Robert Frank argues that the US should replace its income tax with a progressive consumption tax:
By replacing federal income taxes with a steeply progressive consumption tax, the United States could erase the federal deficit, stimulate additional savings, pay for valuable public services and reduce overseas borrowing — all without requiring difficult sacrifices from taxpayers.
Translating the rest of Frank's argument into Canadian, this is how it would work:
Under such a tax, people would report not only their income but also their annual savings. A person would be taxed on income minus savings - just like people are taxed now on income minus RRSP contributions. That amount would be "taxable consumption." Any dissaving - RRSP withdrawals, for example - would be included in taxable consumption and taxed. There would be a personal amount tax credit, like there is now. Rates would start low, as they do now. As taxable consumption rises, the tax rate on additional consumption would also rise.
So basically a progressive consumption tax is just like allowing everybody to make unlimited RRSP contributions.
Frank argues that such a tax would have economic benefits:
With a progressive income tax, marginal tax rates cannot rise beyond a certain threshold without threatening incentives to save and invest. Under a progressive consumption tax, however, higher marginal tax rates actually strengthen those incentives.
Unfortunately, there is a problem with Frank's argument, that can be represented diagrammatically by representing the tax design problem as a game:
Yes, exempting savings from tax would reduce inefficiencies associated with the taxation of savings. That whacks down one mole. But in order to maintain tax revenues, some other tax would have to be raised.
In fact, there aren't a lot of choices.
Allowing people to make unlimited RRSP contributions, and taxing RRSP withdrawals, like Frank suggests, is equivalent to having a zero tax rate on investment income. Here's an example. I used simple interest, so the total return on investments is 60% or 0.6 over a ten year time period (can't do compound interest in my head while standing in front of 70 people).
The after tax amounts in the "no tax on investment income" and the "RRSP" column are exactly the same, unless a person faces a lower or higher marginal tax rate when withdrawing RRSP funds (remember that the principal invested in an RRSP is subject to tax when it is withdrawn).
So a progressive consumption tax is like not taxing investment income.
It follows that a progressive consumption tax is equivalent to only taxing wage income.
[I'd recommend stopping now, getting yourself your beverage of choice, and just mulling that over for a while. But if you really care about such things, here's the math. Let w1=wages in period 1, w2=wages in period 2. Consumption in period 1 = wages in period 1 - savings, so c1=w1-s1. Consumption in period 2 = wages in period 2 plus the return to savings, i.e. c2=w2+(1+r)s1. The present discounted value of lifetime consumption = c1+(1/1+r)c2= (substituting for c1 and c2) w1-s1+ (1/1+r)(w2+(1+r)s1) =w1+(1/1+r)w2. Taxing lifetime consumption is the same as taxing lifetime wages]
[Correction: as a number of people have pointed out in the comments below, this proof ignores the possibility that people might inherit money or win the lottery. Also, switching from our current system to a consumption tax half way through people's lives isn't the same as having a consumption tax all along.]
Getting back to the whac-a-mole problem: if taxes on investment income are lowered (that's what a progressive consumption tax does), and the government wishes to maintain the same level of tax revenue, taxes on wage income must be raised.
The government could raise taxes on the income earned by the highest wage earners. There might be some margin for tax increases here, but not a lot: tax rates on this group are already fairly high (40%+). Taxes on wages create inefficiencies by distorting labour supply decisions, and the higher the tax rate, the bigger the distortions - we don't want people to stop saving, but we don't want doctors to stop working, either. Moreover, a large disrepancy between tax rates on wage income and tax rates on investment could give rise to tax avoidance opportunities: making wage income look like investment returns.
And it must be stressed: taxation of investment income is only an issue for upper-middle and high income earners. The RRSP contribution limit is 18% of earned income up to a maximum of $22,000 per year. Tax Free Savings Accounts (TFSAs) provide another $5,000 per year of tax sheltered savings. Registered Education Savings Plans are, for parents and grandparents, yet another tax-preferred method of savings. Moreover, any returns on investment in your own home are also tax free.
Yes, there are people - those with no mortgages and generous employer-provided pensions -- who are running up against the RRSP contribution limit. Perhaps a number of people reading this blog are in that position. But the majority of Canadians do not come close to exhausting their tax free investment opportunities.
Jonathan Kesselman has advocated raising the RRSP contribution limit above the current $22,000 level. The closer I get to exhausting my own RRSP room, the better this idea becomes.
Another change I would advocate: allowing people more flexibility in terms of withdrawing and re-contributing RRSP funds, so RRSPs can be used for income averaging. Sure, continue to tax withdrawals. But then add withdrawals to an individual's contribution room, so any withdrawal could be made up again later on.
So perhaps some tinkering with our current system is in order. But a move to unlimited RRSP contributions, like Frank proposes? Personally, I wouldn't favour it.
The somewhat ironic thing is that Robert Frank - who is a brilliant behavioural economist - doesn't realize that people's psychological quirks are a major problem with RRSP-type savings programs.
Taxes that people don't see - the European VAT, for example, or income taxes deducted at source, things that come out of your pay cheque before it ever arrives - hurt, but they don't hurt that much. It's like being caught unawares when you're given a shot - painful, but bearable.
But taxes that people see really hurt. You contribute money to your RRSP. You watch it grow. You care for your RRSP, tend it lovingly, daydream about all of the things you're going to do with those funds ("every summer we can rent a cottage on the Isle of Wight..."). Then it comes time to take money out of your RRSP - and it's TAXED! Taken from you!
Skeptical? Look at the enormous political pressures exerted on recent governments to minimize taxes on retirees through income splitting. Look at the recent debate about whether or not special withdrawals from RRIFS (what RRSPs become upon retirement) should be counted as income for the purposes of income support programs.
It's a looming political battle: will the government have the guts to stick with the current system? Or will the political clout of a booming generation of retirees carry the day?
This post is in response to a request from a WCI reader.