Guaranteed Income Supplement (GIS) is not the same as Old Age Security Pension (OAS)
The first step in toning down the rhetoric (since confounding the two and adding the numbers together makes the problem look bigger than it really is) is to distinguish these two sets of federal government payments with completely different purposes and different eligibility requirements. In this discussion I'll stick to figures for a single person. Quoting figures for couples makes them look bigger, without helping to address whether abusive use of a TFSA will take place. It's simpler and more direct to deal with an individual.
GIS (per the official Service Canada website description) is the true low income benefit, the only payment that qualifies in any sense of the word "welfare" (which incidentally is not used officially by any Canadian government).
The maximum GIS payment for an individual in 2014 is $747.86 per month / $8974 per year. GIS starts being clawed back at the rate of 50 cents per dollar of income (other than OAS and GIS itself, which do not count as income for GIS-eligibility testing). GIS is completely clawed back and recaptured by the CRA when taxable income, which includes amounts withdrawn from an RRSP, from CPP, from dividends, interest and gains in taxable accounts plus any pension received, reaches a mere $16,728 in 2014 (TaxTips has details). GIS really is for retirees who little or no other income.
OAS (see Service Canada page) on the other hand is a payment merely for being an old citizen and having lived in Canada for a long time. Anyone 65 or over who has legally parked his/her butt in Canada for ten years or more, and is not in prison now can get it.
OAS currently pays $6765 per year to a single person. OAS is NOT for low income retirees, it's for everybody except really high income retirees. OAS only starts being clawed back through income taxes when a person's income reaches $71,592 and it is only all reclaimed by the government when income reaches about $116,700. That's hardly low income, not even middle income. Low income people do get OAS, but whether or not TFSAs are involved, OAS is clearly not intended for low income people only. So please, CBC, stop referring to OAS when using the term "welfare".
The problem only happens in a brief window, the three years between age 67 and 70. Both OAS and GIS are only for people 65 and over at the moment but in 2023 the age eligibility will begin rising gradually over the following six years to 67. TFSAs will need to have existed a lot longer than till 2029 to grow large enough to generate enough income to support a high-income lifestyle. At age 70, CPP must start (or you miss out, it's not paid retroactively). A rich person would likely have worked in a high-income job and likely therefore to get maximum CPP, which would be 42% more than at age 65, or $17,693, which is more than enough income (CPP counts as income when calculating the GIS clawback) to wipe out the GIS. At age 71, forced withdrawals from registered plans would easily (especially for rich people who would almost surely also have saved lots in such plans) take care of GIS.
TFSA accounts allow the tax-free accumulation of any type of investment earnings like dividends, interest and capital gains. The flat contribution limit is currently $5500 annually, starting at age 18, and carries over and accumulates through any years where no contribution is made. There is no tax refund like RRSPs give for contributions i.e. TFSA contributions are made with after-tax dollars. TFSA withdrawals are completely tax-free, not included in income on an annual tax return. There is no tracking or attribution of contributions vs income/profits in a withdrawal.
If the TFSA is thought to be the problem because it holds a large tax-free balance then it is not the problem. To be large enough to support a high income lifestyle for three years, wealthy people would have had to save / contribute the maximum $5500 every year. Over 35 years from age 32 to 67 that would total $192,500. For three years from 67 to 70, the wealthy steady saver could spend those contributions to support $64,000 in annual spending - tax-free, of course, because it's their own after-tax contributions. It's not income and no further income taxes are due and they could keep all their GIS. No tax-free income from the TFSA is needed for rich people to unfairly grab the GIS. Similarly, a rich person could well have large bank savings or a non-registered account, which could be converted for few years into low- or non-interest cash accounts to have plenty of tax-free assets generating no taxable income to spend. Blaming the TFSA as the threat to GIS is a red herring.
The tax-free earnings within a TFSA would help more middle income people implement a GIS-grab strategy since they would not need to contribute as much to the TFSA. Re-invested growth within the TFSA could be combined with the contribution capital to achieve a similar balance e.g. contributing $2500 annually for 35 years from age 32 (an age where more people start to be in a position to save) at 3.5% compound return produces a balance of $172,500 at age 67. As the FinAid calculator shows with those inputs, half the balance is the investor's own after-tax accumulated contributions and half is income.
Using unlikely assumptions to trump up the argument - It is only by using improbable figures that the CBC article's Kesselman puts together an argument that someone could live entirely off tax-free TFSA income. A 6% sustained return on a balanced portfolio that generates a $1 million portfolio? Not very likely! Historically, Stingy Investor's Asset Mixer tell us a real returns from 1980 to 2013 were 6.1% geometric (compound) for a 50-50 bonds-Canadian equity portfolio. Sounds right .... except those are index returns before fees. Select the Global Alpha Assumption of "Average Fund" for typical mutual fund fees that the Canadian investor faces and returns are down to 4.6%. As the famous Dalbar studies have repeatedly shown, investors on average do far worse than their funds by bad timing of buying and selling. Compound 4.6% for 35 years of $5500 annual contributions and the portfolio total is only $478,500, of which only $286,000 is accumulated income, the rest being the investor's own after-tax contributions. It is only by assuming that the rich investor would have been able from age 18 to contribute an unlikely $5500 per year for 49 years that a 4.6% return would produce a portfolio worth $1,008,000 by age 67, when they could achieve their goal in life of taking $9k of GIS by living off tax-free TFSA income for the next three years.
Is the historic average 6% return likely for the future? Um, no! - But let us suppose that rich investors have become so by being smarter than the average. They will buy and hold with mechanical (no gut-feel!) rebalancing low-fee index ETFs and so lose only 0.1% or so return per year off the index. What return can they reasonably expect?
The future return environment is nowhere near as rosy as the halcyon decades of the 1980s and 1990s when stocks were going gangbusters and bonds benefited hugely from continually declining interest rates. Today's reality - Stocks in future might return 3 to 5% while bonds look set for 2% returns. A portfolio combined 3 to 4% is more reasonable. We'll use 3.5%. That has a big effect. TFSA contribution room accumulates and never expires, but most of the growth in a portfolio comes from long term compounding. Starting an investor at age 30, giving 37 years of savings and growth at 3.5% return, gives a TFSA portfolio of only $418,000. At that stage taking out a 3.5% return as tax-free income would provide only $14,600, hardly enough to sustain a wealthy lifestyle. The investor would need to withdraw some $45,400 of the accumulated capital and past income (each of which constitutes about half the portfolio total) to reach Kesselman's $60k of tax-free spending.
In short, a few investors may be lucky or skilled enough to accumulate huge TFSA balances that will enable them to live entirely off tax-free income while collecting GIS for a few years but they aren't likely to be numerous. Is it worth turning the tax system upside down, such as testing for assets instead of income, or drastically changing the TFSA by making some of the income taxable?
The TFSA exemption for GIS (and OAS) eligibility is not a flaw or a "loophole", it is fundamental design feature and deliberate rule to benefit large numbers of modest income retirees.
The government itself explicitly touts the feature as an advantage. The TFSA was designed to especially help, and various analyses comparing the TFSA to the RRSP (including my own), show that it is most attractive to people with low income. So CBC should stop pejoratively describing this as a loophole and people like John Stapleton quoted in the Financial Post should stop calling it a "policy flaw".
Are the rich likely to try grabbing GIS, even though they could do it? Too many other factors mean likely few will.
Future very well-off retired people will have lots of assets. They probably will have maxed out their TFSA and their RRSP, be entitled to maximum CPP and maybe even have non-registered assets too. Assuming that rich people want to stay rich and have no moral qualms (which might deter a few) about possibly claiming GIS, it will be one of the factors in the mix. Staying rich means considering more than the short period from 67 to 70, it means taking account of the whole of remaining lifetime. It may not be worthwhile taking GIS if that means pushing up into higher tax brackets later and paying more tax in total with less disposable after-tax income. Taking GIS is not a single dominant no-brainer must-do. Sophisticated retirement planning tools like RRIFmetic that figure out the best withdrawal strategies depend on many assumptions and factors.
Is there likely to be a problem with public finances? Nope, according to the Chief Actuary at the Office of the Superintendent of Financial Institutions in this latest 2014 report on the Old Age Security Program. Note that the conclusions quoted below are about the entire expected effect of the TFSA, including especially the mass of lower income Canadians using TFSA tax-free income to get GIS as intended, and not just rich people abusively using the GIS-gambit.
- "The GIS recipient rate is projected to slowly increase from its current level of 32% to 34% by 2030 due to the impact of TFSAs. The GIS recipient rate is subsequently projected to reduce to 31% by 2050." i.e. in the major impact in 2030 is when TFSA have been in existence only 21 years and have not yet grown to large amounts, then it tails off ... is this an ominous problem?
- "... the fact that individuals are also assumed to invest in TFSAs results in GIS and Allowance recipient rates increasing slightly over time. Ultimately, however, the fact that benefits are indexed to inflation as opposed to wages drive the cost of the OAS Program relative to the GDP down over the long term, with the result that annual expenditures are expected to fall to 2.4% of GDP by 2050." Doesn't sound like the sky is forecast to fall does it?
- "... he thinks this [suppressing all taxable income to claim GIS] would rarely be possible. Most high-income individuals have other income sources and would inevitably render them ineligible for GIS. And since TFSAs have only been around five years and so remain mostly small, “this might be too hypothetical to comment on, given that the purported optimal scenario is decades away,” he said."