« Videogames and the sexual division of labour | Main | Excess demands for apples, bonds, and money in a representative agent economy »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Great post - quite enjoyed it.

RE: "But at the same time, there are repeated calls for lower corporate taxes, and raising consumption taxes is political suicide."

How about a carbon tax? Legalizing and taxing marijuana? Lots of other options here.

"And predictably, there will be calls for lower rates and greater exemptions.

Will governments be able to resist? Retirees are a potent political force: they vote, they are organized."

I would be surprised if this ended up working out. In the end, governments also need to finance operations. A very large population of senior citizens with a huge amount of savings is a major source of revenue. Shifting the tax burden on to the middle class between 18 and 65 is likely to undermine the ability to fund programs like the Canadian Pension Plan (or social security if you are state-side).

It is hard to treat a group with massive tax perks when they are a major part of the tax base. It doesn't mean that it won't work but it does mean that this will get more expensive (in terms of forgone revenue) over time. In a real sense, these savings are a claim on the output of workers by retirees. It's a very nice social contract but it depends on the groups involved being moderate in their demands.

On the other hand, I am less clear why a Roth Style savings vehicle isn't a good solution, overall. The taxes were paid upfront, Canada has a value added tax (it was PST plus GST when I lived there, I think it is HST now) so you do recover more money when it is spent. And the psychology seems to work better.

In fairness, there is a reasonable argument that the current RRIF rules don't properly reflect modern demographic (longer life expectancy) and financial realities (lower interest rates). So, the proposals from CD Howe don't realy bother me that much And so long as the proposed changes only affect the timing of the income inclusion (i.e., how much of you RRIF you have to cash-out in a given year), it doesn't affect the long-term financial health of the fisc. Personally, I've always wondered why we need the RRIF rules at all. If people prefer to sit on their assets to defer taxes for another 15-20 years, well, so much the better, it'll be that much larger a tax hit when they die.

In any event, I was underwelmed by the arguments of the "Take your money and run" advocates. Apart from the sketchyness of their tax advice (the suggestion that a Canadian citizen could spend up to 180 days in Canada a year without being considered a resident for tax purpose is embarrassing. It is VERY difficult to lose residency for tax purpose, you basically have to leave the country for an extended period of time and stay out), they didn't seem like the type of arguments that were all that compelling. Everyone hates paying taxes, but you have to be able to tell a story about why your hating to pay taxes entitles you to special treatment.

For what it's worth, a $40/tonne tax on CO2 would raise in the neighbourhood of 20B per year by my very crude estimate. That's 10 times the revenue that would be lost through the CIT cut.

I'm with Stephen Gordon on this - my preference is for more consumption taxes (VAT) and some taxes that act against the accumulation of huge fortunes (especially multi-generational). One answer is an inheritance tax, another some income/wealth taxes with large exemptions. And then the one ring to rule them all of course - a citizen's dividend to make it all progressive. Then we forget all this fiddly acronese.

Joseph: Thoughtful comments, thanks. "I am less clear why a Roth Style savings vehicle isn't a good solution, overall." Going forward, I think it is - it's getting there from here.

Mike, carbon and marijuana taxes - as a bicycle loving Vancouver girl, how could I possibly object? But the central point is: how can we get people to cough up the taxes owing on their RRSPs and RRIFs? I don't see any good case for exempting these funds from taxation.

reason - sure - indeed consumption taxes, especially taxes on services, will also get at those RRSP/RRIF dollars. But even if we agree that, say, it would be a good idea to eliminate the corporate income tax and instead raise the GST by 3 or 5 percentage points, or to lower income taxes and raise the GST, there's still a question: should money withdrawn from RRSPs and RRIFs enjoy special tax treatment? My view is no, but like I said, pressure is building.


"Mike, carbon and marijuana taxes - as a bicycle loving Vancouver girl, how could I possibly object? But the central point is: how can we get people to cough up the taxes owing on their RRSPs and RRIFs? I don't see any good case for exempting these funds from taxation."

You're right - there isn't any. But as you point out, seniors are a powerful political group. My 63 year old father is already talking about the issue, so I expect the discussion to get louder and louder.

This is fairly terrifying. I'm pretty young, and all I see are storm clouds on the horizon as the boomers retire and seek to extract as much as they can from the rest of society. It'll be the biggest con in history if they manage to finagle tax-free withdrawals of RRSPs--a $600 billion theft from their children.

Now, on the topic of tax integration within RRSPs. I can see how the applying the dividend tax credit might work (as a refundable tax credit paid into the RRSP account). How would one manage to reduce the inclusion rate on capital gains in an RRSP? Allow half of the capital gain to be immediately withdrawn?

Very interesting post. One would think that the choicebetween IRA's and Roth-style IRA was mostly a gamble over present vs. future tax rates. But it can be important to look at other factors.

I think "take the money and run" is a workable solution here, because a retiree with accumulated savings can move to a country with lower cost of living and be much better off, even if she has to pay income tax to Canada for some time.

"Economists typically support such plans because of their effects on savings"

I've always been puzzled by this. Is it because they believe that households are too impatient, and so the government needs to increase savings demands above what they would normally be? No one cares that the increased savings demands might lead to excessively low interest rates, demand failures, or dynamic inefficiency?

What is the economic rationale?

Seems like the conclusion is that the savings rate is too low, which would seem like a tenable position, seeing as we're basically at or near historic lows.

Now, on the topic of tax integration within RRSPs. I can see how the applying the dividend tax credit might work (as a refundable tax credit paid into the RRSP account). How would one manage to reduce the inclusion rate on capital gains in an RRSP? Allow half of the capital gain to be immediately withdrawn?

Presumably it would have to be some variation of that idea. You'd have to require RRSPs (and, presumably pension plans, because there's no policy rationale for distinguishing between the two) to keep track of their capital gains and losses (since, presumably, you should only be able to take out net capital gains tax free, not gross gains) and allow people to withdraw amounts attributable to net capital gain at a 50% rate. In practice, that would be an administrative nightmare, particularly where RRSPs or pension plans hold securities on a short-term basis or trade them frequently (in which case, any gains, if they were taxable, would likely be on income account and subject to a full income inclusion, rather than capital gains). And while that might be feasible in, say, a one person self-directed RRSP plan, I can see the issue of how to allocate those capital gains amongst beneficiaries being a nightmare for a defined benefit pension plan. In practice, the idea's probably unworkable.

I wonder, though, if there's really a huge constituency for exempting RRSP withdrawas. I suspect, for most Canadian retirees, the marginal tax rate on the RRSP withdrawals (or pension income) is going to be the same or less than what they were paying while they were working (and they didn't rise up and demand that that not be taxed). I mean, take a retired 71 year old with $1,000,000 in RRSP assets. Assuming he's eligible for CPP and OAS (but subject to clawbacks) he's looking at a an income of 96,000 a year (based on the minimum RRIF withdrawals), but an average tax rate of only 27% (in Ontario). That's not an exhorbitant tax rate (and note, to the take your money and run people, if you become non-residents and take your money out of an RRSP, you're hit with a 25% tax rate anyhow, so query whether it's worth abandoning Canada to save a mere 2%. And querry whether retirees with RRSP assets of $1,000,000 are common. Obviously, this calculus changes for high net worth individuals, but that's not a large group (many of those articles are quite misleading in that they assume that taxpayers will pay the top marginal rate on their RRSP withdrawels, when in fact only a tiny percentage of Canadians, mostly of working age, have incomes that high) and its not a group that's likely to have polical success at pursuading the government to exempt them from tax.

Where I can see there being a sizeable political constituency is tinkering with the OAS clawback threshold (since that may affect a larger number of middle-class retirees) and/or the RRIF withdrawal rate (to better reflect demographics and the return on capital). The latter is really only a timing issue, not a real revenue loss to the government (in fact, it will be revenue gain if people don't blow through their RRSPs and so leave a big lump sum to be taxed at the top marginal rate when they croak - the RRIF withdrawals force a degree of tax averaging).

Andrew, here is a thought experiment:

Assume the government budget is balanced, and that all taxes are income taxes (independent of the source of the income). That means that the household sector gets back in factor payments (which increase income) exactly what it pays out in income taxes. So it is as if, at the sectoral level, there is *no* change in income. But some households will be taxed more than others (because they have higher incomes, or because the tax is progressive). That amounts to an endowment transfer to reduce inequality which is what the economists love. Notice that no one is moved away from the preference for current versus future consumption, so households are free to solve their own euler problem.

Now suppose that the government budget is balanced and that all taxes are consumption taxes. Forget about the inequality issues, now you are making present consumption more expensive, but are also increasing incomes due to factor payments from government by the same amount, so the net effect is to increase savings demands and decrease consumption demand.

I understand that during recessions and booms, the government will want to make short run adjustments, but what is the theoretical case for the need for a permanent adjustment? Are households unable to correctly do intertermporal optimization over the long run?

RJS: "I've always been puzzled by this. Is it because they believe that households are too impatient, and so the government needs to increase savings demands above what they would normally be? No one cares that the increased savings demands might lead to excessively low interest rates, demand failures, or dynamic inefficiency?"

I think one argument is that RRSPs are intended to offset the distortions created by a tax system whihc taxes income and creates a distortion in favour of current consumption over future consumption (i.e., savings). So the argument isn't so much that the government needs to increase savings ABOVE where they would normally be, it's that the government needs to RETURN savings to where they would normally be in the absence of the income tax distortion. I think you're assuming that a distorting income tax is the "normal" baseline, while I'm starting on the premise that it isn't.

RSJ: on the economic rationale for consumption taxes, see http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/12/does-canada-need-a-progressive-consumption-tax.html. Personally, I'm not in favour of moving entirely to a consumption tax, but I didn't want to get into that particular debate today.

Bob: "I wonder, though, if there's really a huge constituency for exempting RRSP withdrawas. I suspect, for most Canadian retirees, the marginal tax rate on the RRSP withdrawals (or pension income) is going to be the same or less than what they were paying while they were working (and they didn't rise up and demand that that not be taxed)."

That's only because they didn't really see how they were paying in taxes when they were working. This is the basic insight of prospect theory: people establish a reference point, and then really hate losses that take them below that point. When you've been doing someone's taxes, have you ever kept on looking for deductions and credits *until you got taxes owing down to zero* and then stopped looking? This is prospect theory. There's no reason why a credit that generates a $10 refund would be less valuable than a credit that reduces taxes owing from $10 to zero, but people don't think in those terms.

One argument for having some kind of mandatory RRIF withdrawals is the interaction between the pension system and the income security system. We don't want people collecting Guaranteed Income Supplement while they have hundreds of thousands of dollars sitting in their RRSPs.

I just gave an argument that income taxes are neutral for present versus future consumption.

What is your argument that income taxes encourage present consumption at the expense of future consumption?

RSJ, sorry, I've taught this so often that, sadly, the thrill is gone. Here's the standard argument.

Present discounted value of lifetime earnings=Y1+(1/(1+r))Y2.
Let S1=net savings (negative for borrowing). Assume discount rate=interest rate for savings and boring=r.
Then C1=Y1-S1, C2=Y2+(1+r)S1.
Present discounted value of lifetime consumption=C1+(1/(1+r))C2
=Y1-S1+(1/(1+r))(Y2+(1+r)S1)
=Y1+(1/(1+r))Y2

If you work through the math again with taxes you'll find that
present discounted value of lifetime earnings=present discounted value of lifetime consumption if there is no tax on interest income.

So this is the first argument - it's fair because it taxes lifetime earnings/ lifetime consumption. Savers and spenders pay the same amount in taxes over their lifetimes.

Actually the effect of taxing interest income on the level of savings is theoretically ambiguous - you'll notice I carefully skated over this issue by saying "their effect on savings" without actually specifying what those effects were! If you go to this page and scroll way down until you come to "taxes on interest income" you'll find a reasonably good discussion (some student has posted his/her prof's notes there by the look of it). Those notes point out that the effect of taxes on savings depends upon the relative strength of the income effect (cutting taxes makes you richer so that you spend more) and the substitution effect (cutting taxes makes saving more attractive so that you save more).


What makes sense is a reduction in the RRIF withdrawal rates. Taxes will be paid but even at the lowest withdrawal rate of 7.38% given to-days savings/GIC rates it can be readily seen that disaster looms for those relying on RRIF income for their retirement existance into their 80s and 90s. The bottom line: taxes will be paid. The downside for the taxing authority is a delay in the tax revenue reaching the government coffers.

Frances, you are modeling the effects of government surpluses, not taxes, and actually, of what would happen if money were thrown away entirely.

Taxes and income (at least for purposes of discussing income taxes) are monetary phenomena. If, in the process of taxing, spending, and providing public works, government consumes more output than it produces, you can think of that as a "real" tax if you want, but it's not the definition of tax appropriate for tax policy questions.


And that means that if the government budget is balanced, then households receive as income payments exactly the amount that they pay in taxes. Therefore if their income is taxed, it is as if there was zero effect, since some income is removed but the same income is supplied. The only effects will be income transfers among groups based on how progressive the tax rates are. That's what makes income taxes efficient.

But key to that assumption is that *all* forms of income are taxed equally.

So for example, assume that government taxes $100 of wage income, and uses the proceeds to purchase $100 of goods. If $70 are paid as wage income and $30 are paid as capital income by the producer of the output, then this amounts to a subsidy of capital at the expense of labor.

This will incentivize households purchase claims on capital rather than consumption goods, because the ownership of capital is worth $30 more than it would be if there was no government and no taxes.

This distortion is only eliminated if all income is taxed equally. You *must* tax interest income in order to avoid excess savings demands than would otherwise be the case.

Now we can ask, what about the distortions of government purchasing the $100 of goods?

But that's a whole separate issue, as it depends on what those goods are used for, and doesn't have much to do with tax policy.

RSJ. O.k., you're making a more macro argument. I'm not very good at macro. Perhaps one of the other bloggers will help out here. Here's my response.

"And that means that if the government budget is balanced, then households receive as income payments exactly the amount that they pay in taxes. Therefore if their income is taxed, it is as if there was zero effect, since some income is removed but the same income is supplied."

No, because people still have choices. They can avoid paying taxing on earnings by spending their time in leisure (or on Second Life rather than first life.) They can avoid paying taxes on interest income by spending rather than saving. Taxes will still have an effect on individual decisions even with a balanced budget.

If you believe that the quantity of savings is very responsive to changes in taxes (high compensated elasticity) whereas the quantity of labour supplied is not responsive to taxes(low compensated elasticity), then it makes sense to tax labour (earnings) and not savings.

On the wage v. capital income - I think the response is: Imagine the capital income is in an RRSP or IRA. It gets taxed when it's withdrawn from the RRSP/IRA and consumed, but not before. Is that a problem? Or if it's in a TFSA/Roth-style IRA, the earnings are taxed before they're invested in the form of capital.

Frances,

-- First, feel free to delete this post, or move it to the whack a mole blog.

-- Second, yes, you are right people have choices, and when I said that the income is unchanged, I was referring to private sector *nominal* income, or real *national* income. But private sector real income can be reduced, and this could cause people to produce less. My argument was not that government can consume an unlimited amount of output before the private sector stops producing it, but rather that for a given size of government, the least distortionary tax is an income tax that does not distinguish between types of income.

-- Third, no I don't support retirement plans or any tax deductions. If you want to help retirees, then cut them a check on the expenditure side. I think we have chronic excess savings demands in part because of these plans.

RRIF's are a nightmare. Frances is right in that they are a psychology failure, but not for the reasons she cites. First, they already receive preferential tax treatment when received. RRIF's are eligible for the Pension Income Tax Credit. That means that the tax on pension/RRIF income is credited against the credit allowance ($2,000) at the lowest rate. This effectively raises the tax threshold for RRIF's receivers to $17,000 when the PITC is tacked on to the Basic Person Exemption. Income splitting only enhances the opportunity to lower the tax payable. However most people are completely ignorant of these features of the tax system.

I seriously doubt that any RRIF benefciary pays more tax on receipt than on their original RRSP contribution.

Second, I have no patience for any person who complains about out-living their RRIF funds. The answer is simple: purchase a Life Annuity from a life insurance company. In the UK this is compulsory by age 75 for personal pension plans. This is one feature of UK tax law I really like. However Canadians are strange in their attitudes to life annuities. They will slobber at the mouth at the thought of an employer DB pension (which is a life annuity during payout) but would sooner roll over and die than purchase a life annuity with their own RRIF funds. Even though they cannot outlive their income in that case.

This makes even less sense when the tax system explicitly integrates DB and DC Pension room with RRSP room. The theory is that every dollar of earned income has a equal opportunity to receive tax-preferred treatment for retirement savings, regardless of the particular vehicles available to you. But personal life annuities are the brussel sprouts of the insurance world.

UK tax policy is much clearer on this point: Tax-preferred retirement plans are to be used for the purchase of a retirement income, not as a wealth accumulation and transfer vehicle, period. This results from the UK's history of class struggle and income inequality. Canada is much less clear on this point.

Not to mention the fact that financial planners are salesmen on commission. They can be good at what they do, and the do have products that are worthwhile, but there is an incredible amount of hidden and no-so-hidden greed there too.

RSJ: Since I'm not up to a math or diagrammatic proof, consider this analogy:

Assume you can spend your $100 income from working on apples or bananas. Apples cost $1 each, and bananas $0.50 each. In general (there are exceptions) you want the same tax rate on your $100 whether you use it to buy 100 apples or 200 bananas, or any combination.

Now replace "bananas" with "apples next year" and assume the real rate of interest is 100%. You want the same tax rate on your $100 whether you use it to buy 100 apples now or 200 apples next year, or any combination.

Interest income isn't really income. Future goods are cheaper than present goods (when the real interest rate is positive), just as bananas are (sometimes) cheaper than apples. We don't say the person who consumed 200 bananas in the first example made 100 bananas income.

It wouldn't surprise me to see children of boomers demanding access to RRSP funds with preferential tax treatment too. Currently, you get deal for buying your first house or if you (or your spouse) want to go university. I can see people demanding it be expanded to pay for university for their children (how many people really use RESP's? Most of my peers don't save for retirement never mind kids education!), but also for expenses for caring for ageing parents. If you're in a position of having improvident (or unlucky) parents who have no retirement savings AND you have kids who will be going to university about the same time your parents enter their dotage and must be cared for at great expense and you have a chunk of money sitting in your RRSP ... Well, the temptation is going to be very hard to resist. But it'll cost you 30% off the top, and then whatever you marginal rate is. Ouch. That's sure to piss people off. "I'm trying to pay for my kids education and look after my old parents and the Gov't is taking away half my savings!!!"

Oops: meant to say 30% off the top, and then again the difference with your marginal rate (so you might get some back, but probably you're going to pay even more).

holy smokes; are you even aware of the tax math of RRSPS

due to the combined effect of front end deduction/ back end tax, return on capital within an RRSP is EFFECTIVELY tax EXEMPT

do the math

so the only downside is a lifetime return on capital that is negative

# 1 that's unlikely

# 2 if it happens that's when you want capital loss tax integation

the only proper comparison is between an RRSP account and an alternative taxable account

end of story

anon: I think people here are aware that money in an RSP compounds at the before tax rate of return, not the after tax rate of return. And that an RSP is a better deal than a taxable account (except if you get a lousy rate of return and also have a higher marginal tax rate at the back end than at the front end).

But that's just the beginning of this story.

Patrick - this raises an interesting point: when people withdraw funds from an RRSP, that RRSP contribution room is lost forever. I don't know exactly how to solve this problem, but it's a serious one. One that doesn't exist with TFSAs - another reason why they're better.

Nick, I don't understand your argument.

Capital Income is income. Wage income is income.

Both capital and labor are factors of production.

Both receive payment arising from the production and sale of output, both have an opportunity cost, and both require time.

Yes, it is "natural" to have a positive interest rate, because there is an economic value add to supplying capital, just as it is natural to have a positive wage rate, because there is economic value add to supplying labor.

You expect both wage rates and interest rates to be positive.

That does not mean that wage payments and interest payments are not income.

When an individual household saves, it is sacrificing present consumption for future consumption. When it supplies labor, it is sacrificing leisure. Both of these sacrifices are compensated by the household receiving a payment in the form capital income or wage income. But the actual interest rates and wage rates have nothing to do (directly) with the sacrifice of the household, but with the marginal value add of the factor input. The sacrifice of the household only comes into play in that value add of the output will change as the quantity of the factor supplied changes, and this has to do with the sacrifice of households.

In the absence of government, the wage and capital income received is enough to purchase all the output.

But with government, it cannot be enough, because government is purchasing some of the output.

The household has to be content with public ownership of some output, but it will not receive enough disposable income to buy all of the output, and will need to be content with public ownership of some output (or having some output be thrown away, depending your opinion of government).

This is perhaps the biggest problem with identifying income with production, but it's not the only problem.

The only question is, what is the best way to reduce the ability of the household to purchase the output it produces -- do you reduce the (nominal) purchasing power of the household by subtracting its income (taxing the individual receiving the proceeds of the sale of output), or do you add a surcharge to the output being produced, taxing the buyer of the output?

Apart from a one-time inflationary hit, there isn't much difference between the two approaches. A perfect economic value add tax would be exactly the same as an income tax, except that it wouldn't be progressive and it would be much more complicated to enforce.

But what is critical is that all output be taxed -- both investment and consumption -- if you are on the output side, or that all income be taxed -- both capital income and wage income -- if you are on the income side.

If you only tax consumption goods, but not capital goods, then you are distorting the production system. You are making capital goods cheaper, and incentivizing the economy to produce more capital than is optimal. If you only tax capital goods, then you are making consumption goods cheaper, and are incentivizing the economy to produce more consumption than is optimal. You need to tax both equally. Same with capital income and wage income.

I personally favor the income side because it allows for a progressive tax rate and is generally cheaper to collect (in advanced economies, but not in developing economies).

Frances: "That's only because they didn't really see how they were paying in taxes when they were working. This is the basic insight of prospect theory: people establish a reference point, and then really hate losses that take them below that point".

Really? I'm pretty sure I see that withheld amount on my pay stub every two weeks. You're right, of course, that there's no withholding on the minimum amount that must be withdrawn from a RRIF (although there is withholding on all amounts that are withdrawn from RRSPs), so there are less favourable optics of having to cut the CRA a big check at the end of the year (or quarterly) as opposed to having someone else do it for you on a weekly basis. That said, I suspect the people who are complaining the loudest about the RRIF regime, would probably complain even louder if you told them they were going to have tax deducted from their withdrawals.

"One argument for having some kind of mandatory RRIF withdrawals is the interaction between the pension system and the income security system. We don't want people collecting Guaranteed Income Supplement while they have hundreds of thousands of dollars sitting in their RRSPs."

Maybe, although presumably the same result could be achieved with an asset test. Indeed, that would be preferable, since it would catch people under age 71 who have hundreds of thousands of dollars sitting in RRSPs, as well as people who might have hundreds of thousands of dollars sitting in non-income bearing assets in unregistered accounts. It isn't clear why it's more offensive, for example, for a person to be collecting GIS while sitting on millions of dollars of non-dividend paying shares in a RRIF than it is for a person to be collecting GIS while sitting on millions of dollars of non-dividend paying shares in a RRSP/non-registered account.

In any event, given the low threshold for clawbacks of the the GIS (and the generally low level of GIS payments) it's not clear how large population of individuals willing to live on poverty level income while sitting on a ton of assets really is. (Although I suppose it would be a bigger problem for OAS)

Bob "it's not clear how large population of individuals willing to live on poverty level income while sitting on a ton of assets"

Perhaps the people I know are cheaper than the people you know!

Frances: "Patrick - this raises an interesting point: when people withdraw funds from an RRSP, that RRSP contribution room is lost forever. I don't know exactly how to solve this problem, but it's a serious one. One that doesn't exist with TFSAs - another reason why they're better."

Presumably, you could solve the problem by amending the Income Tax Act to allow people to preserve their contribution room. It wouldn't be a difficult change.

And, in some respects, it would be a welcome one. One of the problems with a progressive tax system is that people with irregular income flows may end up paying more taxes than identical people with identical, but regular, income flows. The obvious example is a person (Person A) who makes $100K in year 1 and $0 in year 2 will pay more tax than an identical person (Person B) who makes $50K in year 1 and $50K in year 2, as a result of the progressive nature of the tax system. And RRSP regime which allows withdrawals and recontribution would allow Person A to smooth out his income flow over the two years (by making a contribution in year 1 and a withdrawal in year 2). Indeed, even now, I know that some people do this anyhow, notwithstanding that they lose the RRSP contribution room.


RSJ: Single period model with 2 goods. Draw a PPF and an indifference curve that are tangent to each other. Find the optimal tax on the two goods to maximise utility subject to generating a given amount of revenue for the government.

Now take the exact same diagram, but relabel the axes. It's now a 2-period model with one good. It's exactly the same problem.

Generalise to n goods and t periods.

Capital is the time-structure of production.

RSJ: "I personally favor the income side because it allows for a progressive tax rate and is generally cheaper to collect (in advanced economies, but not in developing economies)."

Sorry, but a consumption tax regime can have a progressive tax rate and can be every bit as easy to collect as a income tax regime. You're equating a value-added tax with a consumption tax, but a value-added tax is only one form of consumption tax. Indeed, for most Canadians, Canada's "income" tax system is really a consumption tax system, since they are granted a deduction for their single largest form of savings - namely RRSPs. That's the only difference between a consumption tax system and an income tax system, a deduction for savings. That's neither complex nor inconsistent with a progressive tax system.

And, keep in mind, a consumption tax does tax investment income, but only when it is consumed.

No, I meant that among the (non-discriminatory) tax regimes, I would favor taxing all income versus taxing all output.

This is because in terms of progressivity, to mimic the progressive income tax, you would need to send rebates to all but the highest earning households -- sending a different rebate to each person based on how much income they earned.

This requires that households collect and report their income data, but if they are going to do that, then you might as well collect the income tax. That's inefficient.

The alternative is to try to guess which goods the poor are more likely to buy, which is also inefficient (from a tax-collection/progressivity stand point).

Finally, you have the whole problem of elasticities (which are different for different goods).

It's just much better to tax the income across the board, and let the household desire what to consume or whether to consume.

I guess the worst of both worlds is to have *both* an income tax and an VAT tax, and to make them *both* full of deductions, biasing one type of good over another.

Moreover, a consumption tax does not have a deduction for savings, it has a deduction for *investment* -- I've noticed this conflation of savings and investment is common. Income taxes do not tax savings, either, they tax income, not uses of income. Neither of these taxes is a tax on savings, and yet the distinction between sources and uses of income is not sinking in. A source of income is not a use of income.

No one is talking about taxing savings, because savings are not taxed under any regime. The only example of a tax on savings would be something like a deposit fee, or a stamp tax, or a tobin tax.

It's the fact that investment goods are given a pass that makes the consumption tax inefficient.

In practice, with all the deductions and rebates, it's not clear what is really biased in any given implementation.

Bob Smith: that might be true if the contribution limit was proportional to income. But it's capped, so for high earners the income tax is just that; an income tax.

Bob "it's not clear how large population of individuals willing to live on poverty level income while sitting on a ton of assets"

Cheap or not, it seems plausible that old healthy people would do this as a hedge against being winding-up warehoused in a cut rate home with bed sores and a dirty diaper.

K: As I said, "for most Canadians" the income tax system is, effectively, a consumption tax system. People with high incomes are not "most Canadians". In practice, the subset of Canadians for whom the RRSP contribution limit actually binds is probably pretty small (at least in light of the vast quantity of unused RRSP space), more so now that people can also use TSFAs.

Patrick - possibly, but since you don't qualify for the GIS if your income is greater than $15.8K a year (for singles, $20.9k for couples), you have to wonder how realistic it is that someone will choose to forego the use of their registered savings (particularly if they have significant savings) to live at a poverty level (at least as measured by stats can's LICO - yeah, I know, they're not real measures of poverty) in order to hedge against the possibility that they'll have miserable standard of living down the road. I could see that being a problem with OAS, since the clawback on that occurs at a much higher income level, but not GIS.

"Moreover, a consumption tax does not have a deduction for savings, it has a deduction for *investment* -- I've noticed this conflation of savings and investment is common. Income taxes do not tax savings, either, they tax income, not uses of income. Neither of these taxes is a tax on savings, and yet the distinction between sources and uses of income is not sinking in. A source of income is not a use of income."

I'm sorry, what do people do with their savings if they don't invest it? Put it in their sock drawer? (Well, I suppose for some small subset of the population). This is econ 101, investments equal savings.

As for the claim that income taxes don't tax savings, let's go back to first principles. Let's start from the basis that income (Y) equals consumption (C) plus savings (S) (or you can call it investment if you want).

So,
Y = C + S

Now, let's impose a tax T on income, so after tax income equals

(1-T)*Y

And since Y = C +S

(1-T)Y = (1-T)(C+S)

The individual's tax liability is equal to T(C+S). Now, tell me again that an income tax doesn't tax savings.

The distinction between taxing income and taxing uses of income is a false one (think about it in another context, imagine arguing that a tax on cigarettes is not a tax on smoking cigarettes. Obviously, that's a ludicrous argument). All income is used, either to consume or to save. So if you're taxing income, you're also taxing consumption and savings.

In contrast, a consumption tax, by definition, is only imposed on C.

"The individual's tax liability is equal to T(C+S). Now, tell me again that an income tax doesn't tax savings."

You mean saving, not savings.

And income tax doesn't tax saving in the sense that the composition of Y = (C + S) doesn't affect income tax.

E.g. income tax is the same whether Y = C and S = 0, or Y = S and C = 0.

Anon: And income tax doesn't tax saving in the sense that the composition of Y = (C + S) doesn't affect income tax.

I'm not sure what that means. Yes, you income tax liability is the same regardless of the uses of your income, but that simply means that you're taxing current consumption and savings at the same rate. It is the same as having a consumption tax and a savings tax. But that isn't the same as saying that it doesn't tax savings. In fact, your example proves the point, if C=0 and Y=S, than an income tax IS a savings tax. Put it in a different context, one could also describe income as being equal to wages, business income and income from property (i.e., Y = W(ages) + P(rofit) + R(eturn on capital)), but you wouldn't say that, because the composition of your income doesn't affect your tax liability, that we're not taxing wages, profit or returns on capital.

Income taxes do not even appear in the NIPA. They are invisible.

This is because the loss of income due to taxes paid is balanced by the gain in income due to the government spending the proceeds.

Of course, for each *individual* this may not balance out, but in aggregate it balances out. That means some individuals are winners while others are losers, and all the effects come from these relative transfers -- whether one product is taxed more than another, or whether one person is taxed more than another.

But it's just wrong to assume that government taxes without also spending, and to look at the effect of taxation on income as if there was no spending.

The only taxes that appear in the NIA are taxes on production and imports, not income taxes.


Income taxes may not appear in the national income account's, but so what? Income taxes (and all taxes) are visible to individual taxpayers, who respond to them accordingly. Moreover, you say that the loss of income due to taxes paid is balanced out by the increase in government revenue, but that's only true if you're using a tax that doesn't have a deadweight loss (like a poll tax). You're basically ignoring the cost of taxation.

Economics is just reduced to equatations asuming efficient markets?

Those savening plan things are horrible, all of them. The problems are not reduced to prospect theory which is economics in the first place or not?

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad