment — it is where three in four EI recipients live. Less generous EI in such areas presumes that it is easier to find work when the unemployment rate is lower than in the regions with perpetually higher unemployment rates. It is not clear that this is actually so.

An important question is the extent to which demographic trends are reducing the problem of high EI use in rural areas. As more people move into cities the share of the population in nonurban areas is shrinking. In these areas, EI recipients are somewhat older, and in the coming years proportionately more will come to rely less on EI as they become eligible for CPP/QPP benefits, Old Age Security and the Guaranteed Income Supplement. That said, there are still large numbers of younger EI recipients in nonurban areas, which have about one-quarter of total EI recipients.28

This again raises the issue of making EI less generous in the high-unemployment regions. Many see such a change as necessary to help spur out-migration to where the jobs are now. They argue this would increase GDP per capita, as migrants likely would move from part-year to full-year employment and earn more per hour. This scenario is complicated, however, by several factors.

First, it is very hard for people to move. Perhaps the key factor is the difference in the cost of living between regions with jobs without workers and regions with workers without jobs — say, Edmonton versus Corner Brook. Overall, the average cost of living varies somewhat by region, but for most components not by huge margins. The exception is the cost of purchasing a house, where regional differences can be factors of 2:1, 3:1 or even 4:1.29 This matters. A few moments with a mortgage calculator can show that the hourly wage difference in a full-time job needed to pay off, for example, an extra $100,000 of mortgage debt over 15 years amounts to between $4 and $5 per hour, after tax! This is why the young move most: they do not have a significant part of their net wealth held in the immobile asset of an owned home.

This also likely explains in part why a considerable number of workers in these regions accept work arrangements that allow them to work in another province, particularly in western Canada, without formally moving there. Known as “interprovincial employees,” this group is not officially counted in the definition of “interprovincial migration” because they do not formally change residency from one province to another.30 They are, however, an important source of semi-mobile labour. In 2008 alone, interprovincial workers represented more than 6 percent of the total Alberta employment31 — more than the number who actually moved there. That year, more than 26 percent (32,000) of Alberta’s interprovincial employees came from the Atlantic provinces (Laporte, Lu and Schellenberg 2013). With workers already responding well to these labour demand signals, it is not clear that EI changes are needed in order to encourage actual out-migration.

Second, more out-migration might not be optimal. The regions where EI is most generous are typically those facing the poorest demographic prospects and already experiencing substantial net out-migration. These regions have older populations: many of the young have left, often to pursue PSE and jobs commensurate with that education.

Although population and labour force growth will slow across Canada, the patterns will differ significantly by region.32 East of the Ottawa River, total population growth is expected to fall to under 0.5 percent a year, half the rate of growth west of the river (figure 15). The problem is even more pronounced for labour force growth: provinces east of the Ottawa River are expected to experience no growth or even an absolute decline in the working-age population through the middle of the next decade (figure 16). Greater out-migration of workers from these regions would just compound the problem.

halliwell figure 15

halliwell figure 16

Determining how many immigrants Canada will “need” in the next decade

Immigration policy determines how many and which immigrants Canada should admit. As such, it is, and is not, a powerful lever for labour market policy. It is, because it operates directly on the flow and composition of new labour supply (Kustec 2012). It is not, because there is considerable doubt that immigration does much for the existing population and labour force (Drummond and Fong 2010) — aside from vastly enriching our culture. This doubt arises because, while immigrants clearly contribute to overall levels of population, labour force and GDP, it is less clear what the net effect is on GDP per capita (the commonly used measure of economic well-being). Indeed, immigration tends to have a slightly negative effect on GDP per capita, as immigrants generally have average earnings below those of the Canadian-born, despite their higher average educational attainment. This is a purely arithmetic effect.

At issue also is the indirect impact of immigration on workers who already reside in Canada. We know that immigrants do not “take” Canadian jobs — indeed, they help to grow total output and reduce shortages in occupations that are difficult to staff, such as those requiring highly specialized skills. Nevertheless, having high immigrant intakes when unemployment is high can be problematic for both new immigrants and other new entrants to the labour market. This underscores the importance of having a selection policy and admissions targets that are responsive to labour market conditions — in particular the long-lasting impact of recessions on unemployment.

That said, there will be scope for more immigration. Statistics Canada’s medium population projection, which I use throughout this paper, already assumes some ongoing increase in immigration, to a level consistent with 0.75 percent of the population for its full projection period. This would take immigration from the (latest) target range of 240,000-265,000 annually to 285,000 by 2020 and 313,000 by 2030. Ultimately a few tens of thousands of immigrants more or fewer per year will not make that much difference to the size of the labour market, given that the total population of Canada is 35 million now and is expected to reach close to 40 million by 2030.

Refraining from using temporary workers for permanent jobs

The “T” (“temporary”) in TFW used to modify the job, not the worker. That is, TFWs predominantly filled temporary jobs, whether very temporary (a rock concert) or somewhat temporary (often seasonal jobs in tourism and agriculture). More recently, the adjective increasingly modifies the worker, not the job: temporary workers rotate through permanent jobs. By the mid-2000s pressure from Canadian employers led to a considerable streamlining and expansion of TFW programs and, subsequently, to a significant increase in TFW flows. Many such workers went into jobs that were permanent, or so it seemed before the recession. The context here is important: at the time, the economy had grown to a state of excess demand and there were aggregate labour market shortages. This context is not likely to repeat itself, but if it did, it would not persist for long, as the Bank of Canada would act to correct it within a few years at most.

In theory, employers who apply for a formal labour market exemption (known as a “labour market opinion”) to hire TFWs33 can do so only after demonstrating an inability to find qualified Canadians at the regional median wage for a particular occupation (but never below the wage the employer would pay for comparable Canadian workers). The desire to be responsive to the needs of employers and to reduce the bureaucracy they face in accessing the TFW program was, until recently, an important policy objective, even though hundreds of thousands of Canadians were unable to find work in fields with many TFWs. Of the 202,510 positions34 that held a labour market opinion in 2012 permitting a TFW to be employed, 88,000 (43 percent) were in the four industries with the largest number of unemployed workers, accounting for approximately 390,000 unemployed Canadians.35 With so many Canadians unemployed or underemployed, it should have been much harder to bring in TFWs (Green 2011).

As a rule, we should use TFWs largely when the work is temporary, and seek permanent workers for permanent jobs, through either immigration or improved labour market adjustment. That said, if we wish to sustain seasonal industries, using TFWs remains preferable to having a new generation of Canadians entering the seasonal work/EI cycle. In that respect, the series of changes to TFW programs the federal government announced earlier this year (HRSDC 2013b) could be quite beneficial. They start to make the “T” in TFW also stand for “transitory,” in the sense of providing access to foreign workers while employers seek out and train Canadian workers for available jobs. If done well, these measures could dovetail with the proposed Canada Job Grant. Employers could be encouraged to fund the training commitments they would have to make to transition from using TFWs to employing Canadian residents. It might be appropriate to link the new TFW measures explicitly with at least some of the Canada Job Grant programming.

In addition to these recent changes to the TFW program, some have argued in favour of imposing an annual cap on the number of temporary workers entering Canada (see Worswick 2013). Given the potential wage suppression effects associated with having too many TFWs, this certainly would be a worthy next step. Another way to ensure TFWs are used primarily for temporary employment is to link labour market exemptions and the fees charged under the program to the duration of employment, since the risk that a temporary worker will displace an unemployed Canadian likely rises according to job duration.

More important, we must avoid the impulse to try to stem the pressures that labour markets, employers and real wages naturally will face as the economy regains full employment and labour supply constrains real growth (Emery 2013). A better option would be to invest in the human capital of Canadians through a “second-chance” system, rather than merely turning to foreign workers as a source of contingent, inexpensive human capital.

Jump-starting a real wage growth/productivity growth circle

The past three decades ought to have brought substantial growth in real wages for average Canadian workers. Never in history has the workforce been so highly educated or has there been such a large cohort of workers — in this case, baby boomers — with such deep labour market experience. Yet real earnings did not rise much over the period, other than for those at the top of the income distribution.

In theory, nominal wages should have been increasing on average about 3 percent a year over the past two decades. This would be the sum of the 2 percent target/expected inflation rate and around 1 percent growth in labour productivity — in other words, 1 percent average real wage growth. Since 1992, however, real wage settlements (net of inflation) have run above 1 percent for more than two consecutive quarters only during periods of excess aggregate demand, such as at the end of the high-tech bubble in 2000-01, briefly in 2005 and again through 2007 and 2008 (figure 17). Overall, it is a portrait of slow gains, rather than no gains. A similar story can be told with respect to median weekly wages.

halliwell figure 17

This slow average real wage growth likely reflects global labour market developments. The global labour market is often thought of in terms of global labour mobility. In fact, global labour mobility is now predominantly the mobility of demand, through production location decisions for the myriad components of global value chains. When countries such as China and India made their entry into the global economy, this amounted to a global labour supply shock. It caused massive changes in the location of the production of assembled goods and the provision of many services. It also undoubtedly put downward pressure on real wages throughout much of the Western world.

In general, many analysts, including the CPP’s chief actuary, assume the upcoming period of slow labour force growth will lead to faster growth in real wages. Perhaps, but not because aggregate demand growth will exceed potential: the Bank of Canada’s “day job” is to prevent this. At issue is whether relative changes in the structure of demand and supply will drive up real wages. The consensus view is that this will happen as part of population aging, because population growth will exceed labour force growth; the former fuels demand while the latter fuels potential.

What will happen to the global labour supply?

An issue not yet settled is whether such upward pressure on Canadian real wages could translate into more offshoring of production elsewhere. Generally, I am skeptical:36 much of what can migrate likely already has and much that has not cannot, because it requires proximity to the customer. Moreover, it is not clear how large the global surplus of labour will be over the next decade. China, due to its demographic policies, has a rapidly aging population that will eventually decline, which would only add pressure on already-rising domestic wages starting around the end of the next decade, according to the latest United Nations projections (2010). Other developing regions, such as India or sub-Saharan Africa, however, are still a long way away from such workforce peaks. These areas boast very large youth populations anxious to be part of the global labour supply. The ability of these countries to “activate” hundreds of millions of new workers will depend very much on how they manage structural and demographic challenges that make those in the Western world seem trivial. This must be the subject of further research: it is crucial to the assumption that Western demographic trends will lead to ongoing, ideally widespread real wage gains.

Interestingly, though, prior to the 2008-09 recession, observers often raised two fears: first, that the West would not have enough workers and, second, that the rest of the world had too many. Had Canada not experienced a significant loss of manufacturing jobs well before that recession, however, workers would have been even scarcer in other sectors and demand growth would have been constrained (by tighter monetary conditions) even earlier. When the Western world returns to a long-term state of full employment, further outsourcing of low-paid, labour-intensive work to the developed world would make sense.

The productivity challenges ahead

These demographics also might spur a return to faster labour productivity growth. Certainly, many claim that slower labour force growth will lead to higher productivity.

In Canada, analysts and policy-makers have been worrying about productivity growth since it began to slow in the mid-1970s. Despite extensive research, however, we still do not fully understand what drives it; indeed, we are now even more mystified, as the United States has continued to achieve high labour productivity growth throughout the recession and the subsequent recovery. While it is known that a large part of the productivity gap between Canada and the United States is due to much lower investment in technology in Canada than south of the border, there remains a significant part of this gap for which the causes are still unknown (Rao 2011).

Certainly, during the recession and its aftermath, slower labour productivity growth has not been such a “bad thing”: it arguably has helped support higher employment growth since the recession. Looking ahead, however, it could become a “bad thing”: once Canada regains full employment, its growth prospects are modest — barely north of 2 percent — unless labour productivity growth improves.

Over the past decade and a half, Canada has implemented much of what the experts could call a “productivity agenda,” but there is as yet no evidence of sustained improvement in productivity growth. Nor have we seen real wage growth even commensurate with this productivity growth (Sharpe, Arsenault and Harrison 2008). What is left to do? Perhaps it is time for real wages to grow first, making capital and technology relatively less expensive and therefore more attractive for firms seeking to adjust and become more productive.

Such a strategy would not be easy to implement. In the public and quasi-public sector — especially core public administration, education and health care — it is difficult to measure productivity, but there are reasons to suspect the record is poor. As a result, Canada could begin to face a version of Baumol’s “cost disease”37 in these segments of the labour market as wage pressures take hold across the board. The private sector would face challenges, too. Many firms would become more productive, but others would face lower profitability and declining market share. Some would fail. This reallocation of scarce resources (labour and capital) through changing market share is a key ingredient in increasing productivity (Gellatly and Baldwin 2009)

halliwell figure 18

The policy challenge will come when employers confront these labour market pressures (this is already happening, according to some). This might lead to renewed calls to admit more immigrants and TFWs — relying on the latter is formally sanctioned when domestic workers are not available at the “going” wage. It might be best to mostly ignore those cries in the hopes that upward wage pressures will take hold. Here, though, cultural change might also be required of employers. If a recent survey of executives in Canada’s ICT sector is any indication (figure 18), raising wages is not top of mind as a strategy for addressing labour market pressures. While it is positive to see that training is being pursued aggressively, it is also notable that turning to alternative sources of labour rated higher than raising wages.

Ultimately, if we allow upward pressures on wages to unfold, it could spur a virtuous circle of real wage/productivity growth. This would be especially beneficial from a social perspective if it compressed the distribution of wages by raising the relative wages of those at the lower end.


Canada is already beginning to experience a labour market quite different from that of the past several decades: one of very slow growth of the potential labour force. Combined with several decades of lagging productivity growth, this development must provoke fresh thinking about the current framework of labour market policies, institutions and arrangements across Canada.

Over the short term, the lingering fallout from the most recent recession still requires policy attention. Although Canada weathered the downturn relatively well compared with other countries, it has still seen five consecutive years of inadequate labour demand with potential scarring effects on new labour market entrants (young Canadians and recent immigrants). These secondary effects might linger longer. If nothing else, we should not increase immigration intakes or TFW levels until we are more confident we have returned to full employment. In retrospect, we should also reduce immigration inflows during recessions. We have learned that recessions are not a thing of the past, and that the resulting high unemployment can often last for years after the recession is officially declared over.

Looking beyond these short-term concerns, some of what is happening is new, some is not. First, and not new, is that the labour market will continue to have a healthy inflow of young Canadians in search of their first careers. Some analysts claim that immigration is the only source of labour force growth now. They are wrong. The largest source, by an order of magnitude, is still young Canadians going through the school system and then entering the workplace. The top priority for policy is still to ensure that our “first-chance” K-12 and PSE systems work well, providing those new labour market entrants with the skills and knowledge needed to succeed in the labour market. If we fall behind in supporting the human capital formation of young Canadians, focusing on just quantity rather than quality, the effects could be serious and long-lasting.

That does not mean that we need to push for greater numbers of young Canadians to get PSE. We seem to have got the balance largely right in the past few decades, and likely do now. It is true that about two-thirds of job openings — the newly created jobs plus the many existing jobs vacated by retirees — will require PSE, but about two-thirds of new labour market entrants already have such education. Moreover, many jobs still do not require PSE, even though they are increasingly being filled by workers with (sometimes considerable) PSE. Therefore, we have to be careful not to drift into oversupplying the Canadian labour market with expensive, PSE-equipped young Canadians or immigrants. From a policy perspective, all we need to do at present is monitor this balance through instruments such as the National Graduates Survey.

We do, however, need to pay more attention to what young Canadians are studying. Without entering into the perennial debate about whether PSE should be about job preparation or something more ambitious, there is no doubt that, when PSE does not prepare young people for the world of work, those who find themselves unprepared feel short changed. Addressing this problem requires better labour market information and a highly flexible PSE system, so that students can be better informed and are better able to enter the fields of study they think will best prepare them for their envisaged careers.

Another thing that is not new is that Canada will continue to have a strong inflow of new immigrants. What needs to be new is ensuring better outcomes for new immigrants. Here, Canada has been making many of the selection-process changes that research suggests will contribute to better outcomes. The catch is that these will affect outcomes rather slowly, as they operate only on the minority of new immigrants who are principal applicants within the economic class. They will do nothing for immigrants who are already here and who have -work-skills deficiencies, especially those relating to language skills. Their skills development needs must be addressed squarely. If the payoff is anywhere near that derived from our investing in the human capital of the Canadian-born, it will be worthwhile.

What is new is the substantial cohort of older workers reaching the last few years of their working lives and then transitioning into retirement. The net result will be a relative balance between the labour market inflows of what I have called “starters” and the outflows of “enders,” so that the domestic labour force will not grow much on a net basis. This is inevitable. Prolonging the labour market participation of older Canadians will attenuate this slowdown somewhat, but only in terms of when, not if, it happens.

People are already working longer. There are many reasons for this — the most important likely being concerns about retirement income security in the face of the triple whammy of greater longevity, lower investment returns and greater “individualization of risk.” Older Canadians would benefit from being better informed about their retirement income prospects, in order to plan both how long they will work and how much they should invest in upgrading their skills while still working. Moving ahead with pension reform is critical to addressing both of these goals; policy-makers would be wise to focus on solutions that both improve the adequacy of retirement income (and hence savings) and encourage later retirement.

We also need to rethink our active labour market measures for older workers who become unemployed at later stages of their careers. We should consider, and perhaps pilot, wage subsidies or some form of wage insurance, particularly in those sectors of the economy where job loss results in substantial wage loss even when workers are reemployed. That said, we must not assume that upcoming cohorts of unemployed older workers will have the same experience as those in past decades; they are much more highly educated than previous cohorts of older workers and might be able to reskill more easily.

Looking at active labour market measures more generally, we must also devote greater and better efforts to evaluating what works for whom and when. The results, in turn, should lead to appropriate ongoing changes in the portfolio of measures. This is critical, as fiscal and health cost pressures squeeze available budgets in the years ahead.

What also ought to be new is an effort to create a better “second-chance” system for those whose path through Canada’s admittedly good “first-chance” system of education, from primary to post-secondary, has not turned out so well. No amount of labour market information can protect everyone from bad outcomes, nor can a more flexible PSE system, particularly in the context of rapid changes in the global economy. A “second-chance” system would offer what it says: a second chance. In this respect, the system should be structured appropriately to accommodate the needs of different groups, including those who are older and those who are managing a number of significant responsibilities, such as families and mortgages.

A “second-chance” system would help fill a critical gap that extends across our current PSE system, workplace training and unemployment support services. These systems are not well suited to helping those who are already employed and wish to upgrade their skills beyond the immediate requirements of their current job. Indeed, too much of our PSE system is aimed at those who start and complete a formal curriculum, rather than those who need to fill in some gaps or upgrade specific skills. Access to a comprehensive skills-upgrading system would provide valuable opportunities, particularly for Canadians in low-skill jobs that typically provide little or no workplace training of any kind. To the extent that such a system reduces the supply of lower-skilled labour, greater scarcity actually could improve overall income prospects for this group of workers. Immigrants, many of whom face specific gaps in the skills or credentials required to be able to integrate into the Canadian labour market, also would benefit from a “second-chance” system.

The system I describe is not exactly the Canada Job Grant — a worthwhile initiative that, if done well, has as its focus to design better short-term supports as part of re-employment efforts. A “second-chance” system, by contrast, is more concerned with helping people who are already employed get access to more extensive training and education, which is often needed to move up the skills ladder.

Many of the components of a “second-chance” system are already in place. But to achieve its objectives will likely require a move away from the usual menu of “low-dose” training, which does little to enhance human capital and tends to condemn low-skilled workers to frequent spells of unemployment and low income. Again, the payoff of such a system could be high, but would have to be demonstrated with rigorous evaluation. Another key component would be to have an employment insurance system that better supports long-tenured workers in large urban areas who are affected by major structural change.

What is also new is that, no matter what happens, labour force and GDP growth will slow: 2.5 percent annual growth in real GDP, or less, will become the new norm, if it is not already.38 Importantly, however, this is a slowing down, not a decline, in growth. Slower labour-force growth will constrain growth in potential GDP and, in turn, in aggregate demand. This will not mean “shortages” of millions of workers. The Bank of Canada would never let aggregate demand exceed potential GDP by a wide enough margin to create large aggregate imbalances between labour demand and labour supply.

Over the next decade, some parts of Canada that are now experiencing below-average rates of population growth will actually see absolute declines. This will be a new phenomenon, and it will be most prevalent in rural areas east of the Ottawa River. There, the purpose of making EI less generous might be not so much to entice workers to leave but to fill the full-time job vacancies in those regions. The retraining challenge, though, could be significant.

Slower labour supply growth should also start to put upward pressure on real wages. This pressure might arise as a result of the labour supply growing even more slowly than the total population. This could create an imbalance between labour supply and demand in sectors not normally known for strong productivity gains, such as health care. Another factor contributing to the upward pressure on wages is that demographic trends in several emerging economies, particularly China, likely will lessen the extent to which firms can address demand pressures through offshoring.

Some describe this pressure on real wages as a “risk” to businesses, but for workers it sounds like a “hope.” This push for real wage increases, in turn, might be the impetus for the technology, machinery and equipment investments and production changes necessary to achieve faster growth in labour productivity in Canada. For decades now, we have assumed the productivity “chicken” precedes the real wage “egg” — that we need to increase productivity growth so that real wages can grow faster. It could be that the chicken needs the egg: higher real wages to spur productivity-enhancing investments. This is more than a question of semantics: improving our productivity performance will be necessary just to maintain, let alone increase, Canadians’ standard of living in the coming years.

Perhaps the one area where there could be more of a policy debate is the role that immigration plays as population policy. Many people see more immigration as the way to grow the economy faster, which is true if the metric of success is total GDP. If the metric of success is GDP per person, more GDP generated through more immigration likely would have little impact one way or another. In the end, if slowing labour force growth does put broad and sustained upward pressure on real wages, employers might cry out for increased numbers of immigrants or TFWs, while workers might prefer higher real wages. We should not use immigration as a means to keep a lid on real wages, especially among the lower-skill occupations in the services sector. This is essentially, then, a choice between 1) growing the economy by growing its population, and 2) growing the economy by having labour market pressures translate into faster gains in real wages and greater incentives for firms to boost productivity.

Both options will grow GDP. The first will not grow GDP per capita. The second could.

In the end, I am not setting this out as a binary choice between more immigration and higher real wages. Rather, it is a sequencing issue: let real wages rise before letting immigration (permanent or temporary) rise further. Should that happen, the coming decade could be one of solid real wage and productivity gains for Canadian workers. That would be something truly different.

*A previous version of this study included incorrectly marked labels on figures 5 and 6.

  1. As part of changes to the federal ministry announced by the Prime Minister on July 15, 2013, HRSDC was renamed Employment and Social Development Canada (ESDC), although to date legislation has not yet been adopted. For clarity, I refer throughout this study to HRSDC.
  2. That said, these projections still showed that each factor would continue to grow, just at a slower pace. Alarmist claims that population aging will lower GDP per capita are very misleading.
  3. However, the results also showed that because of the oncoming tidal wave of workers aging, this would only modestly attenuate the slower growth of the labour force and GDP per capita over time.
  4. Other studies, however, have suggested greater scope; see, for example, Baker, Gruber and Milligan (2001).
  5. It is also important to note that the Bank of Canada had started easing monetary conditions well before the onset of the recession in Canada.
  6. Between 2008 and 2012, the total number of landed immigrants who were unemployed grew from 262,800 to 339,300 (Statistics Canada, CANSIM 282-0102).
  7. The unemployment rate for immigrants ages 25 to 54 was 6.8 percent in 2008; for workers born in Canada, it was 4.6 percent. The rates rose to 7.9 and 5.4 percent, respectively, by 2012 (Statistics Canada, CANSIM database 282-0102). These differences in labour market outcomes also result in income gaps. According to the 2011 National Household Survey, the median income of employed immigrants ages 25 to 54 ($35,451) was approximately 85 percent of that reported by Canadian-born workers ($41,700) (Statistics Canada, National Household Survey, Statistics Canada Catalogue no. 99-014-X2011041).
  8. Picot and Sweetman report that whereas male immigrants arriving in Canada during the late 1970s had annual earnings that were roughly 85 percent of those of their Canadian-born counterparts within their first five years, that figure had fallen to 60 percent for those arriving in the early ‘90s and early 2000s (2012, 37).
  9. A previous version of this study included incorrectly marked labels on figures 5 and 6.
  10. Note that one of the reasons why the “starters” are fewer than the “PSEers” is that they encompass five birth years, while the PSEers encompass seven years.
  11. A good discussion of this is McMullin, Cooke and Downie (2004).
  12. Perceived financial security is a very influential factor in how and when people choose to take retirement. Although this is an area where research is lacking, given what we know about retirement planning more generally it is likely true that financial considerations play an important role in decisions about reentering the labour market.
  13. However, work can be a significant source of stress; see Uriarte-Landa and Hébert (2009).
  14. These have been fuelled by stock market declines in the recession, low interest rate returns, ongoing gains in longevity and the “individualization of risk” inherent in the spread of defined-contribution plans and self-funding for retirement.
  15. The one problem with the impact of adverse wealth effects, such as stock market crashes, is that they often coincide with a weak economy and poorer prospects for working longer (Bosworth and Burtless 2010).
  16. A significant percentage of defined-benefit pension plans is linked to the retirement provisions set by the CPP/QPP.
  17. These projections also showed that approximately three-quarters of new jobs created in the future will require PSE. It is important to remember that PSE includes more than university: the trades and college are key components with possibly high rates of return (Boothby and Drewes 2004).
  18. It should also be noted that the percentage of “enders” with completed PSE has risen dramatically in past two decades, as they are now the baby-boom generation.
  19. Between 2008 and 2012, on average 72 percent of immigrants in the labour force who arrived in Canada within the preceding five-year period had completed PSE (Statistics Canada, CANSIM 282-0106).
  20. The United States had seen the earnings premium on PSE soar prior to the recession, likely simply reflecting slower growth in the supply rather than in the demand for high-skilled labour.
  21. The most recently available data (as of mid-2013) are for the class of 2005, covering their labour market experience as of 2007. New data are to be released later this year, although this is clearly not the kind of timely snapshot one requires to provide effective labour market information for policy-making and individual decision-making.
  22. Readers merely should ponder the divergent careers of those with whom they graduated.
  23. There is much heterogeneity: microdata show that many new immigrants quickly converge on Canadian standards of earnings and living, but a lot never do. Of course, there is still an element of outright discrimination involved (Oreopoulos 2011).
  24. In 2012, spending on settlement services was forecast at $975 million; the 2011 inflow of landed permanent residents was 248,748 (Citizenship and Immigration Canada 2012a,b). It is important to note that immigrants continue to qualify for settlement supports for a number of years after arriving. Thus, actual per capita spending is likely lower than this figure. Citizenship and Immigration Canada does not, however, publish data on the number of clients who access settlement supports in a given year as part of its Departmental Performance Report, making it difficult to estimate this number with accuracy.
  25. Earlier this fall, Employment and Social Development Minister Jason Kenney suggested that provinces could potentially also pay for the Canada Job Grant using monies from the much larger pool of federal transfers under the Labour Market Development Agreements, although this is subject as well to negotiation with provinces (Curry 2013).
  26. Marginal effective tax rates measure the percentage of incremental income that is “taxed” due to income taxes and reductions and clawbacks in means-tested benefits (including tax credits) as income rises.
  27. Most estimates show that these reforms had a significant effect on increasing work effort and earnings, while also reducing poverty; see Blank (2002, 2006).
  28. Of the 2.5 million Canadians who received EI in 2011, approximately 657,000 (26 percent) lived outside Census Metropolitan Areas. Among this group, people ages 15 to 29 represented one-quarter of the recipients (CANSIM 111-0019).
  29. See the Canadian Real Estate Association house price index at
  30. A trivial change to the Labour Force Survey would capture this: collect the postal code of the usual place of employment of a worker!
  31. As measured using administrative tax data (see Laporte, Lu and Schellenberg 2013).
  32. These are Statistics Canada’s population projections, in this case medium-growth (M3) projections. Provincial population projections are quite uncertain in the sense that interprovincial migration flows vary considerably over time. Moreover, there seem to have been no successful efforts at modelling them. Statistics Canada sets out several variants on interprovincial migration rates based on propensities in different past periods. Key is that the relative demographic decline of the slower-growing regions is dampened in these projections by using fixed migration rates: as their relative population falls, their out-migration slows relative to in-migration.
  33. Not all TFWs require a labour market opinion (LMO) to work in Canada. In 2012, only 80,615 (38 percent) of the 213,575 TFWs admitted to Canada came under an LMO (Worswick 2013).
  34. This refers to all “positive” LMOs within the year, and includes those for new TFWs admitted as well as for TFWs whose positions continued.
  35. These industries were construction, retail trade, manufacturing, and accommodation and food services (author’s calculations, based on HRSDC Labour Market Opinion Statistics, table 5 (HRSDC 2013a), and Statistics Canada, CANSIM 282-0008).
  36. However, the current competition for low-wage employment in the United States does present a certain challenge.
  37. This is the phenomenon where wages for a labour-intensive occupation rise in response to rising wages elsewhere in the economy but are not accompanied by commensurate increases in productivity in the labour-intensive occupation.
  38. The Bank of Canada now sees potential GDP growth as just over 2 percent (2013).

Abbott, M.G., and C.M. Beach. 2011. Do Admission Criteria and Economic Recessions Affect Immigrant Earnings? IRPP Study. Montreal: Institute for Research on Public Policy.

Advisory Panel on Labour Market Information. 2009. Working Together to Build a Better Labour Market Information System for Canada: Final Report. Ottawa: Human Resources and Skills Development Canada.

Autor, D.H. 2007. “Structural Demand Shifts and Potential Labor Supply Responses in the New Century.” In Labor Supply in the New Century, Conference Series 52, edited by Katharine Bradbury et al. Boston: Federal Reserve Bank of Boston.

Baker, M., J. Gruber and K. Milligan. 2001. “The Retirement Incentive Effects of Canada’s Income Security Programs.” NBER Working Paper 8658. Cambridge, MA: National Bureau of Economic Research.

Bank of Canada. 2012. Monetary Policy Report. Ottawa, October.

—————–. 2013. Monetary Policy Report. Ottawa, January.

Beaudry, P., D.A. Green, and B.M. Sand. 2013. “The Great Reversal in the Demand for Skill and Cognitive Tasks.” NBER Working Paper 18901. Cambridge, MA: National Bureau of Economic Research.

Begin, K., C. Goyette, and W.C. Riddell. 2010. “Revising Canada’s Immigrant Selection System: An Assessment of Recent Changes.” Background paper prepared for Citizenship and Immigration Canada.

Blank, R. 2002. “Evaluating Welfare Reform in the United States.” Journal of Economic Literature 40 (4): 1105-66.

Bonikoska, A., D. Green, and C. Riddell. 2008. Literacy and Labour Position of Canadian Immigrants. Ottawa: Statistics Canada.

—————–. 2006. “Was Welfare Reform Successful?” Economists’ Voice, March: 1-5.

Boothby, D., and T. Drewes. 2004. “Post-Secondary Education in Canada: Returns to University, College and Trades Education.” University of Toronto.  Accessed November 5, 2013.

—————–. 2010. “The Payoff: Returns to University, College and Trades Education in Canada, 1980 to 2005.” e-brief. Toronto: C.D. Howe Institute, August 24.

Bosworth, B.P., and G. Burtless. 2010. “Recessions, Wealth Destruction, and the Timing of Retirement.” Center for Retirement Research Working Paper 2010-22. Washington, DC: Brookings Institution.

Boudarbat, B., and V. Chernoff. 2010. “The Determinants of Education-Job Match among Canadian University Graduates.” CIRANO Working Paper 2010s-14. Montreal: Centre for Interuniversity Research and Analysis of Organizations.

Browarski, S. 2007. “From Baby Boom to Labour Crunch: Quebec’s Impending Labour Shortage.” Executive Action Report. Ottawa: Conference Board of Canada.

Busby, C., and D. Gray. 2011. “Mending Canada’s Employment Insurance Quilt: The Case for Restoring Equity.” Backgrounder 144. Toronto: C.D. Howe Institute.

Bussiere, P., F. Cartwright, and T. Knighton. 2004. Measuring Up: Canadian Results of the OECD PISA Study: The Performance of Canada’s Youth in Mathematics, Reading, Science and Problem-Solving, 2003 Findings for Canadians Aged 15. Catalogue no. 81-590-XPE. Ottawa: Statistics Canada.

Canada. 2013. “Canada Job Grant.” In Canada’s Economic Action Plan 2013. Ottawa. Accessed November 3, 2013.

Carr, J. 2008. “Bridge Payments, Phased Retirement and Employment Choices.” Human Resources and Skills Development Canada, Policy Research Directorate.

Carrière, Y., and D. Galarneau. 2011. “Delayed Retirement: A New Trend?” Perspectives on Labour and Income 23 (4): 1-16.

Centre for the Study of Living Standards. 2013. “Aggregate Income and Productivity Trends: Canada vs. United States.” Database. Ottawa. Accessed November 6, 2013.

Citizenship and Immigration Canada. 2012a. “Canada Facts and Figures: Immigration Overview, Permanent and Temporary Residents, 2011.” Ottawa.

—————–. 2012b. “Report on Plans and Priorities: 2012-13.” Ottawa.

—————–. 2013. Canada Facts and Figures: Immigration Overview, Permanent and Temporary Residents, 2012. Ottawa.

Commission for the Review of Social Assistance in Ontario. 2012. Brighter Prospects: Transforming Social Assistance in Ontario. Toronto.

Commission on the Reform of Ontario’s Public Services. 2012. Public Services for Ontarians: A Path to Sustainability and Excellence. Toronto: Ontario Ministry of Finance.

Conference Board of Canada. 2012. Canada’s Long-Term Economic Forecast: 2012. Ottawa.

Curry, B. 2013. “Ottawa Offers Changes to Sway Provinces on Canada Jobs Grant”  Globe and Mail, October 8, 2013.

Denton, F.T., and B. Spencer. 2009. “Population Aging, Older Workers, and Canada’s Labour Force.” Canadian Public Policy 35 (4): 481-92.

Desjardins, R. 2011. “Literacy for Life: Further Results from the Adult Literacy and Life Skills Survey: Second International ALL Report.” Ottawa: Statistics Canada.

Drewes, T. 2008. “Adult Education and Education from a Longitudinal Perspective.” Ottawa: Human Resources and Skills Development Canada, Strategic Policy and Research Branch.

—————–. 2010. “Postsecondary Education and the Labour Market in Ontario.” Toronto: Higher Education Quality Council of Ontario.

Drummond, D., and F. Fong. 2010. “An Economics Perspective on Canadian Immigration.” Policy Options 31 (7): 28-34. Montreal: Institute for Research on Public Policy.

Duclos, J.-Y. 2007. “A Better Income Security System for All Canadians.” In A Canadian Priorities Agenda: Policy Choices to Improve Economic and Social Well-Being, edited by J. Leonard, C. Ragan, and F. St-Hilaire. Montreal: Institute for Research on Public Policy.

Emery, H. 2013. “Labour Shortages in Saskatchewan.” Calgary: University of Calgary, School of Public Policy.

Estevao, M., and E. Tsounta. 2010. “Canada’s Potential Growth: Another Victim of the Crisis?” Washington, DC: International Monetary Fund.

Expert Committee on the Future of the Quebec Retirement System. 2013. Innovating for a Sustainable Retirement System: A Social Contract to Strengthen the Financial Security of All Quebec Workers. Summary Report. Montreal: Régie des Rentes du Québec.

Expert Panel on Older Workers. 2008. Supporting and Engaging Older Workers in the New Economy. Ottawa: Human Resources and Skills Development Canada.

Ferrer, A.M., and A. Menendez. 2009. “The Returns to Flexible Postsecondary Education: The Effect of Delaying School.” CLSRN Working Paper 20. Vancouver: Canadian Labour Market and Skills Researcher Network.

Ferrer, A.M., G. Picot, and W.C. Riddell. 2012. “New Directions in Immigration Policy: Canada’s Evolving Approach to Immigration Selection.” CLSRN Working Paper 107. Vancouver: Canadian Labour Market and Skills Researcher Network.

Finance Canada. 2012. “Economic and Fiscal Implications of Canada’s Aging Population.” Ottawa. Accessed June 27, 2013.

Finnie, R. 2002. “Early Labour Market Outcomes of Recent Canadian University Graduates by Discipline: A Longitudinal, Cross-Cohort Analysis.” Analytical Studies Branch Research Paper 164. Ottawa: Statistics Canada.

Finnie, R., and M. Frenette. 2003. “Earnings Differences by Major Field of Study: Evidence from Three Cohorts of Recent Canadian Graduates.” Economics of Education Review 22 (2): 179-92.

Finnie, R., and D. Gray. 2011. Labour-Force Participation of Older Displaced Workers in Canada: Should I Stay or Should I Go? IRPP Study 15. Montreal: Institute for Research on Public Policy.

Finnie, R., and R. Mueller. 2008. “The Effect of Family Income, Parental Education and Other Background Factors on Access to Post-Secondary Education in Canada: Evidence from the YITS.” Toronto: Education Policy Institute.

Fortin, P. 2000. “Macroeconomic Unemployment and Structural Unemployment.” Canadian Public Policy 26 (1): 125-30.

Fougère, M., S. Harvey, Y. Lan, A. Léonard, and B. Rainville. 2009. “Incentives for Early Retirement in Canada’s Defined-Benefit Public and Private Pension Plans: An Analysis with a Dynamic Life-Cycle CGE Model.” In Retirement Policy Issues in Canada, edited by M.G. Abbott et al. Kingston: John Deutsch Institute for the Study of Economic Policy, Queen’s University.

Fougère, M., S. Harvey, J. Mercenier, and M. Merette. 2006. “Population Ageing, Time Allocation and Human Capital: A General Equilibrium Analysis for Canada.” Working paper. Ottawa: Industry Canada.

—————–. 2009. “Population Ageing and the Effective Rate of Retirement in Canada.” Working Paper 2005, A-03, HRSDC-IC-SSHRC Skills Research Initiative. Ottawa.

Fougère, M., S. Harvey, and B. Rainville. 2012. “What Are the Economic and Labour Market Effects of an Income Tax Reduction Targeted at Older Workers?” Economics Research International, Article ID 461597.

Gellatly, G., and J. Baldwin. 2009. “Productivity Spillovers from Competitive Reallocation: Evidence from Canadian Manufacturing Plants.” Canadian Productivity Review. Cat. 15-206-X, no. 024. Ottawa: Statistics Canada.

Gray, D. 2004. “Employment Insurance: What Reform Delivered.” Backgrounder 82. Toronto: C.D. Howe Institute.

Green, D. 2011. “Temporary Foreign Workers and the Election: A Major Issue Getting Scant Debate.” Behind the Numbers: A Blog from the CCPA. April 15. Ottawa: Canadian Centre for Policy Alternatives. Accessed October 23, 2013.

Green, D.A., and C.W. Riddell. 2001. “Literacy, Numeracy and Labour Market Outcomes in Canada.” Ottawa: Statistics Canada and Human Resources Development Canada.

Green, D., and C. Riddell. 2013.  “Ageing and Literacy Skills: Evidence from Canada, Norway and the United States.”  Labour Economics  22 (June): 16-29.  Accessed November 5, 2013.

Gunderson, M. 2003. “Active Labour Market Adjustment Policies: What We Know and Don’t Know.” Research Paper 33. Toronto: Panel on the Role of Government in Ontario.

Gunderson, M., and H. Krashinsky. 2009. “Do Education Decisions Respond to Returns by Field of Study?” CLSRN Working Paper 40. Vancouver: Canadian Labour Market and Skills Researcher Network.

Halliwell, C. 2009. “Retirement in Canada: Some Trends … Some Issues.” In Retirement Policy Issues in Canada, edited by M.G. Abbott et al. Kingston: John Deutsch Institute for the Study of Economic Policy, Queen’s University.

Hicks, P. 2012. “Later Retirement: The Win-Win Solution.” C.D. Howe Institute Commentary 345. Toronto: C.D. Howe Institute.

Horner, K. 2011. A New Pension Plan for Canadians: Assessing the Options. IRPP Study 18. Montreal: Institute for Research on Public Policy.

HRSDC (see Human Resources and Skills Development Canada).

Hum, D., and W. Simpson. 2002. “Public Training Programs in Canada: A Meta-Evaluation.” Canadian Journal of Program Evaluation 17 (1): 119-38.

Human Resources and Skills Development Canada (HRSDC). 2010. “Presentation on Labour Demand: Key Conclusions.” Strategic Policy Research Directorate.

—————–. 2011. Labour Market Research and Forecasting, Policy Research Directorate. “Canadian Occupational Projection System: Macroeconomic Scenario, 2011-2020.” Ottawa. Accessed November 3, 2013.

—————–. 2012. Labour Market Research and Forecasting, Policy Research Directorate. “Canadian Occupational Projection System 2011 Projections.” Ottawa. Accessed November 3, 2013.

—————–. 2013a. “Labour Market Opinion (LMO) Statistics – Annual Statistics 2012.”  Ottawa. Accessed November 5, 2013.

—————–. 2013b. “Harper Government Announces Reforms to the Temporary Foreign Worker Program — Ensuring Canadians Have First Chance at Available Jobs.” Ottawa, April 29. Accessed November 3, 2013.

Jones, S.R. 2012. “The Effectiveness of Training for Displaced Workers with Long Prior Job Tenure.” CLSRN Working Paper 92. Canadian Labour Market and Skills Researcher Network.

Knighton, T., F. Hujaleh, J. Iacampo, and G. Werkneh. 2009. “Lifelong Learning Among Canadians Aged 18 to 64 Years: First Results from the 2008 Access and Support to Education and Training Survey.” Statistics Canada Catalogue no. 81-595-M, No. 079. Culture, Tourism and the Centre for Education Statistics Research Papers. Ottawa.

Kuhn, P., and C. Riddell. 2006. “The Long-Term Effects of Generous Income Support Program: Unemployment Insurance in New Brunswick and Maine, 1940-1991.” NBER Working Paper 11932. Cambridge, MA: National Bureau of Economic Research.

Kustec, S. 2012. “The Role of Migrant Labour Supply in the Canadian Labour Market.” Ottawa: Citizenship and Immigration Canada.

Lapointe, M., K. Dunn, N. Tremblay-Côté, L.-P. Bergeron, and L. Ignaczak. 2006. “Looking Ahead: A 10-Year Outlook for the Canadian Labour Market (2006-2015).” Ottawa: Labour Market and Skills Forecasting and Analysis Unit, Strategic Policy Research Directorate, Human Resources and Skills Development Canada. Accessed November 3, 2013.

Laporte, C., Y. Lu, and G. Schellenberg. 2013. “Inter-Provincial Employees in Alberta.” Statistics Canada Catalogue no. 11F0019M. Ottawa: Social Analysis Division.

Lefebvre, P., P. Merrigan, and P.-C. Michaud. 2012. L’évolution récente des comportements de retraite au Canada. In Le vieillissement démographique: de nombreux enjeux à déchiffrer, edited by S. Rheault and J. Poirier. Québec: Institut de la statistique du Québec.

Lehrer, S.F. 2013. “Postsecondary Student Choices and Skill Shortages in Canada’s ICT Sector.” Paper prepared for the Information and Communications Technology (ICT) Talent Workshop, Education Policy Research Initiative, University of Ottawa, March 21-22.

Lin, Z., and J. Chen. 2010. “The Evolution of Earnings Returns to Education in Canada, 1981 to 2001: Empirical Evidence from the Census.” Policy Research Directorate, Human Resources and Skills Development Canada.

Lin, Z., J. Chen, and H. Qiu. 2009. “Education, Experience and Earnings in Canada: Empirical Evidence from an Extended Mincer Human Capital Model.” Policy Research Directorate, Human Resources and Skills Development Canada.

Macdonald, A. 2006. “Alberta’s Labour Shortage: Just the Tip of the Iceberg.” Executive Action Report. Ottawa: Conference Board of Canada.

Malamud, O. 2011. “Discovering One’s Talent: Learning from Academic Specialization.” Industrial and Labor Relations Review 64 (2): 375-405.

Mavromaras, K. 2013. “Measuring Skill Imbalances in Times of Change: Recent Evidence from the Australian ICT Sector.” Adelaide, Australia: National Institute of Labour Studies.

McMullin, J., M. Cooke, and R. Downie. 2004. “Labour Force Aging and Skill Shortages in Canada and Ontario.” Ottawa: Canadian Policy Research Network.

Mendelsohn, M., and J. Medow. 2010. “Help Wanted: How Well Did the EI Program Respond during Recent Recessions?” Toronto: School of Public Policy and Governance, University of Toronto, Mowat Centre for Policy Innovation.

Mendelsohn, M., and N. Zon. 2013. “The Training Wheels Are Off: A Closer Look at the Canada Job Grant.” Toronto: School of Public Policy and Governance, University of Toronto, Mowat Centre for Policy Innovation and Caledon Institute of Social Policy.

Millway, J., K. Chan, and J. Stapleton. 2009. “Time for a ‘Made in Ontario’ Working Income Tax Benefit.” Toronto: Institute for Competitiveness and Prosperity.

Miner, R. 2010. “People without Jobs, Jobs without People: Ontario’s Labour Market Future.” Toronto: Miner Management Consultants.

Mintz, J., and T.A. Wilson. 2013. “Reform Proposals for Replenishing Retirement Savings.” SPP Research Papers, 6 (9). Calgary: School of Public Policy, University of Calgary.

Morneau, B., and F. Vettese. 2013. “What Retirement Crisis? Share the Risk, Bridge the Gap. Globe and Mail, June 24.

Nordicity. 2012. “Labour Supply/Demand Dynamics of Canada’s Information and Communications Technology (ICT) Sector: Final Report.” Report prepared for Industry Canada. Toronto: Nordicity.

Oreopoulos, P. 2011. “Why Do Skilled Immigrants Struggle in the Labor Market? A Field Experiment with Thirteen Thousand Resumes.” American Economic Journal: Economic Policy 3 (4): 148-71.

Oreopoulos, P., and R. Dunn. 2012. “Information and College Access: Evidence from a Randomized Experiment.” NBER Working Paper No. 18551. Cambridge, MA: National Bureau of Economic Research.

Oreopoulos, P., T. von Wachter, and A. Heisz. 2008. “The Short- and Long-Term Effects of Graduating in a Recession: Hysteresis and Heterogeneity in the Market for College Graduates.” Bonn, Germany: Institute for the Study of Labour.

Percival, J., B. Cozzarin, and S. Formaneck. 2013. “Return on Investment for Workplace Training: The Canadian Experience.” International Journal of Training and Development 17 (1): 20-32.

Picot, G., and A. Sweetman. 2005. “The Deteriorating Economic Welfare of Immigrants and Possible Causes: Update 2005.” Analytical Studies Branch Research Paper 262. Ottawa: Statistics Canada.

—————–. 2012. Making It in Canada: Immigration Outcomes and Policies. IRPP Study 29. Montreal: Institute for Research on Public Policy.

Prince, M.V. 2013. “People with Disabilities and Employment: If Canada’s Aging Population Is a Dark Cloud, Is the Social Economy a Silver Lining?” Keynote Address to the Association of Nonprofit and Social Economy Research, 2013 Congress of the Humanities and Social Sciences, Victoria, BC.

Rao, S.  2011. Cracking Canada’s Productivity Conundrum. IRPP Study 25. Montreal: Institute for Research on Public Policy.

Restuccia, R., and C. Urrutia. 2004. “Intergenerational Persistence in Earnings: The Role of Early and College Education.” American Economic Review 94 (5): 1354-78.

Riddell, C.W. 2003. “Why Is Canada’s Unemployment Rate Persistently Higher Than That in the US?” Address to the 2003 Canadian Economics Association, Carleton University, Ottawa.

Riddell, C. 2011. Unemployment Compensation and Adjustment Assistance for Displaced Workers: Policy Options for Canada. Vancouver: Canadian Labour Market and Skills Researcher Network.

Schellenberg, G., and G. Ostrovsky. 2008. “The Retirement Plans and Expectation of Older Workers.” 2007 General Social Survey Report. Ottawa: Statistics Canada.

Sharpe, A., J.-F. Arsenault, and P. Harrison. 2008. “The Relationship between Labour Productivity and Real Wage Growth in Canada and OECD Countries.” Ottawa: Centre for the Study of Living Standards.

Sloat, E., and J.D. Willms. 2000. “The International Adult Literacy Survey: Implications for Canadian Social Policy.” Canadian Journal of Education 25 (3): 218-33.

Stark, A. 2007. “Which Fields Pay and Which Fields Don’t? An Examination of the Returns to University Education in Canada by Detailed Field of Study.” Department of Finance Working Paper 2007-03. Ottawa: Finance Canada.

Statistics Canada. 2005. “Building on Our Competencies: Canadian Results from the International Adult Literacy and Skills Survey.” Ottawa.

—————–. 2007. Labour Market Statistics Division. “The Canadian Labour Market at a Glance.” Ottawa.

—————–. 2010. “Population Projections for Canada, Provinces and Territories.” Ottawa.

—————–. 2013. “Job Vacancies.” Ottawa, February 19. Accessed November 3, 2013.

Sun Life Financial. 2013. Sun Life Canadian Unretirement Index: 2013 Unretirement Index Report. Toronto.

United Nations. 2010. World Population Prospects, the 2010 Revision. New York: Population Division, United Nations.

Uriarte-Landa, J., and B.-P. Hébert. 2009. “Work-Life Balance of Older Workers.” Perspectives on Labour and Income 10 (10): 17-27.

Warburton, W.P., and R.N. Warburton. 2002. “Measuring the Performance of Government Training Programs.” Commentary 165. Toronto: C.D. Howe Institute.

Willms, J.D., and T. Murray. 2007. “Gaining and Losing Literacy Skills Over the Lifecourse.” Ottawa: Statistics Canada.

Wolfson, M.C. 2011. Projecting the Adequacy of Canadians’ Retirement Incomes: Current Prospects and Possible Reform Options. IRPP Study 17. Montreal: Institute for Research on Public Policy.

———. 2013. Not-So-Modest Options for the Expanding the CPP/QPP. IRPP Study 41. Montreal: Institute for Research on Public Policy.

Wood, D. 2013. “Comparing Employment Policy Governance Regimes in Canada and the European Union.” Canadian Public Administration 56 (2): 286-303.

Worswick, C. 2004. “Immigrants’ Declining Earnings: Reasons and Remedies.” Backgrounder 81. Toronto: C.D. Howe Institute.

———. 2013. Economic Implications of Recent Changes to the Temporary Foreign Worker Program. IRPP Insight 4. Montreal: Institute for Research on Public Policy.

Focus on skills, not the number of workers, as workforce ages

Montreal – Employers and policy-makers should view the significant slowdown in the growth of the labour force expected in the next 10 years as an opportunity rather than a threat, says a new study from the IRPP. If Canada improves the development and use of skills across the labour force, the decade ahead could actually be one of rising wages and productivity, which have long been elusive. Such a turnabout is imperative if we are to maintain a high standard of living in the context of an aging population.

In the study, Cliff Halliwell, former director general of policy research at Human Resources and Skills Development Canada, paints a broad portrait of the labour market issues Canada is likely to encounter over the next decade, providing an updated framework for labour market policy in Canada.

Halliwell finds that despite widespread concerns from employers about impending labour and skills shortages as baby boomers retire, the long-term prospects for Canada’s labour supply and demand appear balanced, and ongoing shortages are unlikely. “While labour-force growth will slow, this demographic transition is something to adapt to, not necessarily something to offset. The key challenge is to ensure we effectively develop, match and use the skills of Canadians.”

His analysis also calls into question recent conventional wisdom that says all future labour-force growth will come from immigration. “The place to look for new skills will continue to be the pool of young people ages 25-29, who between now and 2030 will average approximately 2.5 million, far greater than the annual intake of immigrants.”

“Our focus must be on making sure young people get off to a good start in the labour market, and helping people already in the workplace who want to enhance their skills and opportunities while remaining employed.”

He notes that while slowing labour-force growth should put upward pressure on wages and increase efforts to improve productivity, a new policy framework is needed to help make this happen. He recommends these measures:

  • Strengthening and formalizing the “second-chance” system to help Canadians upgrade their education or skills, or get better jobs, over the course of their careers;
  • limiting temporary foreign worker programs to jobs that are truly “temporary”;
  • reducing immigrant intake during recessions, and increasing investments in settlement supports to enhance the language and occupational skills of newcomers;
  • modernizing labour market measures to discourage the frequent use of employment insurance, better support the needs of long-tenured workers and more effectively target the skills development and training needs of the unemployed; and
  • improving labour market information and increasing the flexibility of post-secondary education so that students can more effectively select fields of study and have better access to trades, apprenticeships and occupations in demand.


No Shortage of Opportunity: Policy Ideas to Strengthen Canada’s Labour Market in the Coming Decade, by Cliff Halliwell, can be downloaded from the Institute’s Web site at

For more details or to schedule an interview, please contact the IRPP.

To receive our monthly bulletin Thinking Ahead by e-mail, please subscribe by visiting our Web site.

The Tory Jobs Program Isn’t Enough

Earlier this month, Canada’s provincial and federal ministers responsible for labour market policy met for the first time in several years. That it has been this long is symptomatic of the many challenges facing labour market policy in Canada. Policy directions are too often ad hoc, made up as discrete problems arise and with inadequate coordination.

The purpose of the meeting was to discuss how federal transfers for training and employment supports will be allocated after March 31, 2014, including Ottawa’s proposed Canada Job Grant. This is important, but given how Canada’s labour market is likely to change over the next number of years, we need a much more comprehensive look at labour market policy.

In a paper just released by the Institute for Research on Public Policy, I look at how Canada’s labour market is likely to change over the next decade and the issues policy-makers and employers should focus on. Over this period, Canada’s workforce will age considerably, labour force growth will slow, and the pressure to improve our productivity performance will only mount.

Claims that labour force growth will only come from immigration are wrong. Canada’s labour force growth has and will continue to be dominated by millions of young Canadians. All that is different now is that millions of older workers will be retiring, bringing the net difference between younger and older workers into closer balance.

This means that when thinking about skills development for the future labour market we have to especially focus on the young and those in the middle of their working lives. Here are several things ministers should think about.

First, we need to make sure our “first chance system,” whereby young Canadians move from PSE to work, performs well. While it works relatively well for many, it could do better. Students need better information, better incentives and better pathways in order to select the best of field of study relative to their skills. Enhancing the quality and dissemination of information on labour market outcomes by field of study and occupation is one very important step.

But, more profoundly, we must also address the perception that any university education is better than a trade program. Combined with greater efforts by provinces to make PSE more flexible for students switching programs, this will help achieve a better match between expectations and reality.

Because the “first chance” system will never get it right for everyone, we also need to ensure there is an effective “second chance system” for those who have had bad outcomes, whether due to bad luck (such as graduating in a recession), bad choices, or simply a desire to obtain a better job. For many in these circumstances, however, the challenge of upgrading skills is one of escaping lower-skill jobs. Because employers are reluctant to invest in developing skills an employee could take elsewhere, these workers face a paradox. To get help, many of our public training programs require you to first be unemployed.

That makes no sense.

Canada needs to step back and do a serious rethink of the training and employment support system. As part of the Canada Job Grant proposal, Minister Kenney has argued we need to “stop doing training for the sake of training.” He’s right.

Yet, what quickly becomes evident in such a debate is that we don’t really know what works and what does not work. To fix this, we have to ask some fundamental questions: Who is best targeted by training measures? What kind of skills are we trying to develop? And how much training is “enough” to ensure long-term success?

Then we need more rigorous evaluation of outcomes, to see if benefits exceed costs. We also need a way to compare evaluation results across interventions. Without concrete evidence of payoff, the growing fiscal pressures from population ageing are likely to leave such programs vulnerable to cuts.

The Canada Job Grant could be a positive step in that direction, if it avoids the traditional focus on short-term training, is better connected to the needs of employers and includes rigorous program evaluation. When an agreement is reached between the provinces and the federal government, the discussions should not end. All partners should then seize the opportunity to discuss a new framework of labour market policy in Canada.

There’s still a lot of work to do.

Cliff Halliwell, Retired Director General, Policy Research, with Human Resources and Skills Development Canada.

No Shortage of Opportunity

Is It Time to Reinvent Employment Insurance?

Employment insurance is one of the most important levers for labour market policy in Canada. With a changing labour market, slowing labour force growth and persistent concerns from employers that Canada’s workforce lacks skills in a number of areas, what is the future role for this program? Is it time to reinvent EI? On January 22, 2014, the IRPP asked these important questions in a panel discussion.

New research from the IRPP provides roadmap for achieving universal pharmacare
New research from the IRPP provides roadmap for achieving universal pharmacare