Kevin Milligan is Associate Professor of Economics at the University of British Columbia
According to the Department of Finance Tax Expenditure and Evaluation Guide, savings held in Tax Free Savings Accounts lowered Canadians' taxes by about $45-million in 2009, and $155-million in 2010. The Conservative Party's plan to double contribution room from $5,000 a person per year to $10,000 is estimated by them to cost only $30-million.
I don't have problems with these estimates per se -- the cost of TFSAs right now is likely small. Moreover, the expanded $10,000 contribution room per year may not have large take-up compared to $5,000 because many families might find it hard to save more than $5,000 per year anyway.
The more interesting question, to my mind, are the long-run implications of TFSAs for our tax system. Right now, three years into the TFSA system, each Canadian has $15,000 of room available. Seven years from now, we will reach TFSA contribution room of $50,000 per person. If the annual limit is increased to $10,000 per year, reaching $50,000 would happen even more quickly. At that point, each Canadian couple would have $100,000 of TFSA room available.
Taking data from the 2005 Survey of Financial Security, I asked the question "how many Canadian families have $100,000 or more of taxable financial assets"? The answer is only 9 per cent of families. This means that the TFSA could, in the medium term, wipe out capital income taxation for about 91 per cent of Canadian families, if they shift their existing taxable assets into TFSAs.
There are positives and negatives with shifting away from capital income taxation, but we should be clear on the long-run implications of TFSAs -- they will transform the nature of our personal income taxation system. This transformation begins slowly, but within a few years will be more evident as families have more and more room available to shelter their taxable assets.
What will this cost in terms of foregone tax revenue? If you assume that people now paying tax on their assets will shift those taxable assets into TFSAs, then we can calculate an estimate of the cost. TFSAs are much more flexible than RRSPs, so this assumption is more tenable for TFSAs than it might have been for RRSPs.
In 2005 according to my calculations with the Survey of Household Spending, a $100,000 per family TFSA would have shielded about 46 per cent of taxable assets from taxation, assuming taxable assets were not left outside the TFSA when TFSA room was available. From the CRA data available, I calculate total taxable capital gains, dividends, and interest investment income to be $57.7-billion in 2005. If we apply that 46 per cent sheltering rate to this $57.7-billion, we get $26.5-billion of sheltered income. If we assume the average federal tax rate on capital income is 25 per cent (most capital income is taxed in the higher 22 per cent, 26 per cent and 29 per cent tax brackets), this yields a revenue cost of $6.6-billion, or 7 per cent of federal income tax revenues.
This number is substantially different than the 2010 Tax Expenditure Guide estimate of $155-million. There are two reasons for this. First, as of 2010 only $10,000 per person of TFSA room was available -- we are still in the early phase of the TFSA transformation.
My estimate does not apply to 2010 or 2011. Instead, my estimate is for the time when there is $100,000 of TFSA room per family, not now. I think that is the right question for those interested in the long-run costs of our current tax choices. The second reason is that families may not put their existing taxable assets into a TFSA. While low take-up may have occurred when the total contribution limit was low, my assumption is that few high-savings families will leave substantial taxable assets outside the flexible TFSA when their available TFSA room grows into the 6 digits. I don't have to assume high take-up of TFSAs for everyone in the population to get high revenue costs. I just have to assume that high-savings families with access to good financial advice don't willingly subject the bulk of their savings to high rates of taxation when the TFSA room becomes available. I think that is a defensible assumption.
TFSAs have a lot going for them. They allow lower and middle income families to shield their retirement savings from high rates of taxation and clawbacks of public pensions, leveling the tax 'playing field' compared to high income families with access to many tax-planning strategies. TFSAs also push toward a consumption basis for taxation, which many economists feel is a better base for economic efficiency and for assessing wellbeing. However, if the TFSA system is allowed to mature with hundreds of thousands of dollars of contribution room per person, it will have a substantial cost to income tax revenues. I hope there will be a debate about how TFSA contribution room should grow in the future.
Follow Economy Lab on twitter