On Friday, December 18, 2009 our federal-provincial-territorial Finance ministers and treasurers met in Whitehorse, Yukon to assess retirement income adequacy in Canada. On the agenda was a report prepared by Dr. Jack Mintz, Research Director. The purpose of the report is to expand Canadians’ knowledge of retirement income adequacy and explore related issues. The report concludes that “overall, the Canadian retirement income system is performing well, providing Canadians with an adequate standard of living upon retirement.” The conclusion is something of a puzzle, as it doesn’t reflect how Canadians feel about the situation. The gap between Dr. Mintz’s and the publics’ perceptions is very wide. The question then is; why does the gap exist?
Dr. Mintz uses a very broad definition of the retirement system. He includes pensions, both private and public (C/QPP), transfer payments (OAS, GIS and provincial supplements), RRSPs, Tax-Free Savings Accounts, any other savings that can provide support in retirement, home ownership and all financial assets. Based on this he concludes that the disposable income of those aged 65 or older is about 90% of the average disposable income of all Canadians. On this basis, on average, we do very well indeed.
After that introduction, the report starts with an assessment as to whether Canadian saving has declined. Sufficient saving is important so that enough accumulated wealth will provide a reasonable income on retirement. Saving is typically thought of as the difference between annual income and the consumption of goods and services. On this basis Canadians have been saving between about 3% and 5% of personal disposable income over the last decade. Saving rates were much higher in the early-1980s and we are now back to what they were in 1961, i.e. 5%. This would seem to indicate that saving has declined, but Dr. Mintz suggests that a correction should be made and that purchases of consumer durables should be added into the savings rate. Consumer durables are mass-market heavy goods that are expected to last for some time. These include washing machines, refrigerators, furniture, cars, TVs, etc. When this is done the savings rate jumps up to about 15% and has been reasonably stable over the last 40 years – except in the early-1980s, when people decided to save real money instead of consumer durables. A saving rate of 15% would certainly be enough to fund an adequate retirement income, although I think I would rather have the saving in something other than consumer durables that wear out over time.
The report leaves savings behind and turns to retirement income adequacy. Dr. Mintz discards income replacement rates as a measure of adequacy as needs vary considerably depending on individual circumstances. He expects that people will consume less than available income during the years they work in order to fund consumption after retirement. He then recommends a measure called “consumption smoothing” whereby a person maintains a similar standard of living when they retire. The smoothing takes into account such things as retirees no longer needing to support their children or parents (?), having more time to do household duties (?, if able), having bought a home, car and other consumer durables prior to retirement (?, nothing wears out), and being able to take out a reverse mortgage. He then returns to income replacement ratios and says that 60% of pre-tax income should be adequate to maintain expenditures. This is reduced to 50% for those earning double the average. The reference to “pre-tax” may be a typo since he refers to 60% “after-tax” in his next paragraph.
He refers to a longitudinal study – looking at individuals over a period of time – that indicates that few have inadequate resources at the lowest income levels, 20 to 25% have inadequate resources at the median income level and 35% have inadequate resources at the top quintile level. The study concludes, “a significant minority of Canadians may not have sufficient replacement income.” The key is to have a high income replacement ratio at least until people enter their 70s. Dr. Mintz acknowledges that it would be important to understand what factors play a role in explaining the income replacement ratios, but seems to discard the conclusions since the study did not take into account the role of consumer durables and owner-occupied housing.
Dr. Mintz presents several tables that he feels demonstrates that once the value of owner-occupied housing is taken into account, people achieve retirement income adequacy. The tables are based on either snapshots or hypothetical models, not longitudinal studies. For example, for one table the assumption is that each household buys a home with a value of 3 times earnings while working. Clearly something is missing in this description as no one is going to spend 3 times their total earning over their career. My best guess is that he means 3 times the highest annual earning, which in the model would be earned in the year prior to retirement. This value is then amortized back into income over the 20-year period that the retiree is expected to live. A test is then made as to where the household experienced either a 100% or 90% consumption replacement. There are several problems with this approach:
· The 20-year amortization period is roughly the period from age 65 to life expectancy for an individual. But, 50% of individuals live longer than the life expectancy and that percentage is much higher for a couple. Given that people cannot predict their own life expectancy and would surely not want to run out of resources before they die, a much longer period should have been used.
· Home ownership involves many more costs than the purchase price of the home – maintenance, repair, heat, light, replacement consumer durables, etc. For many, these costs today exceed the yearly costs of purchasing the home and with inflation they will continue to rise in the future. The costs will also rise as the ability to self-maintain diminishes with age and more reliance is placed on repair and maintenance people.
· A home cannot be sold in bits and pieces to match the expenses of the retiree. The retiree needs real income in order to have some sort of standard of living and to be able to purchase the consumer durables – frig, stove, car, TV, etc. - that will wear out over the period of retirement.
· Many people have paid for their house well before they retire. This changes the results of any comparison that is made to pre-retirement income.
· The approach bears no relationship to reality. It’s interesting that in one of the tables a two parent family has a higher consumption replacement after retirement than a couple with no children. Presumably having children gets you used to spending less and saving more.
The effect of adding a home ownership component to assess retirement adequacy is very significant. In 2005 retired Canadians were found to have an average net worth of $485,000. Of this $174,000 per household is pension and tax sheltered savings and $152,000 is principal residence. However, when one adjusts for taxes on the pension and savings – not on the residence – the residence is actually the most important category of worth. These numbers are averages and are therefore influenced by the wealthy. If a median number is used, net worth drops to $300,000. It’s hard to say what this change does to the pension and residence categories. The biggest impact may be on the pension category as later in the report Dr. Mintz indicates that for those without RPPs (pensions) there is some reliance on the GIS even in the third and fourth quintiles.
The balance of the report deals with issues related to retirement income adequacy such as investment performance and risk, the costs of various funds, why passive management is better than active management and overall efficiencies. There seems to be little correlation between costs and size of funds – contrary to the super fund concept that is being promoted by many. This part of the paper is quite good and well worth the read. It clearly points out how the public is likely paying for services it doesn’t need and that these costs affect retirement income adequacy.
Dr. Mintz raises the question as to whether a new savings program would encourage more savings. An example might be expanding the mandatory savings programs. His feeling is that the introduction of a government public pension fund could result in public pension funds being substituted for current private pension plans. Based on his comments earlier about costs, both investment and administrative, and how low they are in the government plans, substitution would clearly be attractive. But, given the paltry amounts that those in the lower to middle income brackets are saving today, I’m not sure substitution would be much of a problem.
So, the conclusion is that “overall, the Canadian retirement income system is performing well, providing Canadians with an adequate standard of living upon retirement.” This is only true if the value of the owner-occupied home is included as an asset. This explains most of the gap in understanding between Dr. Mintz’s report and the views of Canadians. As a policy paper, this one is not of much help. There are too many guesses and assumptions. What is needed is a comprehensive longitudinal study, one that traces what actually happens to retirees’ incomes.