Taxation of the family: everything old is new again

Matt Krzepkowski and Jack Mintz have recently produced a working paper titled "No More Second-Class Taxpayers: How Income Splitting Can Bring Fairness to Canada's Single-Income Families."  The paper argues that higher income single-earner married couples are "disadvantaged by the current system." It proposes to put an end to that by allowing income splitting, so a one earner couple with income of $200,000 a year would pay about the same taxes as a two-earner couple where each person earned $100,000 per year.

The Krzepkowski and Mintz paper is the latest salvo in a long war over the best tax treatment of the family. Alexandre Laurin and Rhys Kesselman produced an analysis of income splitting for the CD Howe Institute in 2011, and back in 2000, Carole Vincent and I wrote a paper on Taxing Canadian Families for IRPP (downloadable here: Vincent_Woolley).

While cleaning out the the bottom drawer of my filing cabinet I came across a number of other, older studies of the taxation of families in Canada. As most are in the public domain, but difficult or impossible to obtain elsewhere, I thought I'd post them here. What is remarkable about them is the degree to which they presage the current policy debates - most of the ideas in the Krzepkowski and Mintz paper were brought forward and discussed in the 1960s and 70s. Understanding why they were rejected then may help clarify the current policy debate.

1. The Carter Commission (1966): Download relevant chapters here (or, for full report, click here)

In 1966, the Carter Report recommended that Canada adopt "joint taxation". Married couples would file a single joint return, and be treated as a single tax unit. According to Carter, Canada's individually-based tax system - the system we still have today - leads to "ludicrous results". Their example: in 1965, a family where each spouse had an income of $4,000 per year paid $200 less in taxes than a family where one spouse had an income of $8,000 per year, and the other earned nothing. "We can see no justification whatsoever for this difference," they wrote, "particularly if all the income is from property and neither spouse is employed."

Joint taxation can be beneficial or disadvantageous for married couples, depending upon the rate at which the joint income is taxed. The rates suggested by the Carter Commission would have allowed married couples, in effect, to split their incomes, producing considerable tax savings for higher income single-earner couples, while causing taxes to increase for couples in which the spouses had equal earnings. Crucially, the Carter Commission would lower single-earner married couples' taxes to US, or lower than US, levels:

I suspect the Carter Commission would have found Canadian tax treatment of married couples a lot less "ludicrous" if it had been in line with the tax treatment of families in the US. But it wasn't, and still isn't. The US acts as a benchmark; a reference point for Canadian tax policy. The existence of joint taxation and income splitting there means that pressure to adopt such policies here will never go away. 

2. Douglas Hartle (1971) "Taxation and Married Women in Canada" Report Prepared for the Royal Commission on the Status of Women. Ottawa: Information Canada.  Download Hartle background paper here; access Royal Commission on the Status of Women (Bird Commission) here.

Doug Hartle was the Research Director for the Carter Commission, so it is not surprising that the Royal Commission on the Status of Women turned to him when they needed a background report on the tax treatment of married women in Canada.

What really struck me, re-reading Hartle's report, is the remarkable similarities between the approach he advocated, and the Krzepkowski-Mintz proposal. Both advocate some kind of joint taxation/income splitting. Both suggest taxing what Hartle called the "imputed income" of housewives, or what Krezpkowski and Mintz call "implicit income." Both advocate limiting the transferability of the basic personal exemption, in order to mitigate some of the negative impacts of joint taxation on married women's incentives to participate in the labour force. 

Today, many feminists oppose joint taxation on the grounds that, when spouses' incomes are aggregated, the secondary earner - usually the woman - effectively faces the primary earner's marginal tax rate, which discourages female labour force participation. Given this, I find it somewhat surprising that the Bird Commission accepted the Hartle/Carter joint taxation recommendation (p. 304 of the report). But perhaps it isn't so strange - Canada's tax policy community is fairly small and tight knit. Today there are people like Elisabeth Gugl, Lisa Philipps or Kim Brooks who can provide an alternative view on taxation. The Bird Commission would have had no comparable experts to call on. Yet even at the time, commissioner Elsie Gregory MacGill strongly disagreed with the joint taxation proposal. Later on, chairman Florence Bird retracted her support for joint taxation.

Supporters of the Krzepkowski and Mintz idea of taxing implicit income might be interested to learn that the Bird Commission rejected Hartle's suggestion of taxing imputed income, "Although the concept of imputed income is theoretically sound, we feel that its implementation would pose great difficulties." Reading between the lines, I suspect that the (then) Department of Revenue said it was a no go - alternatively, it might have been that Hartle's proposal was simply not understood.

The Bird Commission made a number of recommendations with regard to taxation policy, including reducing the spousal amount exemption, and converting deductions, including the child care expense deduction, to refundable credits. In the end, however, those were largely ignored.

3. Louise Dulude (1976) "Background study on women and the personal income tax system" Ottawa: Canadian Advisory Council on the Status of Women  Download here 

For a popular summary of Dulude's argument see: "The Person Papers: Taxation Untangled" Canadian Advisory Council on the Status of Women, 1976 (Download CACSW_taxation_untangled). For the Advisory Council on the Status of Women's Recommendations (based on Dulude's Report) see here (Download CACSW_1978)

Louise Dulude's background study is one of the first feminist analyses of taxation in Canada. Dulude was a lawyer and social activist, and she brings this perspective to her research.

Dulude's paper made a number of recommendations which have subsequently been adopted: that husbands and wives should be allowed to employ each other in unincorporated businesses or farms, or to enter partnership arrangements; that the (then) child exemption should be replaced by a refundable child benefit; that money received as child support should not included in the recipient's taxable income.

What I find particularly interesting about Dulude's work, though, is her focus on property rights, ownership, and control. Joint taxation or income splitting proposals, such as the Krzepkowski-Mintz one, tax married couples as if all income and assets were shared jointly by the spouses. Yet they do not require any actual sharing of income, or change marital property regimes. As Dulude puts it: "These systems do not encourage legislative changes towards more sharing by the spouses, they assume that spouses already share everything" (p. 31). She was talking about the Carter/Hartle joint taxation proposal, but the point is still relevant today. 

Dulude is also critical of the idea of taxing imputed or implicit income.

Hartle states and repeats that spouses save money because of the wife's provision of housekeeping services. In fact, it is the husband who saves, not the housewife who more than likely owns nothing. To require the housewife, who has only a right to basic support, to pay tax on an arbitrarily-determined imputed income, would be adding insult to injury.

This, I think, rests on a misunderstanding of the Hartle proposal. Under Hartle's proposal, as under Krzepkowski and Mintz's, the person generating implicit/imputed income does not pay any taxes. Under Hartle's proposal, each person who is not in paid employment is assumed to have imputed income of $2,000. He or she is also assumed to have basic needs of $2,000, hence receives a $2,000 exemption. So a person with no employment income has $2,000 of imputed income and a $2,000 exemption - hence pays no tax.

Hartle then makes the crucial assumption that earnings reduce imputed income dollar for dollar.  A person with employment income of $1000 has $1000 in earnings, $1000 in imputed income, and a $2,000 exemption - so pays no tax. Only when a person has employment income of more than $2,000, and no imputed income, will he or she pay start having positive tax liabilities. For both Hartle and Krzepkowski-Mintz, the whole imputed/implicit income is dealt with by eliminating the married amount tax credit, and replacing it with an individual earnings exemption.

Overall, however, Dulude's paper is well worth reading for the insights it gives into the tax treatment of Canadian families.

4. Maureen Maloney (1987) Women and Income Tax Reform. Ottawa: Canadian Advisory Council on the Status of Women. (Download Maloney_women_income_tax) and CACSW Brief on Tax Reform Download Brief_on_tax_reform 

Maloney's paper reiterates a number of positions taken in earlier CACSW papers - the need for changes in the tax treatment of child support, the need to reform the child care expense deduction to make it more generally available, and more attractive to low income earners. She also takes a strong position against the goods and services tax, arguing that it is regressive, and income taxes are a more progressive way of raising revenue.

One thing that is interesting about Maloney's paper is that it conveys a sense of how tenuous Canada's individually-based tax system is. She points out how jointness has gradually crept into the income tax system, through transferable deductions and credits, and provisions such as the spousal amount tax credit (then the married exemption), and is critical of this trend. She recommends that, "At the very least, individual taxation should be reaffirmed as the appropriate unit of taxation." 

Twenty-five years later, the prospect of such a reaffirmation seems slimmer than ever.

Matt Krzepkowski responds in the comments below on July 1st.

Comments

[Nick Rowe]

Frances: some years ago you said something about taxation and the family that made it all suddenly clear to me. But now I have forgotten it! It was something about an impossible trinity, or something. You could have progressivity of marginal tax rates, or X, or Y, any 2 out of the 3 (or was it 1 out of 2??) My brain has gone. Do you remember?

[Frances Woolley]

Nick, it was:

1. neutrality with respect to marriage, so tax liabilities don't increase or decrease upon marriage
2. two-earner and one-earner couples pay the same taxes with the same total household income and
3. a progressive tax system.

The Krzepkowski-Mintz proposal more or less does (2) and (3), but at the expense of (1) - higher-income couples with very unequal earnings would see a big decline in their tax liabilities upon marriage. Interestingly, the one key difference between Hartle/Carter and Krzepkowsk-Mintz is that the earlier proposals included a marriage tax - it seemed obvious to people writing then that there were economies of scale associated with marriage, and a married couple earning $100,000 could afford to pay a higher percentage of their income in tax than two singles earning $50,000 each.

That idea seems to have been lost completely, which is interesting - I'm trying to figure out why, but can only speculate. Because the US system has changed to get rid of its marriage penalty, and because this is all benchmarking against the US? Because having a stay-at-home spouse is less of a winning proposition for the person with employment income than was once the case? Think about it- what the Krzepkowski-Mintz proposal is saying is "as soon as I get married, my ability to pay taxes plummets, because I'm having to support this other person."

[Matthew]

The issue with individual taxation is that it means that the state is not neutral with respect to how couples decide to allocate their combined labor. This may be a good thing or a bad thing--for example, if you think that having a stay-at-home parent is good for child outcomes, then filing separately is bad, since it penalizes this. But if, on the other hand, you think that the children will be just as fine, and that it is good to have both men and women working, then it is good to require separate filings. If you really don't care either way about household allocation of labor, but do want to decrease the number of unemployed tax accountants, then the best system is to allow either joint or separate filing, as the US federal government does.

It is worth noting, though, that the US actually has several different systems. At the federal level, we have the dual system that recognizes individual incomes and property separately, but allows either joint or separate tax filing. But many states, like California, actually do not recognize individual incomes or property, and instead treats both members of a married couple as if they earned all income and own all assets equally, making it closer to the Krezpkowski-Mintz proposal. The Federal government, for its part, also follows the Krezpkowski-Mintz-like system, but only for states that have them (which leads to weird things: a married gay couple in New York, which does not have communal property laws, has absolutely no federal recognition, while the same couple, if they lived in california, would be taxed by the federal government as if the federal government actually recognized their marriage--go figure.)

[Andrew F]

Am I missing something? I have never understand how fairness dictates that "2. two-earner and one-earner couples pay the same taxes with the same total household income". Isn't there substantial non-monetary benefits to having a one-earner couple due to increased home production? And how then do we address fairness for one-earner couples vs one-earner singles who would thus pay more tax with income splitting?

[Bob Smith]

"Isn't there substantial non-monetary benefits to having a one-earner couple due to increased home production?"

Perhaps, perhaps not. The home production of two retired people who earn passive investment income is likely to be the same (or more) than the one income couple that earns wage income. Similarly, the same may well be true of a two income couple where one person works (in the market) part-time (a not uncommon scenario). Moreover, if the intention is to tax the value of home production (which our tax system doesn't currently purport to do), it isn't clear why it should be correlated with the income of the sole income earner.

This is one of those areas in tax policy where I think there are reasonable arguments on all sides.

[Matthew]

Update: The Supreme Court weighed in shortly after my comment, partially invalidating it. It is now the case that all gay couples married in any state or country are recognized by the US federal government, including for tax purposes.

[Bob Smith]

It's an interesting SCOTUS decision. Ironically, it's also a case where deference to states rights ensures the protection of vulnerable minorities - a reversal of conventional thinking in US Civil rights cases.

[Matthew]

Did the court defer to states rights? On DOMA, it is quite likely that this precedent will be used to force states to recognize gay marriages in other states. On Prop 8, court has essentially rendered bans on gay marriage unenforceable throughout the judiciary. My interpretation is that these rulings were more of a warning signal to the states: if they don't restructure their laws, a much more forceful ruling on gay marriage is only about 2 to 3 years away. That's how the court has historically operated on civil rights matters.

[Majromax]

> Interestingly, the one key difference between Hartle/Carter and Krzepkowski-Mintz is that the earlier proposals included a marriage tax - it seemed obvious to people writing then that there were economies of scale associated with marriage, and a married couple earning $100,000 could afford to pay a higher percentage of their income in tax than two singles earning $50,000 each.

I don't think that this is a taxable distinction. In previous decades, it was generally assumed that a couple would be married, and not doing so carried a significant social penalty. But with single parentage becoming socially accepted and common-law relationships becoming more of a default (especially in Quebec), marriage is now much more of an at-will status.

Implementing a "marriage penalty" in the current tax code would effectively have to capture common-law relationships as well, and those are much more easily hidden. To the extent that marriage is a social good that should be encouraged at the societal level, implementing a marriage subsidy in the tax system seems like a decent tax expenditure.

[Determinant]

This is one of those areas in tax policy where I think there are reasonable arguments on all sides.

I agree with Bob (for once, yikes!)

My sister-in-law is a stay-at-home mom, my brother earns enough to make it work and she likes childcare. I can't complain, I got a niece out of the deal with another one on the way (gender not yet known).

I also agree with Frances' Impossible Trinity, which I think encapsulates the fact that there is only so much the state can do to minimize the burden for a given level of taxation.

[Matthew]

So can part of US income inequality be explained by the impossible trinity? For the most part, any two middle class individuals who get married are going to see little difference in their tax liability. But if you have a $50,000 person marry, say, Mitt Romney, then allowing Romneys to file jointly is basically like giving Romney 50% off his taxes, because 50,000 is nothing compared to 350 million.

[Frances Woolley]

Majromax "to the extent that marriage is a social good"

Why? What is the argument for regarding marriage as a social good? The only argument I can see is that marriage is good for kids, in which case any kind of marriage subsidies should be targeted towards families with children.

Yet if marriage truly is a social good, why do we have large implicit marriage taxes at low income levels (e.g. if a single parent earning $20,000 a year moves in with someone earning $40,000, they will experience a substantial loss of Canada child tax benefit and other benefits)? If we're going to start encouraging marriage, let's start there, where relationships are more vulnerable, not at the $100K, $200K, $300K single earner households.

[Bob Smith]

"Why? What is the argument for regarding marriage as a social good? The only argument I can see is that marriage is good for kids, in which case any kind of marriage subsidies should be targeted towards families with children."

Beyond that, I think there is a case for marriage as a social good on two grounds. One, I think there is literature which suggests that married men are better behaved than their single peers (in terms of criminality, violence, anti-social behaviour generally). Granted, I don't recall whether those studies weeded out the impact of children or addressed the obvious causation/correlation problem (maybe single men are single for a reason - i.e., they're generally more likely to be anti-social), but as a hypothesis it seems plausible that marriage may have a civilizing impact on the males of the species.

Second, and I think this is one of the conservative arguments (and there are many) in favour of gay marriage (appropriately, given today's judgements), is that the institution of marriage is an institutional arrangement which provides a degree of income support, insurance, etc. without the need for state assistance. Again, I think there is a fairly robust literature to suggest that married couples generally suffer lower incidences of poverty than their single friends and colleagues. Obviously that takes on greater concern when children are involved, but should be of concern regardless.

Granted, that doesn't address your point that there may be better ways to promote marriage than permitting income splitting - fair point - but it does open the door to different types of discussions.

Mathew: "On DOMA, it is quite likely that this precedent will be used to force states to recognize gay marriages in other states. On Prop 8, court has essentially rendered bans on gay marriage unenforceable throughout the judiciary.

Not to hijack Frances' thread, but... on DOMA, the court basically left the definition of marriage up to the states and said that the US federal government couldn't pick and choose which state marriages it would recognize, marriage being an area of state law. On Prop 8, it was a purely procedural decision, the majority held that if the state government wasn't going to appeal the lower court decision (and California didn't), the appellants (who were merely intervenors) didn't have standing to do so (which is probably right, which is why, ironically, several of the conservative= judges supported that position), which means they didn't touch the substance of Prob 8. I have my doubts as to whether Prob 8 could have survived a substantive challenge (particularly in light of the DOMA decision), but in event the issue still remains outstanding.

[Bob Smith]

Mathew: "For the most part, any two middle class individuals who get married are going to see little difference in their tax liability. But if you have a $50,000 person marry, say, Mitt Romney, then allowing Romneys to file jointly is basically like giving Romney 50% off his taxes, because 50,000 is nothing compared to 350 million"

But keep in mind, for the Mitt Romney example, the benefits of income splitting disappear if both spouses have incomes in the top income bracket. I can't remember what that is for the US, but if Mitt Romney could split his income with his wife, she'll likely pay the same marginal tax rate on his share as he does, so it likely woulnd't make much of a difference in their total tax bill.

For example, in Ontario, the top marginal tax rate kicks in at ~$130,000. So, assuming one spouse has no income (and ignoring other tax benefits), the benefits of income splitting max out at a family income of $260,000. The benefits are the same for the family with an income of $260,000 as they are for a family with an income of $26,000,000 (who probably wouldn't even notice them). I think the political appeal of income-splitting is that the people who really benefit from it are the people who are decidedly middle-class and whose top marginal tax rates (think people in the $50,000-$260,000) income range would fall with income splitting (i.e., not gazillionaires).

Also, keep in mind that we already permit income splitting for some forms of income (through spousal RRSPs, pension splitting, etc.) and sophisticated taxpayers with investment income (or who carry on business through private corporations) can dividend split between spouses, children, family trusts, etc., despite an abundance of anti-avoidance rules to discourage it (for example, by issuing different classes of shares of the family business to your spouse or adult children to allow the business to sprinkle out dividends amongst your family members). So, in that respect, permitting income splitting really only levels the playing field between people who earn employment income and people who earn business or investment income.

[Majromax]

Re: Bob Smith
> Not to hijack Frances' thread, but... on DOMA, the court basically left the definition of marriage up to the states and said that the US federal government couldn't pick and choose which state marriages it would recognize, marriage being an area of state law.

Right. The decision did not touch on the "full faith and credit" section of DOMA, which permits states to not recognize same-sex marriages performed in other jurisdictions. The reasoning used in this decision was a combination of "discrimination is bad" and "states rights." The former would argue that singling out SSM as grounds for state-level discrimination is impermissible, while the latter would permit the arbitrary-but-optional exercise of state-level powers. Expect further legal challenges.

Re: Frances Woolley
> Yet if marriage truly is a social good, why do we have large implicit marriage taxes at low income levels (e.g. if a single parent earning $20,000 a year moves in with someone earning $40,000, they will experience a substantial loss of Canada child tax benefit and other benefits)? If we're going to start encouraging marriage, let's start there, where relationships are more vulnerable, not at the $100K, $200K, $300K single earner households.

I think you could say the same about many other areas of the tax code, especially as they begin relating to social benefits. The nominally progressive tax system looks a lot flatter (and even perhaps a bit regressive) at low income levels when you start to look at benefit clawbacks. It's not optimal in any way, but touching it is a matter of social policy rather than strict economics and the political optics can look terrible.

[Frances Woolley]

Bob: "I think the political appeal of income-splitting is that the people who really benefit from it are the people who are decidedly middle-class and whose top marginal tax rates (think people in the $50,000-$260,000) income range would fall with income splitting (i.e., not gazillionaires)."

First of all, people in the $50,000 income range will hardly benefit at all - their benefits in absolute terms of as % of income are much much less than those in higher income ranges.

Second, since when was an income of $260,000 decidedly middle class? That's right up in the top few % of the individual income distribution.

With respect to investment income: i think there is a good case for easing the attribution rules. But that's what we should be doing. If people want to transfer assets to their spouse, that's fine, but to transfer the tax liabilities without transferring over control of the asset reveals that, no, people don't really want to be an economic unit when and where it counts.

On levelling the playing field between employment and self-employment/business income - yes, I think that this is part of what's really going on here. And there is a big disparity between the tax-reduction opportunities available for employment income and those available for self-employment income.

[Bob Smith]

"First of all, people in the $50,000 income range will hardly benefit at all - their benefits in absolute terms of as % of income are much much less than those in higher income ranges"

You're right, in one sense, but it look at it from the perspective of a single-income family making $50,000K a year. Based on the CD Howe's estimate of the impact of the federal conservative party's income splitting proposal, they would save ~600 bucks a year - a good chunk of their federal tax bill. Significance depends on the perception of the recipient.

" Second, since when was an income of $260,000 decidedly middle class? That's right up in the top few % of the individual income distribution."

And if you ask the doctors, lawyers, dentists, etc. making $260,000 a year, they'll identify themselves as being thoroughly middle-class. I doubt the people protesting against the 1% had in mind their dentist.

"With respect to investment income: i think there is a good case for easing the attribution rules. But that's what we should be doing. If people want to transfer assets to their spouse, that's fine, but to transfer the tax liabilities without transferring over control of the asset reveals that, no, people don't really want to be an economic unit when and where it counts."

I wouldn't be so sure about that. First, because transferring legal control doesn't mean transfering defacto control. (And, as aside, not want to give up control, doesn't imply a rejection of the notion of the family as an economic unit, it implies a desire to control the economic unit, which may or may not be a good thing). Second, our family law has evolved in such a way that, in the event of a break-up, your spouse is going to get half of everything, regardless of legal title anyhow, so transferring legal title doesn't really mean much. In the absence of those attribution rules, you'd see a heck of a lot more income splitting.

Moreover, I think there's a good case to me be made that the same reasoning that informs the evolution of our family law (that marriage is a partnership, the assets of which belong to both parties, regardless of legal ownership) should also inform our tax law. Ironically, there's a disconnect between the feminist analysis of family law (which has largely pushed the evolution of our family law structure) and the feminist analysis of tax law (which seems to reject the partnership analogy and adopts a strict focus on legal ownership). It's certainly arguable that the former approach is the better one. Certainly, I suspect it better reflects the reality of most family relationships.

[Frances Woolley]

Bob - actually, the Krzepkowski-Mintz proposal to tax imputed income would wipe out the gains from income splitting for someone at $50K - the elimination of the spousal amount (the way that imputed income is being taxed) would increase taxes owing by about $2500, depending upon the actual provincial income tax rate. I need to look at this aspect of it more closely, but if that proposal actually increases taxes for a single-earner family at $50K, that would be totally unacceptable as far as I'm concerned.

" In the absence of those attribution rules, you'd see a heck of a lot more income splitting." I don't know about that. People were allowed to split CPP for a long time but the take-up rate was always very low. Yes and no on your point re the division of assets in event of a divorce - inheritance, assets acquired prior to marriage, etc.

Your point re the disconnect between feminist analysis of family law and feminist analysis of tax law is an interesting one.

On people seeing themselves as middle class when they are in the top 1 or 2 or 5 percent - if you remember I had a post on that a little while ago.

[Patrick]

"And there is a big disparity between the tax-reduction opportunities available for employment income and those available for self-employment income."

Depends on what the self-employed person is doing, no? If you're sitting at a desk at home isn't it pretty minimal? I think you'd have to be incurring significant expenses, and if you're not generating a ton of income, you're not really ahead.

Someone (who made good money doing contract work from home) once told me it was better to set-up a corporation, make yourself an employee and pay yourself as small a wage as you could get by with, and otherwise keep money in the firm which you can dole out to yourself by declaring dividends. Or something like that. I dunno. Seemed like a lot of work to me, and after paying the accountants I'm not really sure he was ahead (he wasn't earning THAT much). But I suppose he liked the game.

I've often wondered if it could be possible to set-up a marriage like a corporation and game the tax system that way. I suppose that's why child care expenses can't be claimed if the care giver is a relative.

[Bob Smith]

"I need to look at this aspect of it more closely, but if that proposal actually increases taxes for a single-earner family at $50K, that would be totally unacceptable as far as I'm concerned."

I think you're right that, politically, any realistic income splitting proposal has to see the taxes of the single-earner $50,000 family fall, otherwise it's a non-starter (which consideration, no doubt, shaped the federal conservative's proposal, and may doom the Mintz proposal - I haven't done the math on it).

[Andrew F]

I'm really not overly concerned with easing the tax burden of those single-earners with high incomes. Why is this a policy priority?

[Bob Smith]

"I'm really not overly concerned with easing the tax burden of those single-earners with high incomes. Why is this a policy priority?"

Two reasons. One is a cynical one - a good chunk of the conservative party base is concerned about relieving the tax burden on single-earner families (maybe cynical isn't the right word, partisan). The other is a principled one - horizontal equity demands that similarly situated people be taxed equally. If the tax burden on a one-income family is greater than that on a two-income family with the same income, that doesn't happen.

[Declan]

I hope they implement this so I can publish my paper, "No More Second-Class Taxpayers: How Ending Income Splitting Can Bring Fairness to Canada's Single Individuals."

[Frances Woolley]

Declan - that's the comparison that's never made in this debate, isn't it! And no one ever mentions the Carter Commission on this either: it's obvious that two can live more cheaply than one, that there are economies of scale from living together, and a married couple with income of $100,000 should pay a higher rate of tax than a single with $50,000.

[Bob Smith]

"And no one ever mentions the Carter Commission on this either: it's obvious that two can live more cheaply than one, that there are economies of scale from living together, and a married couple with income of $100,000 should pay a higher rate of tax than a single with $50,000."

Which is a fair observation, but which doesn't explain why a two income couple should pay lower taxes than a one income couple (since presumably they both enjoy the same economies of scale). While that issue is related (the solution to address would involve tinkering with personal exemptions for spouses - raising similar issues to tinkering with income splitting), its distinct.

Mind you, strictly speaking, single people can address the economies of scale issue by living with other people (i.e., roommates) so it isn't stricly linked to marital status and its an issue that people can resolve without monkeying with the tax law (although at the cost of sacrificing your privacy - hey, welcome to married life).

[whitfit]

Bob Smith: "Which is a fair observation, but which doesn't explain why a two income couple should pay lower taxes than a one income couple (since presumably they both enjoy the same economies of scale). While that issue is related (the solution to address would involve tinkering with personal exemptions for spouses - raising similar issues to tinkering with income splitting), its distinct."

I think in order for fairness to dictate that a one income couple is taxed the same as a two income couple, the two income couple should be able to get tax deductions for the value of house work that a full time stay at home spouse provides. Perhaps a capped deduction of professional house work, take out meals and laundry services.

Alternatively, why stop at a couple being able to file as one economic unit? Why not a household, or extended family?

[Bob Smith]

Whitfit: "The two income couple should be able to get tax deductions for the value of house work that a full time stay at home spouse provides. Perhaps a capped deduction of professional house work, take out meals and laundry services."

But you're assuming that two-income familes don't do homework (I can't speak for you, but my wife and I do our own laundry, cook are own meals (well, mostly), etc.. I suspect there are few two-income families that don't engage in significant home production). Moreover, as I noted earlier, having two incomes, doesn't mean that both spouses don't engage in home production. I suspect many retired couples are two income familes and both partners engage in home production (since their income isn't derived from labour). Similarly, it isn't uncommon for two income families to have one spouse (or both) who works part-time (in the market) and works at home the rest of the time. Why should the tax rate vary between two otherwise identical families because in one, one spouse works full time and earns $50,000, while in the other both spouses work part-time and earn $25,000 each? Can you credibly claim, in that instance, that the difference can be justified by differences in the value of imputed income from home production?

Moreover, the assumption is that income splitting only favours one-income families. But of course, it would also favour the two income family where one spouse earns more than the other. Turn your argument around, what's the fairness reationale for taxing two otherwise identical families earning $100,000 with both spouses working full time at different rates because in one case the income split is 50/50 (say, two teachers), while the other its 80/20 (accountant and artist)? Surely that distinction can't be founded on a claim that the home production of the one family is greater than that of the other.

As for why to use familes, rather than households or extended family as a basic unit of taxation. First, our society (and our laws) recognize the importance of families in a way they don't recognize, say, roommate relationships. Second, our laws and society recognize families as a core economic unit and recognize (and encourage) the economic interdependance between family members. Indeed, we impose legal support obligations amongst family members, in a way that they don't between, say, roommates or extended family members (and short of legal obligations, you know that the father who doesn't support his children is generally viewed as a POS). You can be legally compelled to provide economic support for your children and/or spouse, in a way that you can't vis-a-vis your roomate or your uncle. Apart from that, our society imposes similar moral and social obligations (or at least, expectations) amongst family members.

[Seamus Hogan]

Frances: The impossible trefector.

"1. neutrality with respect to marriage, so tax liabilities don't increase or decrease upon marriage
2. two-earner and one-earner couples pay the same taxes with the same total household income and
3. a progressive tax system."

Can't you have a common marginal rate of taxation across all individuals plus a demogrant to all individuals (giving negative tax at low incomes)? The common marginal tax rate ensures 1 and 2, and the intercept shift of 3. gives progressivity.

[Matt Krzepkowski]

Great piece Frances. It's important we don't forget the work of past commissioned work on this.

A quick point about our proposal to clear up some confusion in the comments: Elimination of the spousal amount was only recommended for couples electing to split income. At the worst, couples would pay the same taxes they do now. Drastically changing the distribution of tax payments wasn't on the table. This targets the proposal a little more towards dual-earner couples that unequal incomes, where I think there is a more solid argument in favour of similar taxation, rather than single-earner couples.

I'd also like to note that secondary earners are usually assumed to make a discrete decision to either work or not (or make a 0/20/40 hour a week decision), which makes the average tax rate the relevant tax rate, not the marginal tax rate. As our proposal increases the marginal tax rate up to the primary earner's level (bad) but increases their personal deduction (good), the average tax rate of a secondary earner actually decreases for a large number of individuals.

Some other points on the topic in general and through the comments:

1. Our way of looking at joint income wasn't so much the assumption that individuals share all of their income, the question we asked ourselves was whether the tax system should take this into account or not. In the end our answer was that it shouldn't, or can't. That a couple *could* share their incomes between each other is enough that these couples should be comparable for a tax system in our opinion. Transfers that are means-tested on family income try to go somewhere between the two extremes but I'm not sure they currently do so in the best way. That said, I think its important the tax system doesn't increase income inequality within a household. A higher income spouse dictatorially shifting income to their spouse, saving themselves taxes and forcing their spouse to write a large check from their personal bank account, would certainly be problematic. Any implementation could, and should, make sure issues like this don't occur.

2. I think one of the reasons including a measure of implicit income was not done, and may not be done in the future, is the difficulty of measurement. The empirical work from the US Bureau of labour suggested that unpaid household income is constant across the income distribution of households, and our paper estimated what this value may be by updating some older work from Statistics Canada. There is no way we picked the 'correct' value, but do not know if we picked a value that is too low or too high. Misestimating this value will effect whether the tax system favours single- or dual- earner couples

3. The individual/couple taxation comparison is an important question as well. Using an income-equivalent scale of root 2 (a couple only needs 1.414 times the income to maintain the same standard of living as a single individual, which is the value used by StatCan for some low-income measures), a dual-earner couple's average tax rate is always under an individual's. But the average tax rate of a single-earner couple is greater than an individual's above around $57000 (couple's joint income). However, as far as I can tell, trying to match individuals and couples in terms of some equivalence scale less than 2 will always cause additional distortions in marriage decisions.

[Franceswoolley]

Matt - thanks so much for taking the time to respond, and for those extremely thoughtful comments.

That average v. marginal tax rate argument is an interesting one. How much push-back are you getting on that? Personally, I would not want to stand up in front of a gathering of public finance economists and say "marginal tax rates are less important than average ones." I'd anticipate a fair bit of flack.

The thing that I'm still struggling to figure out is how you picture unwaged or (relatively) low-earning spouses. Are they like my mother, who still grows enough vegetables to feed the family at least 8 months a year, cooks most meals at home, paints and does home repairs, and generally produces goods and services that would cost several tens of thousands a year to replace? If so, equity between one and two-earner couples would require a far higher tax on implicit income than is contained in the current proposal.

Does one spouse have a relatively low-income because they're low-waged? I don't know if that's really plausible - how many people who earn $100,000 per year have spouses who work minimum wage jobs?

Is one spouse relatively low-earning because they don't work full-time or full-year? Again, we get back to the implicit income issue.

Is one spouse low-earning or unwaged because they're basically unemployed? Hidden unemployment - especially of women or men who have taken time out of the labour force to raise children - is a really serious issue. But I would think it would be better dealt with through training - there are some serious issues with EI, and its lack of responsiveness to this demographic - or other policies that focussed more on encouraging work effort on the margin, or otherwise easing the home-to-work transition. Another possibility way of recognizing hidden unemployment is to provide greater eligibility of at-home spouses for benefits - kind of like the law commission's Beyond Conjugality proposal.

Or is there a big earning disparity because the higher wage spouse is earning well into six figures? In this case - zero sympathy there.

It seems to me that, in your model, the low-earning or unwaged spouse is essentially consuming half of the couple's joint earnings and contributing an utterly paltry amount of goods and services. Are there really a lot of people out there like that? Who are they? I simply couldn't imagine doing that or being like that myself.

[Franceswoolley]

Seamus - good to see you here. Yes, a flat tax with a refundable credit for those with zero income would do it - so basically a flat tax is the limit of progressivity one can have and still satisfy the no marriage tax/same taxes for single and dual earner couples criteria.

[Matt Krzepkowski]

Perhaps that is why others refer to a 'participation tax rate', but it is still an effective average tax rate. The empirical work on the effect of taxes on hours/participation seems to support this.

The unwaged contribution of a stay-at-home spouse is not their contribution of goods and services, but the additional contribution of goods and services they provide over what they would if employed. Lots of work needs to get done somehow. There is an argument on potential tradeoffs in work done between spouses (or additional value of leisure) in single- and dual-earner households, but I haven't seen decent numbers.