Tax cuts should deliver equity, or efficiency, or both. Income splitting does neither.

The ideal tax system reflects a compromise between two conflicting goals: equity and efficiency.

Equity requires that those who are able to pay more taxes do so. It means taxing the rich and giving to the poor, in thereby reducing inequality and ameliorating poverty. Hence equity demands relatively high marginal tax rates. 

Efficiency, on the other hand, argues for lower marginal tax rates. Taxes are inefficient to the extent that they distort people's choices. But choices are made on the margin - to work an extra hour, to save an extra dollar. For example, when deciding whether to work overtime, a rational decision-maker will compare the pain of working longer hours with the gains from doing so. These gains are shaped by the tax rate on the last dollar earned. Hence what matters for efficiency is marginal tax rates.  

Intelligent people disagree about how much efficiency should be sacrificed in the name of greater equity, and vice versa. Some favour higher rates and more progressivity; others place greater weight on efficiency, arguing for a broad tax base and a low tax rate.

But economists agree: tax reductions should deliver improvements in equity, or efficiency, or both. Income splitting does neither.

I am using the term income splitting here to refer to policies that allow a person to transfer money to his or her spouse for tax purposes. Income splitting reduces a couples' tax liabilities when one partner has a low income and other has a high income, because the income transferred to the low income spouse is taxed at that person's own, relatively low, marginal tax rate.

Income splitting reduces efficiency because it raises the effective marginal tax rate faced by the "secondary earner" - by which I mean the lower-earning of the two spouses. Suppose, for example, income splitting allows Ahmed to transfer $50,000 of his earnings to his partner Betty - an at-home spouse - for tax purposes. If Betty decides to enter the labour force, some of the tax savings achieved by income splitting will be lost. At the same time, Betty will have to pay EI and CPP premiums. Adding together the loss of tax savings and social insurance premiums, Betty will face a considerably higher marginal tax rate than she would have faced in the absence of income splitting. 

One could note, of course, that Ahmed's effective marginal tax rate is reduced by income splitting. But the point is: the labour supply of "secondary earners" is much more elastic than the labour supply of primary earners. What the vast labour economics literature on this subject tells us is that, if Ahmed's marginal tax rate falls, he probably won't change his labour supply very much. Betty, on the other hand, is much more likely to be affected by a change in her marginal tax rate. 

Income splitting reduces the marginal tax rates of people who aren't very sensitive to changes in tax rates, and it increases the marginal tax rates of people who are sensitive to changes in tax rates. That's why it has efficiency costs.

Income splitting also compromises equity. One interpretation of equity focuses on the vertical: relieving poverty and reducing inequality. The benefits of income splitting go primarily to higher income Canadians (see, for example here), so it does nothing for vertical equity.

Some might say income splitting promotes horizontal equity, because it allows single earner couples and dual earner couples with the same total income to pay the same total taxes.

Yet the proper basis for making equity assessments is ability to pay taxes. A single earner couple always has the potential to increase their income by having the second spouse enter the labour force - this is how Canadian families have been maintaining their standard of living in the face of declining male wages and rising house prices for the last several decades. The earnings capacity of the single earner family at $100,000 is greater than the earnings capacity of the dual earner family at the same income level - and with greater earnings capacity comes greater ability to pay taxes. 

Moreover, the single earner couple enjoys the benefits of having an at-home spouse - having someone there when the children come home from school, someone to stay home with the children when they're sick, someone with time to organize all the stuff in life that needs to be organized.

Dual earner couples have a lower standard of living than single earner couples at the same income level. Therefore horizontal equity requires that they pay lower taxes.

Income splitting delivers neither equity nor efficiency benefits. Also, income splitting at the federal level would create pressure for income splitting at the provincial level, further compromising the ability of provinces to balance their budgets.

It's a bad idea. My message to the government: don't do it.

Comments

[Taxrage]

It's been a thorn in the side of 1 and mostly 1-earner families since 1968 (Carter Commission), who are about the only group which has no special tax accomodation for *their* choice.

If the issue hasn't gone away after 45 years, it's never going to.

Let's get it done, so we can move on.

[Jeremy]

Would your analysis of the ability to pay taxes be modified by incorporating childcare costs and their tax treatment? I would guess that the tax incentives arising from income splitting are overwhelmed by the tax treatment of childcare? As in the marginal rate for a spouse working is much higher because they can not deduct all of their childcare expenses.

Furthermore income splitting currently exists for those with proper employment structures and for those with large amounts of invested assets. Does the existence of income splitting at higher levels of income change your analysis of what is fair? Should all income splitting strategies be stamped out? Should pension splitting be allowed?

Also, that C.D. Howe paper doesn't seem correct, it seems to ignore tax credits in its analysis (which would have a bigger impact on lower income levels). (Look at the analysis for Alberta, There would be a benefit for capturing the non-refundable tax credits lost each year the spouse didn't work, but the paper indicates the maximum benefit for the spouse of zero). Maybe that's a better idea, allowing for the carry forward of non-refundable tax credits that are unused?

[Bob Smith]

Agree wholeheartedly with your analysis Frances.

It's interesting that there was a story in today's Globe suggesting that the original income splitting proposal was being sharply scaled back (the comment by Mintz is telling). As I suggested in an earlier discussion I wouldn't be surprised if we saw some sort of income splitting for families with infants (when typically one parent takes time out of the workforce anyhow, so the efficiency concerns are less problematic, and it would make it easier for parents to make full use of their maternity leave) or a change to the child care credit regime to give recognition to the opportunity cost of being a stay-at-home parent. Either of those would give the Tories some political cover, while avoiding the hefty cost and nasty political optics of their initial proposal.

In terms of the politics, I know lots of Tories who would really benefit from income splitting. None of them will change their vote over the issue. They'd be quite content with an increased UCCB.

[Livio Di Matteo]

Frances:
I agree with your analysis but bear in mind that there already effectively are some forms of income splitting being allowed in the Canadian system:
CPP sharing for spouses and partners who are at least 60 years old.
Loaning money to a spouse.
Hiring your spouse or child.
Spousal RRSPs.

[Chris S]

Here's my example, but I cannot see how to analyze it in your framework. I call it the "job splitting paradox".

Family A: Parent 1 works in both the morning and the afternoon, and earns $100,000 per year. Parent 2 stays at home in both the morning and the afternoon, and earns $0 per year.

Family B: Parent 1 works in the morning and stays home in the afternoon, and earns $50,000 per year. Parent 2 stays at home in the morning and works in the afternoon, and earns $50,000 per year.

But Family A pays significantly higher taxes than Family B. In Ontario, a rough estimate is that Family A would pay $26,440 in taxes, while Family B would pay $17,400 in taxes. (* https://simpletax.ca/calculator ).

Under both circumstances, the family always has one person at home, one person at work, and earns family income of $100,000 per year. Why should Family A pay over $9,000 more in taxes?

[Bob Smith]

Livio,

That's true, but Frances' reasoning really only applies to income splitting with respect to labour income, rather than with respect to investment income. After all, the ability to pay taxes, as she defines it is the same for a retired couple with two pensions (or RRSPs/RRIFs) as it is for a retired couple with one pension (RRSP/RRIF).

It's worth noting that in other areas, the tax act has strict policies against income splitting strategies (for example, attribution rules for income derived from property gifted to spouses/minor children, attribution rules for certain trust income, etc.). Moreover, while you can achieve income splitting by hiring a spouse or a child, that only works if they are paid a reasonable fee for the work they do - i.e., they have to be bona fide employees. Obviously, if they are doing actual work, Frances objections to income splitting are mitigated. Similarly, you can lend money to a spouse but only if you charge a prescribed interest rate (which is, admitedly, quite favourable).

[Bob Smith]

Livio,

Presumably there are more plausible examples than that one. Why does a couple who work full time and make $100K and $50K respectively pay more in tax (roughly $2K) than the couple of who make $75k each. That's surely a fairly common scenario.

Moreover, there's a broader counter to Frances argument, if higher taxes can be justified on ability to pay (measured by time endowment), surely it would make sense to impose tax at a higher rate on the person who works part time and makes $30K a year than on the person who works full time for the same income. Can we tax endowment or does that open-up a whole can of worms?

As a practical matter whether you allow income splitting or not, you're going to end up with results that are "unfair" by some reasonable measure of fairness.

BTW, good series of posts Frances.

[Bob Smith]

Sorry, that last comment should have been address to Chris.

[Chris S]

Bob;

Where I find Frances' argument lacking is that apparent focus on the individual. Only briefly touched on is that the economic unit a family sees foremost is - the family. As a society, we would disapprove strongly of an individual in a family who primarily makes economic decisions based on their individual preferences.

The starting point of income splitting is that there is a family - treating the two people as individual earners, one who happens to be primary and one secondary, leaves until much later the point that this is not how that pair will likely make their economic decisions.

I agree that my example is not usual. But it is possible, and it expressly structured to show that a one-income and two-income family where all other things are equal except which partner earns what fraction of the income leads to family level inequality. It's also simplified - it doesn't touch CPP and EI, which would reduce the tax inequity, but potentially raise the later inequity in that the split income couple ends up with both a higher level of insurance and pension.

This leaves almost entirely untouched the question of quality of childcare; that is, parent vs hired caregiver. This is important because many of the proposals on the table making splitting dependent on their being dependents. This gets us deep into non-financial value (which must still be assessed financially as part of any decision), along with a dump truck load of worms that gets opened.

[Livio Di Matteo]

Bob:
Labour income versus investment income - as Kenneth Carter might say - a buck is a buck no matter where earned. Why should one be denied income splitting and another granted it given that both forms of income represent accretions to economic power?

[Leo]

Chris,

What job pays $50k a year for 20 hours of work a week and lets you work 4 hours a day?

[Bob Smith]

Chris,

I agree that your example (or mine, for that matter) illustrates a real unfairness of the current tax system. On the other hand, I think the example given by Frances equally illustrates the unfairness of a system with income splitting (i.e., two families with identical incomes, paying identical taxes, but one has the value of home production). Given the diversity of family arrangements, I don't think there's an easy way to close that gap. And I do worry about the incentive effects of imposing high marginal tax rates on secondary earners (which, heretofore, has generally been woman, but given recent trends, likely won't be in the future).

I suppose the theoretically "right" answer would be to allow income splitting but tax home production/leisure - it's non-taxation is the source of the distortion. I suppose you could do that, though how you value that is beyond me (can I argue that my tax bill should be lower because my stay-at-home spouse is a lousy cook?) and the inability to pay your tax bill with freshly baked cookies raises practical concerns.

I'm with you with respect to raising children, and given the recognition of the cost of caring for children that the Tax Act gives for paid care, I think there's a good principled case for given similar recognition to reflect the opportunity cost of staying home to care for children. As I suggested above, I could see the government making a change along those lines as a sort of fig leaf for breaking the income-splitting promise.

As an aside, the mention of CPP raises one of the more significant inequalities IN FAVOUR of one income families. Under the current regime if you die, your spouse is entitled to survivors benefits. Fair enough, but they can't collect CPP benefits in excess of the maximum. In other words, if you worked and your spouse didn't, and you die, he gets to collect your full pension until he croaks. On the other hand, if you worked and your spouse worked, and you both get the maximum CPP benefit, when you die, your spouse gets what he would have received anyhow. In effect, the value of CPP benefit is greater for a single income family than a two income family.

[Bob Smith]

"Why should one be denied income splitting and another granted it given that both forms of income represent accretions to economic power?"

Fair point. Although, presumably if you were taking the Carter prespription seriously, we would also be taxing the value of home production (why treat a dollar of home production differently than a dollar of market income?). In the absence of an ability to tax home production directly, that might suggest having higher tax rates on,investment income than on labour income (since in the case of the latter you have to forego home production, while in the case of the former you don't). Though there are compelling efficiency rationales for not doing that.

I haven't thought it through, but what are the incentive effects of income splitting for pensions vs. income splitting for labour income. In the latter case, it's likely to distort the labour/leisure decision, does income splitting for pensions have the same impact on savings? I don't think so, but I don't know.

Of course, one might argue that the fact that we allow pension splitting for seniors doesn't make it good policy (though I don't see anyone trying to reverse it).

[Taxrage]

Yes, gotta love pension splitting. No limit or age restrictions. Max $30K benefit for someone like former Nortel CEO John Roth. Not a peep of opposition from the mainstream media.

Whoa! Rewind. Are we talking the same mainstream media that has been fighting income splitting from the get-go?

Which represents more fairness: a $2K break for a $60K bus driver supporting a wife and 2 kids, or a $30K break for John Roth, who walked away from Nortel with $100M and a $750K/year pension? Which is facing more opposition from the media and progressives?

[Determinant]

Frances, could I have you saying this on a Youtube video so I could repeat it to my brother? My brothers works while my sister-in-law is a stay at home mom. My brother is all in favour of income splitting and thinks he has all the objections worked out. I believe he would limit it to people with children under the age of 18. I think he's crazy, but he is an accountant.

Guess what we talk about at family dinners? (yeah, yeah, we're a family of political geeks).

[Bob Smith]

Since when has the CD Howe Institute been labeled a collection of "progressives"? Certainly, I've been called lots of things in my life but that's a new one for me.

You know, when the Broadbent Institute and the CD Howe institute agree on something, maybe they're on to something.

[Bob Smith]

Determinant,

Don't know if you saw the piece in the Globe this morning, but the accountant who was the poster boy for the Tories on this issue in 2011 has changed his mind on the issue - see there's hope for accountants.

It was also interesting to see that Rhys Kesselman, who was one of the originators of the TFSA concept (and so someone who the Tories have listened to in the past) come out strongly against this proposal.

[Frances Woolley]

Determinant - we try to avoid talking about this at family dinner parties, though not always successfully. There is absolutely no other topic that causes bitter division within my family the way that income splitting does.

Taxrage, Bob, others writing about Carter: Carter advocated adopting the family as the tax unit. It did not advocate income splitting. The report suggested having a separate tax schedule for families that would be higher than that paid by two individuals with the same total income. Consequently, two people with equal incomes would face a marriage penalty upon marriage.

There is absolutely no reason why adoption of the family as the tax unit should necessarily involve large tax savings - the family could be taxed, as under the British system, at close to the rate of a single individual.

[Taxrage]

It has caused division for over 45 years. Like bilingualism, this is a political issue, more so than an economic one.

I'm not sure if $3B is significant at the federal level. $3B buys you a 12 km LRT line in Ottawa (that goes nowhere). What I can tell you, though, is that $7,000/yr is a lot of money for a family, especially a 1-earner family.

What simply has to happen, now that the issue is front and centre, is that this differential must be scaled down to a level that removes it from the policy radar for the next 45 years. Perhaps something like $2,500 (higher tax paid by a 1-earner vs a 2-earner) penalty would be more palatable; I'm even being generous here. Few 1-earners would begrudge their 2-earner counterparts a minor tax advantage, but not to the tune of $7,000.

Economic arguments will come up short at the ballot box. The 30% of families who look after their own children rather than pay someone else to do it have waited a long time...and the iron is hot.

[Frank Restly]

Frances,

"The ideal tax system reflects a compromise between two conflicting goals: equity and efficiency."

"Equity requires that those who are able to pay more taxes do so. It means taxing the rich and giving to the poor, in thereby reducing inequality and ameliorating poverty. Hence equity demands relatively high marginal tax rates."

Is equity a nominal measurement? Normally we don't think of governments taking real goods (houses, cars, land, etc.) and giving them to the poor.

"Efficiency, on the other hand, argues for lower marginal tax rates. Taxes are inefficient to the extent that they distort people's choices. But choices are made on the margin - to work an extra hour, to save an extra dollar."

Is efficiency a real measurement? If tax policy increases / decreases hours worked and thereby increases goods produced, this is a real effect.

Can a tax system achieve both a nominal object (equity maximization) and a real objective (increased efficiency)? I think you need to combine tax / fiscal policy with monetary policy to achieve both.

[abhay]

I don't think the objective of income splitting is about equity or efficiency (that is what progressive taxes/ Laffer Curve is all about). Income splitting provides an incentive for couples to live together and (to put it more bluntly) for secondary earner to focus on raising kids. So oppose it for what it is, a policy designed to promote the conservatives' view of ideal family structure.

[JF]

In my view the characterization of progressivity and the notion that it sacrifices "efficiency" is parroting forward messaging, and would prefer better.

Good tax policy wants tax burdens fairly shared. This means that the tax's design features must try to make sure the burden felt by a high net worth person approximates the burden felt by others. Tax rates often progressively decrease in order to equalize this weighting, but it often cannot change the plain fact that a tax that collects a low income person's "marginal" dollar establishes a huge burden on this person (distorting basic demand freedoms-of-order in the economy, in my opinion). The same tax burden for high net worth people is only approximated by design features - a valid understanding of utility concepts. BUT, unlike the text of the blog author - this is not about "giving to the poor" and its purpose is not to ameliorate poverty - this is mis-characterization and propaganda to say it this way. Use of progressivity provisions in tax systems and tax regimes is "equity" in the classic sense however - tax burdens should be shared and we should try to equalize them and a tax regime can also use features that are just, as well (for instance, the tax base definition can exclude the basics from the calculation of the base so little to no collection arises from the poorest).

Efficiency is not a code word for allowing policy makers to have design features that favor those of high net worth. These individuals can and do: split, shift between income and wealth, afford to move, afford to buy elsewhere to avoid taxes, etc. etc. They are legally permitted to do so and have the net worth to back themselves up. Agree though, the productivity of any individual tax regime needs to figure out when its rates or other design features influence avoidance and non-compliance behaviors because of behavioral responses from those who bear the incidence. It should be obvious that I personally feel that people of high net worth or current income like to work and such distortionary behaviorial responses to a tax policy are overblown concerns and do not deserve the notion that there is much of a trade off here.

But it is true in all tax policy that you cannot burden an object of taxation to the point that you bury it - though cigarette taxes are an example of a taxed object where this policy is being pursued. Fair, open debate and valid analyses can inform, and productivity and tax-administration-efficiencies are to be considered.

Hopefully, the debate in Canada is better informed and less accepting of these simple misleading characterizations of tax policy making.

You know, I trust that to be the case, but would appreciate more questioning of the author, as found in the Frank Restly's comment above.

[Frances Woolley]

JF "Efficiency is not a code word for allowing policy makers to have design features that favor those of high net worth"

Absolutely. Base-narrowing measures are generally bad for economic efficiency, and schemes that favour those of high net worth not infrequently have the effect of narrowing the tax base.

Moreover, as Nobel Prize winning economist Amartya Sen has noted, an economy can be efficient and still be perfectly horrible.

But if a policy increased both equity and efficiency, it would be brain dead not to implement it. Which means either (a) all such policies have already been implemented, hence we face a trade-off or (b) policy makers are brain dead.

I will refrain from making any comments about the state of the US political system.

[Bob Smith]

"Few 1-earners would begrudge their 2-earner counterparts a minor tax advantage, but not to the tune of $7,000.":

And few 1-earners would see anything like a savings of $7000 with income splitting. The CD Howe study suggests that the average tax relief for all single earner families (and families where the second earner makes less than 15% of family income) would be ~$2,700, assuming that the provinces follow the federal lead (unlikely). That's the average benefit to high income earners (people making more than $125K) in that catagory is only $6,400 (again, which assumes the provinces follow the federal lead). So let's not pretend that your $7000 number has any meaning.

[Bob Smith]

"Can a tax system achieve both a nominal object (equity maximization) and a real objective (increased efficiency)? I think you need to combine tax / fiscal policy with monetary policy to achieve both."

That'a good point. We've got multiple policy objectives (efficiency and equality) and multiple policy instruments (tax policy, government spending, regulation, etc.) why does one policy have to advance both objectives? Could you, for example, choose an efficient, but less equal, tax policy with a view to raising revenue (for example, by taxing consumption rather than income, or taxing immobile wage income at a higher rate than highly mobile capital income), but use the revenue to advance the equity policy objectives. I think a lot of the problem with tax discussions in North America, is that tax policy is being treated as the only policy tool to address inequality.

[Frances Woolley]

Bob "Could you, for example, choose an efficient, but less equal, tax policy with a view to raising revenue (for example, by taxing consumption rather than income, or taxing immobile wage income at a higher rate than highly mobile capital income), but use the revenue to advance the equity policy objective"

This is, to some extent, the Scandinavian model, which Stephen Gordon often advocates. Here we could shift to expenditure taxation, but the idea of raising serious amounts of revenue through a value-added tax is a non-starter because of cross-border shopping, (puzzling, to me) inability of governments to impose tax-inclusive pricing.

[JF]

Bob Smith - Did you really mean to suggest that because capital is mobile that tax burdens should shift therefor to immobile wage/salary earners? Is this tongue-in-cheek? Do you really think that way?

Smith and Woolley - Tax policies have not all been implemented. A better revenue system, in my opinion, would also include a net worth tax regime among the various tax types, one that puts all net-capital positions into the base regardless where they are in the world, as it does in fact define the taxpayer's net worth (so by-by mobile tax administration concerns). This system of tax regimes would allow lower rates across all tax types simply because net worth is so large, so the tax rates in this regime can be quite small.

The public seems to accept benefit-type tax regimes readily (e.g., fuel taxes). This type of tax is something to look at carefully - and who in fact can argue that a net worth tax is not simply a benefit-tax - someone's net worth has current and temporal value because of civilization and the public's laws and government - something of immense daily value - the tax simply reflects the benefit the taxpayer is receiving, as measured by their net worth (of course, design features would likely bring basic exclusions and progressivity ito its design). So as long as revenue systems can't get at net worth, by political choice in some countries, it is limited to income streams as a tax base, and this plainly shows that not all tax policies have been implemented.

I realize I am a pollyanna-thinker here but in terms of mythology, I just hope people will recognize that there is nothing mystical about income taxation, or limiting yourself to cash flows as a tax base. There are many possible tax regimes that could be used.

Oh, and absolutely agree - the US political system is benighted. I look to Canada for ethics, morality, common sense - and leadership. So go advocate for the addition of a modest net worth tax regime so you can compete for job-creation by lowering payroll taxation burdens (ok, that is my thinking about how to use the revenues, you don't have to agree). If you did this, then a better discussion would ensue here in the US where I reside. Shine some light for us, please.

[Miguel Madeira]

"Yet the proper basis for making equity assessments is ability to pay taxes. A single earner couple always has the potential to increase their income by having the second spouse enter the labour force "

But if we are talking, not about a family with a stay-at-home wife, but about a family where both partners work, but one earns more than the other (for example, a lawyer married with a nurse)?

[Bob Smith]

"Here we could shift to expenditure taxation, but the idea of raising serious amounts of revenue through a value-added tax is a non-starter because of cross-border shopping, (puzzling, to me) inability of governments to impose tax-inclusive pricing."

I'm not sure cross-border shopping is really that obvious a barrier. The CBSA has the authority to collect GST (and HST) at the border on importation. In practice, they usually waive it for small amounts, but that could be changed as a matter of policy. The only real gap for commodity taxes would be online intangible property (e-books, music, etc.) which are difficult for governments to tax.

[Majromax]

> The CBSA has the authority to collect GST (and HST) at the border on importation. In practice, they usually waive it for small amounts, but that could be changed as a matter of policy.

If anything, it acts a protectionist measure. Customs collections for those small-value items that *are* assessed is generally a small amount for the G/HST plus a customs-broker-specific flat fee. For smaller items, that flat fee is easily as large as the tax itself.

[Bob Smith]

"Bob Smith - Did you really mean to suggest that because capital is mobile that tax burdens should shift therefor to immobile wage/salary earners? Is this tongue-in-cheek? Do you really think that way?"

Why not, that's what the Swedes do (see http://www.kpmg.com/Global/en/IssuesAndInsights/ArticlesPublications/taxation-international-executives/sweden/pages/income-tax.aspx#2). Earned income and business income is taxed at marginal rates of between 51-56% on average (depending on municipal taxes), admitedly with generous allowances so that the tax system is neverthless quite progressive (Sweden has a very flat, and high, rate structure, but you only start paying tax above a reasonably high threshold). Capital income is taxed at a flat rate of 30%.

There's a good theoretical and empirical basis for taxing capital different then income from labour - the marginal cost of public finance on taxes on capital tend to be significantly higher than that on employment income or consumption. Since a lower marginal cost of public finance allows governments to raise more revenue without crippling their economy, countries with larger government sectors generally rely more heavily on payroll taxes and VATs than small government economies. Sweden isn't alone among European countries in taxing capital income more favourably than labour income (or in heavily taxing consumption. If you insist on taxing all income at the same rate, in effect you sharply limit the amount of revenue that governments can raise.

While the efficiency arguments for lower taxes on capital on untouchable, they obviously raise equity arguments (since capital is distributed less equally than labour) adn political barriers. Mind you, that make a compelling case for using a consumption tax base rather than an income base. While a consumption tax doesn't have the same efficiency implications of taxing capital income, it nevertheless allows you tax the owners of capital, at least when they choose to consumer their capital.

A VAT like the GST/HST is one form of consumption tax, but so is, ironically, the Canadian personal income tax system (at least for most Canadians). To the extent that your only savings are in your RRSP or your TFSA (which is probably a fair discription of many Canadians), our tax system only taxes you on your consumption, not your income.

[Bob Smith]

"A better revenue system, in my opinion, would also include a net worth tax regime among the various tax types, one that puts all net-capital positions into the base regardless where they are in the world, as it does in fact define the taxpayer's net worth (so by-by mobile tax administration concerns). This system of tax regimes would allow lower rates across all tax types simply because net worth is so large, so the tax rates in this regime can be quite."

That's a theoretically sound suggestion, but it raises a host of practical implications. How do you measure net worth? It's a tough question - one the Canada Revenue Agency often struggles with when it tries to assess tax evaders (since often the only way to figure out how much income they didn't report it to take their net worth, substract reported income, and treat the balance as unreported income). How do you deal with non-marketable assets? It's easy enough to figure out what my RRSP is worth, it's all invested in listed securities. How do you figure out what someone's private business is worth? Not so easy.

How do you deal with liquidity concerns. If you're a 90 year old who bought a house in Toronto in the 1950's you may have a hefty net worth, with little income. It's hard to monetize a private company. How do you deal with volatity in net worth (i.e., if you tax increases in net worth, presumably you have to give refunds for declines in net worth - that would have had a nasty impact on the fisc in 2008 when markets plumetted).

And of course the problem with taxing offshore income is not that we can't tax it - we do (and, in fact, in Canada we impute income on certain offshore passive investments which don't otherwise give rise to reportable income). The problem is identifying offshore income when people don't report either their offshore income or assets (evaders). A net worth tax doesn't address that problem.

[Taxrage]

"assuming that the provinces follow the federal lead (unlikely)"

Did you miss the recent media stories about how the provinces would pretty much have to follow the feds lead, unless they're going to amend their provincial ITAs?

Did even one of them try to opt out of pension splitting?

[Taxrage]

"So let's not pretend that your $7000 number has any meaning.".

Assuming the provinces don't change their ITAs, this is the approximate tax penalty that any Ontario family a single $100K income pays. Why is that meaningless?

[Taxrage]

Here was one of the media stories on the topic of provinces being in or out of income splitting: http://business.financialpost.com/2014/09/18/income-splitting-canada-provinces-cost/

Note 2nd paragraph.

[Kaleberg]

What is the argument that taxation cuts efficiency? I think that is an unquestioned assumption with no evidence.

Untaxed markets aren't particularly efficient in any sense of maximizing utility. They can only perform local, short term gradient based optimization, and they are easily gamed by accumulated capital and various forms of economic power. (At least that's what the mathematics shows. Real world optimization algorithms usually have to limit the mis-allocations caused by allowing accumulation or avoid them completely by not allowing accumulations.) In the economic world taxation that prevents capital accumulation usually results in higher efficiency as it allows higher utility of asset allocation.

[Frank Restly]

Kaleburg,

The argument is that progressive taxation distorts efficiency. Progressive taxation has the effect of increasing taxation on each additional hour worked and each additional good produced. In that way it creates an incentive to work fewer hours and produce fewer goods - that's the theory anyway.

Many tax experts recommend a system of flat taxation. Whether that taxation is applied to income only or net worth, whether that flat taxation is high or low relative to GDP is open for debate.

Bob,

"How do you deal with liquidity concerns. If you're a 90 year old who bought a house in Toronto in the 1950's you may have a hefty net worth, with little income. It's hard to monetize a private company."

Good points. Most governments don't try to address differences in real equity, only nominal equity.

[Bob Smith]

Kaleberg,

There's, literally, an ocean of empirical research on the efficiency implications of taxes - distorting choices between labour and leisure, consumption and savings and between alternative forms of consumption or investment. That taxes generally have adverse efficiency implications (generally, because Pavlovian taxes, such as cigarette taxes or carbon taxes can be efficiency increasing) is as much an unquestioned assumption as the assumption that apple's fall down.

Moreover, in the context of tax policy discussions, efficiency has another connotation apart from (but related to) allocative efficiency. Different tax regimes are more or less efficient at raising funds (in terms of the cost to society of an extra dollar in tax revenue - which reflects, in part the distorting effects of the tax instrument). Whereas taxes likes VATs and payroll taxes are generally more efficient tax instruments, while taxes on capital or narrow sales taxes tend to be less efficient.

[Squeeky Wheel]

To play devil's advocate:

"Moreover, the single earner couple enjoys the benefits of having an at-home spouse - having someone there when the children come home from school, someone to stay home with the children when they're sick, someone with time to organize all the stuff in life that needs to be organized."

Given all the advantages of single earner couples, perhaps there should be an *incentive* towards that family structure. I've seen some evidence that children in single-earner families do better in school, and have higher wages after school. One could frame the question as an investment in future efficiency and productivity. Alternately, one could propose positive externalities to single earner families such as higher parent-teacher involvement, higher volunteerism for local schools and charities, lower rates of hooliganism, etc. Where positive externalities exist, it generally makes sense to apply government funding.

Finally, why do we want to so much work anyway? Increased economic activity my not lead to greater happiness http://theconversation.com/why-our-happiness-and-satisfaction-should-replace-gdp-in-policy-making-30934

[Bob Smith]

"Assuming the provinces don't change their ITAs, this is the approximate tax penalty that any Ontario family a single $100K income pays. Why is that meaningless?"

Taxrage, you're just wrong. Let me guess, you went to some online tax calculator and typed in $50K and $100K and from that you conclude that a single-income family making $100 pays $7000 more than than a two income family with the same income. Am I right, because that's the only way you can come up with that number.

First, as a starting point, federal income splitting has no impact on either Alberta or Quebec. Alberta has a flat tax system, so income splitting has no impact, and Quebec administers its own tax system so wouldn't be affected by federal tax changes. Second, among provinces with progressive rate structures administered by the feds, your calculation doesn't actually reflect how the tax system works. Among the obvious defects, it doesn't take into account the spousal amount that can be claimed by a one income family, but not a two income family. The value of that credit varies from province to province ranging from $2200 to $3600, depending on the province. Again, read the CD Howe paper - they've done the calculations properly - http://www.cdhowe.org/pdf/Commentary_335.pdf.

[Bob Smith]

"Did you miss the recent media stories about how the provinces would pretty much have to follow the feds lead, unless they're going to amend their provincial ITAs?"

Did you miss the media stories about the provincial governments (notably Ontario) running massive deficits with no obvious simple way of balancing them? Did you miss the media stories about provincial governments (notably Ontario) increasing tax rates on high income earners. From that do you conclude that they'll just sit idly by and let the feds reduce their tax base by a few billion dollars by cutting taxes for the same people whose taxes they've been recently increasing? Yeah, right.

The Mowat instittue is right that, if the province (other than Quebec) did nothing, a federal change would affect them. But why would we assume they would do nothing? I can think of easy ways of amending Ontario (or other provincial) tax law to avoid that result:

"income" means... in respect of an individual for a taxation year the total determined under paragraph 114(a) of the Federal Act IF THAT ACT WERE READ WITHOUT REFERENCE TO SECTION[THE INCOME SPLITTING PROVISION}..."

The non-capitalized language is the current definition of income, the capitalized language is all it would take to carve income splitting out of the provincial tax regime. You think the provinces can't add 10-odd words to their tax acts?

I can think of easy ways of amending Ontario (or other provincial) tax laws to avoid that result in a manner that is wholly consistent with the Federal/Provincial intergovernmental tax collection agreements (more complicated, but still easy). And keep in mind, those agreements are not legally binding on the either level of government - note the recent termination of the Federa Government - BC comprehensive integrated tax cooperation agreement governing GST/HST. More likely, if the provinces decided not to permit income splitting under their provincial laws, the feds would likely either look the other way (if it technically violated the relevant tax collection agreement) so as not to provoke a fight with the provinces during an election year or, just as likely, agree to a technical amendment to those agreements.

In short, anyone who believes that federal income splitting will be automatically matched by the provinces is kidding themselves.

[JF]

Above there is misinformation about the practicality of Net Worth tax base descriptions. Net worth calculation is done ALL the time by people who have net worth, and a tax regime that asks one to fill out a two page description annually (subject to audit) makes it very easy to do and maintain. Bob Smith's points are intentionally misleading.

The 90 year old's scenario is a classic red-herring too. Small groups of taxpayers with well-defined issues of fairness can be simply addressed in a tax regime's design, protecting such small populations without affecting the productivity of the tax or any sense of fairness. But of course, a multi-million asset in downtown Toronto, perhaps this person can afford to borrow on their equity, regardless of their age. Even these common sense design provisions have limits, I doubt we need to protect millionaire net worth people from a 1% tax rate. And no, one would not refund money, how baldly disinformative this point is - a net worth tax regime is essentially a benefit-tax, and if the value of a person's net worth goes down because of the markets, they pay less tax (at very low rates). The public provides for civil and market order; high net worth individuals ought to be able to take care how they array their wealth in terms of marketplace effects - and isn't it great to have a currency that can be a storehouse of 'value' and the ability to trade assets easily, quickly, and safely or hold them safely. The point is that Net Worth really is 'net worth' because of the public's governments, it receives a huge benefit just doesn't pay anything to finance it.

What is the theoretical reason to tax capital lower than labour? This is presumption and theory. Plus it seems you are actually talking about income-taxation, a sleight of hand. Capital is Net Worth, not income. You could tax net worth until people decide they are better off moving (Tiebout), not anyone's suggestion, anywhere, of course, but leaving capital untaxed is inviting this net worth to be idle (if it is idle, it is just wealth then, not capital, I suppose) making the economy less dynamic in my view.

I think people read this blog, I hope they will be careful in their readership, and thoughtful (and this goes for my comments too).

Marginal dollar discussions are associated with income taxation, and you seem to see progressively lower rates for people of lesser means as somehow wrong (and note how I've flipped the exact same logic to speak of the rates as progressively 'decreasing' in order to address concepts of burden and utility). Your bias sees progressivity from the view of high income people, and it is showing. And note that defining your world to be only related to cash-flows (income taxation) is a framing that allows you to avoid the fact that high net worth individuals are under-taxed for the value they get from civil and market order - worrying about their marginal income dollar is sophistry and artistry in discussion of 'on-the-margin' theory - plainly, there is no such thing as wealth being 'on the margin.'

Pointing to "many tax experts" supporting flat taxes, is also disinformation. A benefit-tax-type as in the case of motor-fuels is a flat rate and many experts would support this design feature, and have for most of modern history because this type flat-tax makes common sense to people. But people need to know that many other tax types deserve attention to progressivity. I am of the opinion that most tax experts support progressivity as a design feature in a tax regime, the far-majority of experts and economists.

Taxation is part of all markets. Taxes are not per se "distortionary" unless you believe that all or most marketplace rules are distortionary and self-government can't set some rules (isn't this libertarianism, not tax expertise or economic expertise speak?). Taxation is uniformly applied, and the court's enforce this for affected taxpayers, and taxes are just part of your marketplace (labour doesn't see it this way though). They are NOT distortionary per se, just because. This is theoretical, a theoretical viewpoint that has its obvious bias.

Poorly designed revenue systems and poorly designed tax regimes can cause unintended effects, and these may in fact distort people's behaviors, something that should then be corrected in the statutes when a consensus agrees that the distortion isn't what they want in society. Of course some want the distortion they get - for instance, some want taxation to lay on labour and on basic cash-flow or income streams while favoring capital income and ignoring Net Worth as a tax base. I'd say that a revenue system like that is a very distortionary one, facially so to my way of thinking.

So read the comments carefully. Hopefully people can align themselves more with these JF comments than some of those above.

[Taxrage]

"Taxrage, you're just wrong. Let me guess, you went to some online tax calculator and typed in $50K and $100K and from that you conclude that a single-income family making $100 pays $7000 more than than a two income family with the same income. Am I right, because that's the only way you can come up with that number. "

Bob, are you really a tax lawyer?

Sorry, I'm not wrong. It's actually $7,339 (2014).

I created a Visual Basic macro for Excel, and used the federal brackets (cut/paste here: Array(0.15, 0.22, 0.26, 0.29), Array(0, 43954, 87908, 136271)) as well as provincial (Array(0.0505, 0.0915, 0.1116, 0.1316), Array(0, 39724, 79449, 509001)). I apply credits for personal, spousal, CPP/EI and employment, as well as the $3,500 that is exempt from CPP.

I'm a software developer. What would convince you, the actual buckets of fed/prov tax calculated in each bracket? Like, this is very simple stuff.

I don't need online calculators, which usually don't try to make assumptions regarding your credits. I'm perfectly capable of *building* them.

[Majromax]

@JF:

> The 90 year old's scenario is a classic red-herring too.

Not really. The problem of net-worth taxation is that it creates a duration mismatch.

Taxes must be paid in highly-liquid money. However, net worth is often held in long-duration and/or illiquid assets.

For long-duration assets, taxation requires continual liquidation of parts of the asset in order to avoid taking on debt. That's extremely inefficient, since it will incur multiple transaction fees. (Imagine selling 1% of a 30-year GIC yearly!) We like to imagine that taxpayers are lumps of assets and incomes and are not credit-constrained, but that's not true.

For illiquid assets, we also have the problem of speculative valuation. Ask anyone who's sold their house, and you'll find that the final sale price is not necessarily in a tight range around the property's municipally-appraised value. It's something of a LIBOR problem, in that valuations are often what people think the asset is worth rather than a standing offer to purchase.

There are ways around these problems, but they ultimately involve the taxpayer assuming some of the duration risk (or valuation risk). Imagine you could pay taxes-on-the-house in shares in the house, for example. But this is getting ridiculously complicated, which is why it makes sense to only tax gains when realized (and thus available for consumption).

> leaving capital untaxed is inviting this net worth to be idle

What does "idle net worth" mean?

Idle cash at least has a hard meaning to it: it's buried in a mattress and unavailable for circulation. But "idle net worth" doesn't have such a tidy interpretation. Passively-invested stock holdings, for example, don't mean that the company itself is doing nothing.

> high net worth individuals are under-taxed for the value they get from civil and market order

Is this necessarily true? Outside of protection of their balance sheet, it seems that the lives of the poor are most-protected by civil order. The wealthy suffer financially in disordered times, but it's the poor who starve or get caught up in violence and riots.

To me, it seems that the vast majority of the benefit of high net worth is realized when the asset is liquidated for consumption. That suggests that it is reasonable to tax that worth at that time. Aside from that, there are three-and-a-half other, more idiosyncratic benefits:

*) The first is income security. If I had a billion dollars in the bank, even without spending it right now I'd have perfect financial peace of mind. Without financially realizing that net worth, I'd be receiving a significant psychological benefit. Should this benefit be taxed?
*) The second is control. Owning equity allows me control over the corporation's activities, which is a direct benefit. Should this benefit be taxed?
*) The second-and-a-half is personal use. This is where the homeowner comes in: our retiree living in a $1mil house is receiving housing services from the property, whereas if she rented a space in a $150k house and in turn rented her $1mil house to others she'd be receiving (presumptively lesser) housing services herself and paying tax on the rental income. Is this appropriate?
*) The third is leverage. Even without selling my hypothetical $1bil, I can turn some of it to consumption by borrowing against it. The gain is still technically unrealized, but I've now consumed prior to paying taxes. Should the tax code correct for this? (Incidentally, funding taxation through consumption rather than income taxes negates this point entirely)

[Bob Smith]

You apply the credit for EI and CPP, do you take into account the tax actually paid - that's not reflected in your chart. That's not unimportant - the two income family gets more EI/CPP credit because it pays more CPP/EI (to the tune of an extra $3,500). If you're going to take the CPP/EI credit into account (fair) in computing the income tax penalty, you have to reflect that that's a real cost.

[Bob Smith]

I'd go further, as even on your numbers you beautifully illustrate France' case against income splitting.

Let's say your one-income family pays $7000 more in tax (ignoring CPP/EI and any other provincial payroll taxes that the two-income family pays but the one-income family doesn't) - who takes care of the two-income family's kids? That has a cost - not unadjacent to $7000, at least, in most parts of Canada (i.e., before and after school programs, care in the summer). Per kid. I'm not talking nannies or high-end daycares, I'm talking ordinary average daycare. Sure they get some tax recognition for that, but only partially (if you make 50k, it might be 20 cents on the dollar).

[Majromax]

@Bob Smith:
> who takes care of the two-income family's kids?

Well, now we're back to how and whether home production should be taxed.

I'm taxed if I earn wages to buy jam for my toast, but not if I grow strawberries and make jam for my toast.

From the perspective of the tax system, this is pretty trivial, especially because industrial jam has efficiencies of scale that I can't replicate at home. On the other hand, childcare is both a significant cost and (as I believe you've pointed out) not something where we see huge efficiency gains.

Double-income families and single-parent families are fairly equivalent here, since they simply don't have the time to devote to tremendous efforts of home production. Single-income, double-parent families are the odd situation out.

[Taxrage]

"You apply the credit for EI and CPP, do you take into account the tax actually paid - that's not reflected in your chart. "

My algorithm works as follows: Family has spouse A and spouse B. One or both work. Aggregate family income is apportioned either 100% to A or 50% to each (if both work then, I calculate A's taxes on his/her 50% then, at the end, I just double the tax for A). For A, calculate all fed/prov taxes in each bracket, including ON surtaxes (admittedly, I didn't calculate ON health levies - note to self). Subtract all (non-refundable) personal credits for A (personal, employment, CPP/EI), as well as fed/prov spousal credits if only earner. You had me worried that maybe I doubled the spousal credit, but it's only added in the 1-earner case, hence would not be doubled.

So, the only inaccuracies should be due to my having left out the FSHL, but both 1 and 2-earners pay this, perhaps to slightly different degrees. Perhaps I'll add this tweak over the weekend.

[Taxrage]

"I'd go further, as even on your numbers you beautifully illustrate France' case against income splitting.".

CPP and EI aren't taxes though. CPP builds a pension (which 1-earner loses out on) and EI is employment insurance, which provides income in the case of job loss. One could view these as just other forms of work-related expenses. You pay them only if you have employment income.

At the end of the day, if the government is going to help families with $400K incomes to pay their daycare expenses (or provide subsidized spaces), then they have to put something on the table for families that look after their own kids. Simple as that.

Income splitting is one way to put something on the table that helps achieve this, but it's the proverbial bird in the hand.

[Bob Smith]

"Well, now we're back to how and whether home production should be taxed."

Actually, we're back to how we measure income. If I have to spend $14,000 in child care to earn $100,000, is my income $100,000 or $86,000? If child care expenses are cost of earning income - clearly they are - the real income of a two-income family with salaries of %50K is less than than a one-income family with a salary of $100K.

"You pay them only if you have employment income". So if you split employment income, should your spouse have to pay CPP adn EI (and qualify for benefits) - that would make sense for CPP (since it would force spouses to save for their own retirement, rather than relying on survivor benefits for which they haven't paid), and would probably make income splitting more palatable.

"At the end of the day, if the government is going to help families with $400K incomes to pay their daycare expenses (or provide subsidized spaces), then they have to put something on the table for families that look after their own kids."

Which might be a compelling argument, if the Conservative government was proposing to provide subsidized child care or if the credit for child care (for most Canadian) was particularly large (not sure why you fixate on the treatment of people who account for less than 1% of the taxpaying public). As neither of those propositions are true...

More to the point, if the conservative governments is trying to help families with children, why on earth would they pursue a policy where most of the benefit accrues to families who need it the least. Apart from being difficult to justify on policy grounds, politically it's suicidal.

[Taxrage]

"More to the point, if the conservative governments is trying to help families with children, why on earth would they pursue a policy where most of the benefit accrues to families who need it the least."

I think we can say that (gov't trying to help families with children) w.r.t. UCCB and other benefit payments.

Can't say that w.r.t. daycare subsidies and deductions, since they don't apply to 30% of families with young children who look after them by themselves. We could say that w.r.t. helping families put a 2nd (or in the case of a single parents, the 1st) spouse back to work. Where is the objective for daycare stated anyway?

Why do we (to use my 1% example) help a $400K income family put a 2nd spouse back to work? Makes sense, I guess, if the only objective is to get the 2nd spouse back to work. Just don't dress it up as it being for the kids.

Why do we let a retired bank (or Nortel) president save $30K through pension splitting? Again, can you provide a reference/link where I can look it up?

In the case of a the free $2,100 ghost spouse credit we provide a single parent, I can see that the government wants to tax the single parent the same as a 1-earner couple family, even in the case of a $400K income.

We could perhaps say that parental leave is to benefit the child, but this is an expensive (52-week) benefit.

There are probably a few goodies that I left out, but you see a pattern, which is there is a targetted benefit for just about every type of family with young children except the type that looks after their own kids.

If it's really for the kids, then why is the only thing thing that accrues exclusively to this group, is a $7,300 tax penalty? Is that all this group deserves from the tax system? Currently, the answer is apparently YES.

[Bob Smith]

So, as expected, the Tories gutted their proposal to allow income splittting by capping the maximum savings at $2K. Also as expected,they structured it as a credit, so that it doesn't affect provincial revenues. They've increased the UCCB and extended a limited UCCB to children over 6 and increased the maximum amounts that can be claimed for child care expenses. A little something for everyone. The UCCB change is the big dollar amount, and that'l be hard for either opposition party to campaign against.

As an aside, I'll be curious to see the forms you need to fill out to benefit from the income splitting - one of my colleagues joked that it will cost more than $2000 in accounting fres to claim it.

[Taxrage]

Yeah, it's definitely going to have to be creative...and in place for the 2014 return!

I looked at the ways & means legislation (4-5 pages). It basically requires filers to do some sideline calculations (like RRSP room) to determine the hypothetical federal savings that would result from an income transfer.

I didn't quite understand how the spousal credit worked. Didn't have a dark room and a single light bulb to study that one.

[Pithlord]

I am not necessarily in favour of income splitting, but I think both these arguments are wrong.

Efficiency

The error in the argument is that while the market labour supply of lower-income spouses may be more elastic to taxation, the market labour of the higher-income spouse is more productive, and the non-market labour of the lower-income spouse has to be accounted for. The analysis doesn't do this. If the higher income spouse works more in the market, and the lower income spouse works more at home, then each is probably pursuing their comparative advantage.

Equity

Equity is pretty hard to define, but it whatever it is, it can't just be overall higher marginal taxes. Rawls' suggestion is that equity is improved by a rule that parties behind the veil of ignorance about whether they would be the beneficiaries would pick. On that basis, the question is what rule you would pick if you didn't know whether you would be single, in a couple with relatively similar incomes, or in a couple with relatively divergent incomes. I don't find it easy to answer that question. Rawls assumed people in the original position are risk averse, and people can end up in one-major-earner families for reasons outside their control, so they might well want at least a compromise that gives some room for splitting.