Month: March 2003

Spring 2003 issue of the Expert Witness newsletter (volume 8, issue 1)

Posted on By economica

Contents:

  • Selecting the Discount Rate – An Update
    • by Christopher Bruce, Derek Aldridge, Kelly Rathje, and Scott Beesley
    • This article extends the work done by us in issues 5(3) and 6(4) of The Expert Witness, we conclude that it would be appropriate to revise our existing 2½ and 3½ percent two-part forecast of real interest rates. We propose to use a rate of 2¼ percent for the first five years of all calculations. For all subsequent years we propose to use a rate of 3¼ percent.
  • Retirement trends in Canada
    • by Kelly Rathje
    • This article compares retirement ages of Canadians over the five year period 1991-95 to retirement ages of Canadians over the five year period 1996-2000. The findings show that for most educational and industry categories, Canadians are retiring earlier than they did even five years ago.

Retirement trends in Canada

Posted on By economica

by Kelly Rathje

This article first appeared in the spring 2003 issue of the Expert Witness.

The value of an individual’s potential earning capacity depends in part on how long she will be in the labour force. That is, for the calculation of lost earnings, it is necessary to make assumptions concerning the age at which the individual would have retired (and will now retire).

The focus of this article is to compare retirement ages of Canadians over the five year period 1991-95 to retirement ages of Canadians over the five year period 1996-2000.* The findings show that for most educational and industry categories, Canadians are retiring earlier than they did even five years ago.

In the early 1980s, the median retirement age was close to age 65. However, retirement ages decreased steadily from 1986 to approximately 1993. Notably, in 1987, the Government of Canada reduced the age (from age 65 to age 60) at which Canada Pension Plan benefits could be collected (albeit with a reduced pension amount). I compare the retirement patterns of Canadians over the 1991-95 period to the retirement patterns of Canadians over the 1996-2000 year period to see if the trend for earlier retirement has continued. In Table 1, the distribution of retirement ages in Canada over the two five-year periods (1991-95 and 1996-2000) is summarized.

As shown in Table 1, in the five year period 1991-95, the highest percentage of individuals retired in the 60 to 64 year age category. In the five year period 1996-2000, 60 to 64 still remained the most popular age group for retirement, however the percentage of individuals retiring within the age category had dropped by 6 percentage points (37 percent to 31 percent). The percentage of individuals retiring “earlier”, in the 50 to 54 age category and 55 to 59 age category both increased – the former having the largest increase (9 percent to 15 percent, or 6 percentage points). This suggests that although many Canadians are still choosing to retire at a “normal” retirement age (60 to 64), there is a shift to earlier ages.

Table 1

Although the percentage of people retiring in their 60s decreased within the time period considered, the percentage of individuals retiring at age 70 or above remained unchanged.

Table 2 compares median retirement ages in Canada over the periods 1991-95 and 1996-2000. (The “median” retirement age is defined such that 50 percent of individuals retire at ages above that age and 50 percent below it.)

As shown in Table 2, the overall median retirement age has decreased by one year, from approximately age 62 (1991-95) to 61 (1996-2000). The median retirement ages of private employees also decreased by one year (from age 63 to age 62), however the retirement ages of public employees decreased by two years (from 60 to 58). Self-employed individuals experienced no change in retirement trends from 1991-95 to 1996-2000. The earlier retirement ages of public employees are likely due to the generous pension plans available to most public employees, which often offer incentives to retire at earlier ages.

Table 2

Table 3 summarises retirement ages by industry. As shown in the table, many industry categories also experienced decreases in median retirement ages between 1991-95 and 1996-2000. However, self-employed individuals did not alter their retirement patterns in most industries and, if anything, the retirement age of self-employed individuals may be increasing.

Table 3

Considering the “employees” category for the goods-producing occupations, retirement ages remained constant for all industries except for utilities and construction. These two experienced a decline in retirement ages of 2 years. For service-producing industries, all declined except management, which remained constant at 65. Educational services showed the largest decline in retirement ages over the two five year periods – from age 61 to age 57, a drop of four years. This may have resulted from restructuring within the education sector that led to the offering of early retirement packages to many teachers. If so, there may be a reversal of this decline in the future (as fewer early retirement incentives are offered).

Self-employed individuals again show higher retirement ages than employees. Within the goods-producing industries, most show an increase in the age of retirement (there was a decline in manufacturing), which goes against the overall trend for retirement patterns of Canadians. The service-producing industries, however, showed a one year decrease in retirement ages for the trade, management, and other industries; constant retirement ages for transportation and professional industries; and a one year increase in retirement ages for health care and accommodation industries. The largest increase was in the finance, insurance, and real estate industries, which saw a four-year increase in retirement ages over 1991-95.

The above tables show retirement trends by industry, and the statistics combine both male and female statistics, and also do not consider specific levels of education. The statistics for males and females, by education levels are summarized in Table 4.

Table 4

Over 1991-95, men and women overall had similar retirement patterns. That is, there was only one year difference in the retirement ages at each education level. In 1996-2000, the male retirement patterns saw very little change from 1991-95, with most education levels having unchanged retirement ages. Only males at the high school diploma level experienced a decline in retirement (age 61 to 60).

Females, on the other hand, have followed a decline at each level by at least one year. Females with a university education resulted in the lowest median retirement age (57) over 1996-2000. Also note that at the high school diploma level, males and females experienced the same retirement age (60).

I note, however, that the female retirement trends may be underestimated. Retirement trends are based on historical retirement, and may not reflect the trends of future generations. That is, young women in the labour force now may experience different retirement patterns than women who were of retirement age in the last five years. The current generation of women are obtaining higher levels of education, and are participating in the labour force more, as compared to those women of retirement age now. This suggests that their labour force attachment may be greater than the attachment of the older cohorts. Therefore, it may be theorized that young women now in the labour force will retire later than women who faced the retirement decision in recent years.

Also, there is evidence to suggest that in the future, there may be pressure for Canadians, both male and female, to delay retirement. For example, in a paper entitled “Future Age of Retirement”,** Brown argued that as the baby boom generation moves into the retirement ages, they will attempt to liquidate assets in order to buy goods and services. This will reduce the value of the assets due to the number of retirees attempting to do this. Also, the smaller “baby bust” generation will be the source of labour within the economy. The production in the economy may slow due to the smaller labour force, resulting in price inflation. This may force some potential retirees to postpone retirement since the value of their assets will have decreased. Thus, Brown’s prediction for retirement in Canada is that the median retirement age will fluctuate between age 60 and 61 over the next 47 years. That is, the overall median age of retirement and the trend of decreasing retirement ages may not continue on into the future.

Conclusion

Based on the above information, it seems that the median retirement age in Canada has fallen from age 62 to 61 and, depending on educational attainment and gender, the average retirement of an individual may be as low as 57, or as high as 65. Self-employed individuals continued to retire at approximately age 65.

For the purposes of loss of income calculations, it seems reasonable to consider the education level of the individual, since there are statistics available for both males and females. In addition, if a career path has been established, one should also consider retirement patterns of the specific industry. For minors, the overall or educational statistics would be appropriate.

Footnotes

* The source of the data provided in this article is the Statistics Canada publication Perspectives on Labour and Income, Summer 2002 Vol. 14, No. 2, and Summer 1997 Vol. 9, No. 2. [back to text of article]

** Brown, Robert L. “Future Age of Retirement” Canadian Investment Review, Fall 2002, pages 32-37. [back to text of article]

leaf

Kelly Rathje is a consultant with Economica and has a Master of Arts degree (in economics) from the University of Calgary.

Selecting the Discount Rate – An Update

Posted on By economica

by Christopher Bruce, Derek Aldridge, Kelly Rathje, and Scott Beesley

This article first appeared in the spring 2003 issue of the Expert Witness.

In the Autumn 2000 issue of this newsletter, we conducted an extensive review of the various methods of measuring the real rate of interest, or discount rate, and presented evidence concerning the movement of those measures over the period 1995-2000.

That survey was subsequently updated in our Winter 2001/2002 issue (Vol. 6, No. 4). What we found was that interest rates had begun to fall, relative to the historically high levels that had persisted over most of the 1990s.

At that time, we concluded that the best estimate of the long-run discount rate was 3½ percent. But we also argued that, as interest rates on short-term bonds and GICs were lower than those on longer term investments, it would be appropriate to employ an interest rate of 2½ percent on the first five years of any investment.

The purpose of this article will be to provide five additional quarters (15 months) of data to determine whether the trend we observed at the beginning of 2002 has continued, or whether a revision in our recommended interest rate is appropriate.

Revised data

Tables 1 and 2 provide updates of the information contained in the equivalent tables of the Winter 2001/2002 article. In particular, we have added data for all four quarters of 2002 plus the first quarter of 2003.

Table 1 reports the “raw” data from which some of the real interest rate figures in Table 2 have been calculated. The first column reports the “core rate of inflation” – a measure of the rate of inflation that removes the effects of change in those components of the price index that often move erratically, such as food, energy, and taxes. It is often argued that this measure offers a more reliable predictor of future changes in prices than does the “standard” measure of price inflation. (See the Autumn 2000 article for a detailed description of the core rate of inflation.)

The next three columns in Table 1 report the rates of return on Government of Canada 5-year and 10-year bonds and on 5-year Guaranteed Investment Certificates (GICs). The former represent the minimum rates of return that investors can expect on safe investments. The rate of return on GICs, on the other hand, represents the interest rate available on a mixed, low-risk portfolio of stocks and bonds.

Table 1

Table 2 reports seven measures of the real rate of interest – that is, the rate of interest net of the expected rate of inflation. The first of these is the market-determined rate of return on “real rate of return bonds” – bonds whose value is denominated in terms of the real rate of interest. These bonds are of particular importance because they are purchased by sophisticated investors and because they tend to held for long periods of time.

The second, fourth, and sixth columns report the 5- and 10-year government bond interest rates and 5-year GIC rates net of the core inflation measure.

Finally, columns three, five, and seven report the government bond and GIC rates net of the Bank of Canada’s target rate of inflation of 2 percent. As the Bank has managed to keep the core rate of inflation within a small band around this target for the last eight years, it is widely believed that 2 percent is the rate that is expected by most investors. That is, investors are believed to act as if the real rate of interest is the observed, nominal rate less 2 percent.

Table 2

Interpretation of the data

The data in Table 2 indicate that real rates of interest have continued the downward trend that began in 1996/1997. Whereas we concluded a year ago that long-term interest rates were approximately 3½ percent and short-term rates approximately 2½ percent; it appears that those rates have now fallen to 3 percent and 2¼ percent, respectively.

Note that the latter rate is close to the rates reported in the Bank of Canada’s Monetary Policy Report of April 2003 (Chart 19, p. 24).

In addition, 3 percent is the rate at which the Bank of Canada recently issued a new set of 33-year real rate of interest bonds. As we argued in the Autumn 2000 issue of the Expert Witness, the rate of return on these bonds is a particularly reliable estimate of the expected real interest rate as they are purchased primarily by large institutional investors (like pension funds) that have made considerable investments in the prediction of future rates of interest and inflation.

For this reason, we believe that it would be appropriate to revise our existing 2½ and 3½ percent two-part forecast of real interest rates. Based primarily on the observed rate on 5-year Government of Canada bonds, we propose to use a rate of 2¼ percent for the first five years of all calculations. For all subsequent years we propose to use a rate of 3¼ percent – though we note that a rate as low as 3 percent could be supported based on the most recent observed rates on 10-year Government of Canada bonds and based on the Bank of Canada’s current issue of real rate of return bonds. Our long-term rate is perhaps slightly conservative, but we will re-examine this issue next year and decide then if changes are warranted.

leaf

Christopher Bruce is the President of Economica and a Professor of Economics at the University of Calgary. He is also the author of Assessment of Personal Injury Damages (Butterworths, 2004).

Derek Aldridge is a consultant with Economica and has a Master of Arts degree (in economics) from the University of Victoria.

Scott Beesley is a consultant with Economica and has a Master of Arts degree (in economics) from the University of British Columbia.

Kelly Rathje is a consultant with Economica and has a Master of Arts degree (in economics) from the University of Calgary.