:
Great. Thank you very much for having us here this afternoon.
The deck that you have in front of you is organized into two broad sections. In the first few slides we focus on working-age women and men in the paid labour force and the proportion of them who belong to an employer-sponsored pension plan or a registered pension plan, in slightly different terms. We're going to consider data from three different sources, focusing on the trends in pension coverage drawn from each of those data sources to piece together the best picture we can.
In the second part of the presentation we're going to shift our focus from working-age women who may or may not have pensions to women aged 65 and over and the income that they receive. There the key story is that over the past 25 years we've seen quite a profound shift in the working lives of women, as evidenced in their increasing labour force participation and their contributions to retirement savings programs, and these changes are reflected in the amounts and sources of income they receive.
If you turn to the first line graph, data on pension coverage can be drawn from a number of sources, each of which have various strengths and limitations. We were going to consider three of those sources. This chart is based on administrative data compiled on all registered pension plans in Canada, and it shows the share of paid workers aged 17 to 64 who have pension coverage.
Among men, as shown in the blue line, there has been an ongoing decline in pension coverage rates in Canada over the last 15 to 20 years, from around 47% to 48% in the late 1980s to about 38% in 2007. Among women, there was an increase in pension coverage rates between the mid-eighties and the early to mid-nineties, and as you see here, these data suggest that the pension coverage rate for women has remained fairly stable, at around 39%, over the last decade.
Turning to two other sources of data on the next table, which are taxation data and household survey data, what I want to do is focus on the trends within age groups and focus particularly on the trends and coverage rate between 1997 and 2006-07.
Considering, first, men on the right-hand side of the table, data from the taxation date in the surveys confirmed the downward trend in pension coverage rates, particularly among men 35 to 44 and 45 to 54 years, and you see here a magnitude of decline in the range of four to six percentage points over the last decade. When we look more closely at men aged 25 to 34, we see that pension coverage rates have stabilized or perhaps increased very modestly.
When we look at taxation data for women on the left-hand side of the table, we see that the tax data show again stability in the pension coverage rates among women 35 to 44 and 45 to 54 years. Although the survey data show something of an increase for women, this could do with the way that respondents to household surveys report group RRSPs, for example, and that's the difference between these data sources. Interestingly, both the taxation data and the survey data show increases in pension coverage rates among women aged 25 to 34 over the past decade.
Turning to the next table, in addition to trends and coverage, changes in the characteristics of pension plans have received considerable attention and public discussion. Here what we're considering is the shift from defined benefit plans to defined contribution plans. In a nutshell, a defined benefit plan is one where retirement benefits are established by a formula that's laid out in the pension plan, such as 2% per year of service based on earnings during some period, and employer contributions to the plan are not predetermined but based on actuarial valuations. A defined contribution plan is one where the contributions are based on a fixed amount or a percentage of the employee's earnings, and while contribution amounts are known, the amount of the retirement benefit is only known when the employee reaches retirement and that depends on things such as the rate of return.
Between 1991 and 2007 the number of pension plan members who belong to a defined contribution plan more than doubled, increasing from about 466,000 to 935,000. Those numbers do not include group RRSPs. Evidence from the 2005 Workplace and Employee Survey indicates that about 18% of employees have a group RRSP, which is something akin to a defined contribution plan.
There are two other points I would draw from this table.
First, among both women and men, the proportion of paid workers who have pension coverage is far higher in the public sector, well above 80%, than it is in the private sector, at 22% for women and 29% for men.
Secondly, most of those individuals who have a pension plan—over 90% in the public sector—have a defined benefit plan, while those in the private sector who belong to a pension plan less often have one. Among women, 59% have a defined benefit plan, 26% a defined contribution plan, and increasingly we're seeing the emergence of mixed plans, wherein you have characteristics of both DC and DB plans or employees within the same firm who may belong to one or the other of those plans; we see that in the mixed numbers.
Turning to the next slide, I want to shift our focus from working-age women and their pension coverage to women who are now aged 65 and over, and to the retirement income they receive. Over the past three decades, median incomes of the elderly in Canada have increased significantly. Between 1980 and 2007, the median incomes of elderly couples increased from about $30,000 to $47,000, and those are inflation-adjusted dollars. Among elderly women living alone, the median income increased from just under $15,000 in 1980 to about $22,000 in 2007.
As displayed on the next slide, women's increasing participation in the paid labour force and their contributions to retirement savings programs, such as the Canada and Quebec pension plans and registered retirement savings plans and RPPs, are one factor underlying these trends. In this slide, we consider the role played by the Canada and Quebec pension plans. On this chart, the black line shows the percentage of women aged 65 or older who receive income from CPP or QPP as measured on the percentage axis on the left. Between 1980 and 2006, the proportion of senior women in Canada receiving CPP or QPP benefits more than doubled, from 35% to 84%, and of those who are receiving benefits, the median amount, as shown by the red line and the axis on the right, increased from $3,100 to $5,500.
We can also look, as shown on the next slide, at these same trends in terms of RRSP pensions and superannuations. Again we look at the proportion of women aged 65 and over receiving income from this source and at the median amounts. As shown on the black line, the proportion of women receiving retirement income from RRSPs, pensions, and superannuations increased from 20% to 55% over this period, and the median amount received by recipients increased from $4,600 to $7,400.
Overall, the income that Canadian women now receive in old age reflects changes in their working lives and their participation in both retirement savings programs, both publicly and privately administered.
As seen on the next slide, the overall effects of these changes can be assessed by considering the total aggregate income received by all senior women in Canada. In 2006, women aged 65 and over received about $54 billion, up from about $20 billion in 1980. That overall increase, as we move towards the right of the chart, reflects increases in average income as well as the fact that there are more elderly women in Canada today now in absolute terms than there were 26 years ago. But what's particularly interesting about this chart is the growing proportion of total aggregate income received from the Canada and Quebec pension plans, as shown by the yellow area, and the growing proportion of income received from RRSPs, pensions, and superannuations, as shown by the red area.
To conclude, I want to highlight two points regarding the composition of income received by seniors.
On the next table, we look at how the composition of income received by seniors varies by where they are in the income distribution. This slide is taken from a recent Statistics Canada study that looked at seniors who had significant labour force attachment when they were younger; specifically, they had earnings of at least $10,000 at age 55.
The key point to be made here is, if we look at the column entitled “bottom quintile at age 55,” that people who were at the bottom of the income distribution at age 55 received the majority of their income from old age security, guaranteed income supplement, and the Canada/Quebec pension plans—62% of it—by the time they were aged 75 to 77. For those who were at the middle of the income distribution when they were 55, the proportion of income received from OAS, GIS, and C/QPP was around 43%, while pensions, superannuations, and other investments accounted for almost half of their income at age 75.
Finally, those who were at the top of the income distribution relied primarily on pensions, superannuation, and investments. The importance of transfers on this chart would be more evident for those at the bottom of the distribution if this study hadn't looked at people with strong labour force attachment at age 55.
From the final slide in this afternoon's presentation, we want to make the point that in spite of the growing share of income that seniors and senior women receive from retirement savings programs such as the Canada and Quebec pension plans, pensions, and RRSPs, many also receive income from the guaranteed income supplement. In 1981, 55% of women aged 65 or older received the guaranteed income supplement. In 2008, 40% did so. The share of women aged 65 to 69 receiving the GIS is indicated by the dotted red line; it's about 10 percentage points lower. This suggests that GIS receipt is higher among women who are in their seventies and eighties.
Thank you very much.
Thank you for being here today.
Before getting started, I would like the committee to send its condolences to the family of the worker who lost his life on the Hill today in the boiler explosion yesterday.
I listened carefully to what you said. However, with regard to the figures in the second table, concerning "Tax filers with annual earnings greater than $1,000," I wonder to what extent that would be significantly different if we put: "Tax filers with annual earnings greater than $10,000." Do you have those figures as well?
Also, in May 2009, the Department of Finance published a news release in which it stated that the federal, provincial and territorial finance ministers had recommended amendments to the Canada Pension Plan starting in 2011. Among the new measures they would like to introduce, there is the main proposal, which is to maintain the contribution rate at 9.9%, but there is also some question of increasing the pension adjustment when a person retires before 65 years of age, from 0.5% per month to 0.6%. So it would be 7.2% per year instead of 6%.
The document also proposes to increase the pension adjustment starting after the 65th birthday, from 0.5% to 0.7% per month, that is from 6% to 8.4% per year, as also proposed in the discussion paper of the Régie des rentes du Québec.
Do you have any projections based those data? If these proposals are implemented, what will the population look like, with regard to women, pension plans and all that?
During this consultation—which is still underway, I imagine—certain women's groups have submitted briefs to ensure that their concerns are taken into account. One of those concerns—my colleague Ms. Neville referred to it—is women at home with children under seven years of age who cannot contribute to a pension plan while they are at home. These women's groups asked that they be granted pension credits equivalent to 60% of maximum eligible earnings.
Do you believe that measure would recognize the "invisible" work of those women and also enable older women to have better pension incomes?
:
In terms of answering that question, I would focus on three sets of characteristics that I think have implications. The first set of characteristics is what's happening with the labour market and demographic characteristics of women who are now in their thirties, forties and fifties, and coming up on retirement. What are the key trends we're seeing there? Again, I think, on the positive side, we're seeing high and sustained rates of labour force participation and, as a result of that—and acknowledging the fact of years out of the workforce—higher lifetime earnings than any preceding cohort.
The second factor that I would point to there is the tremendous increase in educational attainment among women in their thirties, forties and fifties. If we look at workers aged 45 to 59, even at the beginning of the 1990s—men included here—the proportion who had high school or even less than high school was quite substantial. I don't have the number off the top of my head, but it was in the range of probably 40% to 50%. The proportion with a university degree was quite small. That has been fundamentally reversed over the last 20 years. To the extent that things such as earnings and pension coverage are correlated with earnings, that would certainly have a positive impact.
On the negative side around the demographic and the labour markets, I think the incidence of later-life divorce is something that we hadn't seen before, and that would be another implication. So there are those types of trends in the demographics on the labour market side, plus the fact that they're having fewer children than used to be the case.
That's all fine and well, but the next set of characteristics is what's happening within the workplace. Here we have trends in pension coverage, the characteristics of pension plans. In the United States and the U.K., from what I read in the literature, the shift towards defined contribution plans has surpassed what we've seen in Canada. If we were to move in that direction of greater prevalence of group RRSPs, that would reshape the financial preparations that people would be making and the vehicles they would have to make those financial preparations, regardless of their education and these factors.
The third one I would point to is what's happening in financial markets in terms of rates of return on these things—I'm not in a position to speculate on that—and also in terms of whether people are in institutional savings arrangements like defined benefit plans or in individualized retirement accounts, which raises issues such as management expense ratios and things like that that people at the C.D. Howe Institute and others have pointed to.
So I think those three sets of characteristics—the demographic and labour market, what's happening within workplaces, and what's happening in the financial markets—all come to bear.
:
The fundamental issue you're getting at here is, what are the implications for retirement for individuals who do not have coverage within a registered pension plan? Somewhat surprisingly, the issue of income replacement rates for people who do and do not have pensions has not been addressed up until now, largely because of lack of data and the longitudinal data of the duration that we need to address that question.
But on that point, we could ask what the implications are for retirement of having or not having pension coverage. Last year we released the 2007 General Social Survey, and in that release we focused on non-retired Canadians who were 45 to 59 years and we asked them a whole battery of questions about what their plans and expectations were around retirement.
What we found was that pension coverage was associated with a number of plans and expectations. Individuals who did not have a pension plan were far more likely to say they did not intend to retire. There was a difference of about 12 percentage points between those two groups. Of those who said they planned on retiring, those who did not have a pension plan were far more likely to express uncertainty or to say they didn't know about the timing of their retirement.
Clearly, if you have a defined benefit plan and you know at age 60 you will get x% of your salary, it will add certainty to one's retirement plans. And similarly, among those who stated a planned age of retirement, the likelihood of saying they were going to retire before age 60 was 17% of those who did not have a pension, but 39% of those with a pension. So the age of retirement and the age at which people leave the labour force is fundamentally different for those with and without pensions.
Finally, we asked these near-retirees: at the time you expect to leave the workforce, do you think your retirement will be more than adequate, adequate, barely adequate, or not adequate to maintain your standard of living? What we found was that 74% of people with pensions said they expected their retirement income to be adequate or more than adequate, compared to 60% of those with no pension. So that was a difference of about 14 percentage points. Quite clearly, there are differences.
In another study we have currently under way, we're just looking at a very narrow cohort of individuals using the tax file data. These are people who did or did not have RPP coverage in 1991 when they were 55. And by the time those individuals were 70 to 72 years of age, the difference in the proportion who were retired and who had left the workforce was 5 to 15 percentage points, depending on where they were in the income distribution. So quite clearly there are implications.
The final point I would draw from that last study, though, was that of those individuals who had left the workforce--those with no RPP, they had retired--we didn't find a significant difference in the earnings replacement rate achieved by RPP members and non-members. RPP non-members were far less likely to be retired--as I've just said, there was a difference of 5 to 15 percentage points--but of those who had left, at age 70 to 72 they were replacing a comparable share of their earnings that they had had at 55 years.
That's only one study, and given the importance of that issue, it certainly warrants more and deeper analysis, and that's certainly an important question to keep pursuing.
:
The issue of income replacement rates is one that I would approach as a basket of indicators that one uses to assess the financial well-being of seniors. It is very possible to have a woman who has very modest attachment to the paid labour force through her working life, then at age 65 receives the guaranteed income supplement and old age security and meagre CPP benefits, and her earnings replacement rate is far above one, it might approach two, but her absolute level of income may be $15,000 or $18,000. So when we look at the very bottom of the income distribution, the concept of income replacement rates provides some information, but if your income was low during your working life and your income is comparable during your retirement, you will have a replacement rate of one. At that point, I think the additional indicators one ought to consider would be absolute levels of income, composition of income, and the extent to which one would rely on the guaranteed income supplement and other income support programs to maintain income in old age.
When we get further up the income distribution or, more accurately, the earnings distribution, the earnings replacement rate begins to make more sense in terms of a measure of the replacement of the standard of living. So at the middle of the distribution, if your earnings are $80,000 at age 40 or 50 and your income at age 70 is $50,000, you have a replacement rate of 0.67, and that in conjunction with some indicators around absolute levels and composition of income provides you with a picture. Similarly, you could have a hockey player with the Ottawa Senators who has a huge income at age 30 and huge lifetime earnings and has an income replacement rate of 0.3 at age 70, when in fact that tells us very little about the standard of living he is able to maintain.
In that sense, then, I would flag income replacement rates or earnings replacement rates as one of a series of measures that are useful to assess financial characteristics, but on it's own, I think, it ought to be interpreted with a degree of caution.
The other point I would make here is that while the concept of earnings replacement rates is frequently used and is an attractive benchmark for financial planners and policy-makers, there is no single agreed-upon definition of the concept or how to measure it. When we go back to the literature on this, you find people use very different things in the denominator. For pre-retirement income, is it income that is in the denominator in the calculation or is it earnings? If it's either of those, what do we count as pre-retirement? Is it the five years before receiving CPP? Is it average earnings between 45 and 64? Is it average annual earnings over a lifetime? What one puts in the denominator makes a very big difference.
Similarly, a numerator of the replacement rate, that is income in retirement, is not exactly straightforward either. For example, we have a significant proportion of people at age 70 who are receiving earnings. The very concept of an income or earnings replacement rate means that earnings are being replaced with something else. So if they are still working, do we count them in an earnings replacement rate calculation and do we count their earnings in that? Different people take different approaches to both of those issues.
One more that I would flag here is coming back to your issue of the CPI versus the average industrial wage. How do we adjust the dollar figures from the pre-retirement to the post-retirement period of time? Do we put them in constant dollars using the consumer price index, or do we put them in constant dollars in terms of average industrial wages or some other wage measure? That also can have very big implications for replacement rate outcomes.
To further complicate matters, sometimes in the calculation of replacement rates people take into account declining family size--the fact that children leave home and hence a dollar at age 70 may go further, or not--and family size at, say, age 40, and how does one take that into account?
Finally, I would throw into all that mix, how does one treat home equity in the replacement rate? If you live in a home at age 70 that you own outright, do you calculate some form of imputed rent? In a sense you're deriving an invisible income stream.
Thank you very much. We could keep you here for another two hours. This is very, very interesting.
You will get a note from the clerk, obviously, asking you for some data that everyone wanted, but Ms. Guarnieri wondered if you could provide us with any data to support the answer to her question, which was specifically about when a woman loses a spouse. She talked about her six-month interim, and you said that you noted it was later on and not in the immediate time after the loss of a spouse. Do you have any data to support that? She wanted to know. If you have, could you send it to us, please?
Secondly, I think Ms. Wong and Ms. Hoeppner wanted the work on the general survey in 2007, about how near-retirees may be lacking in financial information, and also perhaps the retirement plans and expectations of non-retired Canadians between 45 to 59, which you did, Mr. Schellenberg. Those are a couple of the things, but we will send you other information later on.
You did not give us any information on women in the unpaid workforce other than the fact that obviously they probably receive survivor benefits, OAS, and GIS. Is there a way to get any information on that, or is that really all that is available to them? If one wanted to look at ways to value unpaid work, would one have to come up with a brand-new policy? Is there no data on that?