Vision

You are here

Why We Can’t Retire Early Anymore

By 2029, Old Age Security (OAS) pension will not become payable until age 67. After nearly half a century of steady improvements in government retirement programs, this is the first significant take-away that affects middle-income Canadians. It may not be the last. For the last four decades, demographics and capital markets have worked in our favour, enabling us to enjoy ever-longer periods of retirement. That era appears to be coming to an end. The change to the OAS retirement age is not the cause, it is a symptom.

Canada’s retirement situation in recent years has been a case of the glass being nine-tenths full but described as one-tenth empty. In spite of sub-optimal levels of coverage in pension plans and RRSPs, Canada’s retirees have been doing well for the most part. Poverty rates among seniors are extremely low (see sidebar) and one way or another, most people in their early sixties have found ways to retire with at least adequate retirement income; a majority of them have the same discretionary consumption as when they were working if not more. There are three main reasons for this favourable situation:

  1. Recent retirees accumulated significant wealth through home ownership. They can tap into this wealth by downsizing their home and using the difference to supplement their retirement income.
  2. The recent financial crisis notwithstanding, the capital markets have performed remarkably well since the early 1980s, and this has padded RRSP account balances.
  3. The rapid growth in the size of the labour force since the 1960s exerted pressure on governments and employers to facilitate earlier retirement in an effort to contain the number of unemployed.

These factors largely explain why the average retirement age in Canada dropped from 64.9 in 1976 to 61 in 1999 (see Figure 2). We will explain later on what has happened since 1999.

Unfortunately, the future looks less auspicious. The combination of factors that created this positive result will soon start to work against us.

THE HOUSING MARKET’S BEST DAYS ARE BEHIND US

The last 30 years have been a particularly good time to own a home in Canada. In Toronto for example, house prices rose by 500% between 1980 and 2010 while the CPI rose just 154%. Other cities enjoyed an even bigger percentage increase. Housing equity has contributed significantly to the net assets of most Canadians. The wealth we have locked up in real estate equity, net of mortgages, actually exceeds all of the assets held in the Canada and Quebec Pension Plans, all RRSPs and all registered pension plans combined.

This is all about to change, as it is a near-certainty that house prices will rise much more slowly over the next decade or longer. Consider the danger signs:

  • Price increases in recent years have outpaced income growth by a wide margin. In Canada’s four biggest cities average incomes rose 23% to 32% between 1996 and 2009, depending on the city, while home prices went up between 100% and 200%1.
  • Rents have been rising far more slowly than house prices. OECD figures from 2008 showed that Canada’s house price-to-rent ratio was at an all-time high and was also the second highest among OECD countries. A high price-to-rent ratio is considered to be an important predictor of future price movements in the housing market2. When the ratio strays too far from historical norms, it eventually re-aligns with those norms but does so by downward adjustments to house prices rather than by increases in rents. It is sobering to note that this observation was made in a rather prescient commentary3 on the US housing market in 2004, two years before prices started to fall!
  • The percentage of Canadians owning a home has jumped from 64% in 2000 to 70% today, which means it is now 1% higher than it was in the US when the housing bubble burst in 2006. Amongst the top 40% of households by income level, over 90% own their homes. We are very near the point where there will be few net new buyers ready and willing to push prices higher, assuming we haven’t already reached it.

What allowed house prices to soar was falling mortgage rates. 5-year fixed rate mortgage rates which peaked in 1982 at 18%4, had fallen by 2012 to 2.99%. This drop in interest rates means that the monthly payments for a $100,000, 25-year mortgage dropped from almost $1,500 to under $500! Mortgage rates are to house prices what interest rates are to bonds; as rates fall, prices rise. The decline in mortgage rates has kept a lid on the cost of home ownership even as house prices kept rising to record levels. Those rates are now bottoming out and that affordability is about to become precarious. The average cost of owning a bungalow in Canada in the fourth quarter of 2011 was 42.2% of pre-tax household income5, which equates to 60% or more of after-tax income. It would have been higher yet for a two-storey home. If mortgage rates were to return to their 60-year average of 8.8%, the average cost nationally of owning a bungalow would reach 80% of after-tax income6. In Vancouver and Toronto, the average cost would exceed 100%!

Clearly, market forces would kick in before we ever reached this point. Families that are straining to cope with their mortgage burden would not be able to meet these higher payments and would have to sell and either downsize or rent. The immediate result is downward pressure on housing prices. Secondary results are less money available to put into retirement savings vehicles, and less home equity as a retirement asset.

The worst-case scenario is a housing collapse similar to the US experience in 2008, in which case a recovery could take ten or more years as we learned after the 1989 housing crash here in Canada. The more likely scenario is a prolonged period of flat or very slowly increasing house prices. Either way, home ownership will not be the wealth generator it has been, which will hurt most the people who are transitioning to retirement.

POVERTY AMONG SENIORS

The most commonly used measure of poverty in Canada is the Low Income Cutoff (LICO) as reported by Statistics Canada. The LICO measures the income needed to buy a basket of goods that is deemed to be essential. In 2009, only 5.2% of seniors had income below the LICO compared to 10.5% of Canadians age 18-64. If we use an international measure, the OECD sets the poverty line at 50% of the median national income. On this basis, 5.9% of seniors are below the poverty line compared to 12.0% for the general population. By comparison, highly developed countries like the US, Japan and Switzerland have much higher poverty levels for seniors than for the general population. Canada’s challenges in dealing with poverty affect mostly working-age Canadians, not seniors.

CAPITAL MARKETS WILL BE LESS HELPFUL

The financial crisis that began in mid-2007 and ended in early 2009 was the most devastating event in the capital markets since the Great Depression. In spite of it, the average return before fees for the median pension fund over the 25-year period ending December 31, 2009 was still a healthy 9.1% per annum and compared favourably to inflation which averaged just 2.5% during the period.

Anyone who was making regular deposits into an RRSP or other capital accumulation plan during this period and who was fully invested in stocks and bonds would have done very well. Just as falling mortgage rates propelled housing prices upward, falling interest rates enhanced the returns on both stocks and bonds, especially bonds. The average annual return on long-term Canada bonds over the 30-year period ending 2011 was 11.5% which beat inflation by nearly 9% annually. Most of the 11.5% return stemmed from capital gains on bonds as yields kept dropping. This stunning bond performance cannot repeat any time soon. Having dropped to historic lows, bond yields can only go up or sideways. One theory is that bond yields trend in the same direction for 30-year periods, which signals we are about to embark on a 30-year rise in yields with a corresponding drop in bond prices. Even if that is not the case, we are still likely to see more frequent capital losses on bonds in the years ahead while capital gains will be scarce. Equities, too, are expected to produce smaller real returns as the population ages. A maturing population probably means more modest productivity gains and slower growth in GDP, which some believe will translate into lower real returns on equities. Certainly this is what seems to be happening in Japan and Europe. Most actuaries would agree that the most likely return for a balanced fund will be between 5% and 6.5% per annum, depending on asset mix, with 6.5% reflecting a fairly heavy exposure to equities.

To give an idea of the impact of the expected lower returns within a capital accumulation plan, assume that an RRSP-saver’s salary had been rising steadily with inflation plus promotional increases over the past 30 years and that she is currently earning $100,000 a year. If she contributed 8% of pay to an RRSP the entire time and earned 11.5% annually on her investments, her balance today would be $643,000. If the annual investment return was half that, her balance today would be $327,000.

That is only half the picture. Retirement income will be lower not only because the account balance at the point of retirement will be smaller but also because investment returns after retirement will be lower. Let’s assume that rather than purchase an annuity with the RRSP balance, our now-retired saver transfers her RRSP assets into a Registered Retirement Income Fund (RRIF) in which she continues to hold stocks and bonds. Assume she draws a steady annual income from the RRIF. If she had $643,000 and expected to earn 7% in the RRIF—a figure which would have seemed a little conservative before the financial crisis—her level annual income would be about $52,0007. On the other hand, if she had $327,000 and expected to earn 5.75%—a target which now seems almost aggressive, especially for a retiree—her annual income would be just $23,000, less than half.

This leaves the RRSP–saver and the participant in a DC pension plan with a difficult choice: start saving more than twice as much or retire later. Given the difficulty of saving more when the cost of home ownership is so high, the more likely scenario is that people will save about the same, work longer and retire later.

DEMOGRAPHIC PRESSURES ENCOURAGED EARLY RETIREMENT

In 1961, only 33.4% of women were employed outside the home8. As shown in Figure 1, that percentage climbed dramatically over the next forty years to a recent peak of 67.6%. This phenomenon has had a major impact on the workforce that is sometimes under-appreciated. Had the participation rate remained at the 1961 rate, there would be four million fewer women in the workforce in Canada today.

Figure 1 - Female participation in workforce (ages 15-64)

The other major demographic event that was unfolding at the same time was the coming of age of the baby boomers. The boomers first started entering the workforce in the late 1960s with the last ones coming in by the early 1990s.

Either one of these demographic events by itself would have been significant. Combined, they created a major strain on the economy to absorb such a large influx. Even with the robust economy we enjoyed in the 1960s, we could not possibly create jobs quickly enough to absorb an influx of this magnitude. The inevitable result is that unemployment would start to rise. Table 1 shows that unemployment rates rose markedly in each decade up to and including the 1990s.

Table 1 - Canada's unemployment rate

As a surplus in the labour supply started forming by the 1970s, employers generally wanted to encourage workers to retire earlier in order to avail themselves of the plentiful supply of younger and less expensive workers who were looking for their first jobs. So, apart from the usual goals of attracting and retaining employees, pension plans were in vogue since they made it easier to entice workers to retire earlier, which better suited corporate needs at the time. Hence, the 1980s and early 1990s were a time for early retirement "windows" under which employees were offered financial incentives to retire early.

In their pursuit of coaxing older employees to leave the workforce, employers had the tacit co-operation of organized labour, which was equally keen to secure jobs for new workers, provided of course that older workers were well-treated on the way out. Organized labour negotiated unreduced early retirement benefits through collective bargaining which reinforced the mindset of early retirement even in the absence of an early retirement window.

The government for its part also wanted to see older workers retire to make room for new entrants to the job market since high unemployment was perennially a sensitive election issue and helping workers leave the work force is easier than creating jobs. The federal and provincial governments thus introduced, and then improved, a number of retirement programs to facilitate retirement, including:

  • RRSPs which were introduced in 1957,
  • legislation to regulate registered pension plans, enacted in the mid-1960s,
  • Canada and Quebec Pension Plans, launched in 1966,
  • reducing OAS retirement age to 65, phased in between 1965 and 1969,
  • the introduction of the Guaranteed Income Supplement in 1967, and
  • making 60 the earliest age for pension eligibility under the Canada and Quebec Pension Plans in 1987.

Of course, some of these programs would have been introduced in any event but the demographic pressures contributed in a major way to the political will to make this happen.

As a nation, we therefore enjoyed a win-win-win-win situation as employers, young job-seekers, organized labour and government all benefited from easing older workers out of the work force. The older workers themselves benefited to the extent they were given hefty incentives to retire earlier than they had planned. The net result was a long-term trend toward earlier retirement. Figure 4.2 shows the average retirement age declined steadily for the last quarter of the twentieth century. Later, we will show what has transpired since then.

Figure 2 - Average retirement age in Canada

THOSE DEMOGRAPHIC PRESSURES ARE EASING

By the early 1990s, all of the baby boomers had reached working age and by 2006, the female participation rate in the workforce plateaued. For the first time in forty years, no new identifiable, large group of people was coming into the work force. The pressure on government and on employers to encourage early retirement had finally eased. Surprisingly, the workforce kept on growing at a steady rate during the 2000s, up until the 2009 recession (see Fig. 3).

The answer is that we had a strategic reserve of workers, one that is made up of the ranks of the unemployed. This pool of labour was built up through years of high unemployment. The unemployment rate had more than doubled from the 1950s to the 1990s and when it topped out in 1993, there were over 1.6 million unemployed. As we showed in Table 1, unemployment averaged 9.6% in the 1990s. In the first decade of this century, it has dropped to an average of only 7.0% as our excess labour supply is being drawn down. A 40-year trend of rising unemployment has therefore started to reverse.

Figure 3 - Total labor force (millions)

Aside from three major recessions when job creation was flat or negative, the growth in the labour force has been remarkably linear since 1976. In the period 1993-2008, which was the last prolonged period between recessions, the labour force grew by a steady 280,000 workers a year. But how did the workforce continue to grow at this pace when the two secular forces previously responsible for much of the supply were drying up?

THE NEXT 20 YEARS

We can expect the unemployment rate will continue to decline until it reaches the limits of structural unemployment, which is probably around the 5% level we last saw in the early 1960s. Based on recent trend rates, that should take about ten years barring any major recession. When it happens, we will be in a situation that is outside the living memory of readers: low unemployment and no obvious large pools of labour to draw from. Certainly we can expect immigration to ramp up to some extent but as Table 2 shows, without a change in our retirement age patterns, the best guess is that we will be adding only half as many workers per year in the next thirty years as we did in the last thirty. A crisis is looming, but it is economic in nature rather than the retirement crisis we imagine.

Table 2 - Labour force growth is slowing

When our unemployment rate is low again for the first time in sixty years, it will be a watershed moment that will focus our attention on people in their sixties who are still capable of working. That is because they will represent the one large pool of potential workers that can be deployed. Today many of them choose not to continue working and can afford to retire early given the factors that have been working in their favour, as described earlier. Since this is the only group of potential workers that is large enough to make a difference we can expect that employers and governments will do whatever is necessary to keep them working longer, just as they did everything possible from 1965 to the 1990s to get 60-somethings to leave the workforce. The measures they will take will include enticements such as phased retirement and part-time work that meets the preferences of older workers. As we have recently learned, it will also include takeaways such as later retirement under OAS and perhaps eventually, under the Canada/Quebec Pension Plan.

Other takeaways are possible such as changes in the Income Tax Act to reduce tax deductions that incent earlier retirement, but this is mere speculation at this point. How draconian the government will have to be in its actions depends on the extent of the labour shortages which eventually develop.

ADVICE FOR EMPLOYERS

It is difficult to imagine a labour shortage after several decades of high unemployment but it was equally difficult to imagine 2% annual inflation back in 1990 or 4% yields on long-term bonds. Yet those happened and labour shortages also have a demographic near-certainty of happening in the not-too-distant future. Employers can prepare by starting to modify their pension plans now to ensure the plans are incenting behaviour that best aligns with the interests of their organizations as the labour pool starts to shrink. For instance, many DB plans still have generous early retirement provisions that made eminent sense in the 1980s and 1990s but may be working against the interests of employers in the near future.

We note that one plan design change that encourages later retirement has been happening with great regularity for nearly twenty years now, that being the steady migration from defined benefit (DB) pension plans to defined contribution (DC) plans. DC plans indirectly encourage later retirement because account balances keep growing the longer one puts off retirement and the cost of annuitization shrinks. This factor may be starting to affect retirement ages already. After a steady decline in the average retirement age that lasted until 1999 (Fig. 2), it has been trending modestly upwards ever since, and in 2011 it reached 62.1. For all the reasons given earlier, we can expect the average retirement age to keep rising.

The biggest challenge for employers is to find suitable employment for older workers who would otherwise leave the workforce. This could be problematic even when labour shortages become a reality. Many older workers want to keep working but only on a part-time basis. Most employers can accommodate some part-time positions but they would have to re-organize their operations to absorb many more part-timers and to do so in a way that protects the interests of shareholders. After decades of rising salaries and increasing vacation time, older workers are not always seen as being as cost-effective as their younger counterparts. This is a problem we will need to solve in the years ahead.

WHAT IS CANADA’S SUSTAINABLE RETIREMENT AGE?

With looming labour shortages, our ever-increasing life spans and shrinking returns in the capital markets, all signs point to later retirement. It appears that age 62 is no longer affordable. The question is, how much later? In our next Vision, scheduled for June, we will explore the question: Is there a sustainable age by which all Canadians can retire without imposing a burden on the next generation?

1 Alexandre Pestov, "The Elusive Canadian Housing Bubble", Schulich School of Business, February 2010
2 John Krainer and Chrishen Wei of the Federal Reserve Bank of San Francisco, 2004
3 John Krainer and Chrishen Wei of the Federal Reserve Bank of San Francisco, 2004
4 5-year fixed rate mortgages, Canada
5 Source, Royal Bank of Canada, March 2012
6 Alexandre Pestov, "The Elusive Canadian Housing Bubble", Schulich School of Business, February 2010
7 This assumes annual increases to cover 50% of expected inflation.
8 Statistics Canada data, number employed as a percentage of female population ages 15-64
9 From the December 31, 2009 actuarial report on the Canada Pension Plan
10 From Statistics Canada Cansim table 282-0002
11 From the 2010 Actuarial Report on the Canada Pension Plan


For the complete article download: Vision - April 24, 2012 (PDF)