Give the Rich More Money!

Against Means-Testing

This article was co-authored by guest writer Daniel Pareja and James Heaney. Calculations for this article are available in this Google Sheet.

---

See? Look how happy he is!

One day a few weeks ago, James and Daniel were having a friendly chat about universal basic income studies (as one does). Offhandedly, Daniel matter-of-factly remarked, “The real danger, of course, is that such programs might be income-tested, which would be a disaster, since all income testing is actually just a tax cut for high-income earners.”

James stared blankly for a few moments and then said (paraphrasing), “Daniel, what the hell are you talking about?”

Daniel explained. James found the explanation so interesting he asked Daniel to write this article, because he thought De Civ readers would be interested, too.

Four Ways to Look at Your Taxes

There are (broadly speaking) four different percentages that can be said to be someone's "tax rate." (For simplicity and clarity, this section considers only U.S. federal taxes on income.¹)

The first is what we’ll call your simple tax rate. This is just the percentage of your income that goes to the government. For example, in 2023, James’s total household income² was $95,904.³ His overall income tax was $15,281.⁴ We divide those two numbers to get James’s simple tax rate: 15.9%.

However, we don’t just give money to the government. Many of us also get money from the government. This may come in the form of tax credits, a welfare check, a student loan forgiveness, or a covid stimulus. This subsidy from the government “cancels out” some of what we have to pay them. What we really want to know, then, is the second kind of tax rate: your effective tax rate. This is exactly like your simple tax rate, except, once you find your overall income tax, you subtract all the money you got from the government.⁵ For example, this past year, the Child Tax Credit was worth $2,000 per child.⁶ James has two daughters, so the government gave him $4,000. That means James effectively paid only $11,281⁷ on his income of $95,904—much less than the $15,000 we were talking about before! We divide these two numbers to get James’s effective tax rate: 11.7%.

Third, we should consider the marginal tax rate. This is the amount of additional tax paid on your next dollar. If we had a flat tax, where the government simply took, say, 20% of everything, our marginal tax rate would be the same as our simple tax rate. However, most of the Western world uses a graduated (or “progressive”) tax system, which means the poor pay a lower rate and the wealthy pay a much higher rate. If you get more wealthy, you pay more taxes. For example, suppose James’s employer raised his salary by $1, so his income becomes $95,905. The federal government would take 30 cents of that dollar.⁸ That tells us James’s current marginal tax rate: 30%.

Notice this is much higher than James’s simple tax rate. This matters, because the more money you make, the less of it you get to keep. That’s frustrating. Why do the work at all, if the government is just going to take more and more of the fruits of your labor? You might even decide it’s not worth doing, and do something else with your time instead (like write a blog post about tax rates).

Marginal tax rates discourage people from earning more money, because earning more money causes the marginal tax rate to go up. This is an unfortunate but unavoidable cost of a tax system where the wealthy are taxed more heavily than the poor. I repeat: in a graduated tax system like ours, the richer you are, the higher your marginal tax rate.

At least, that’s how it’s supposed to work.

Finally, we come to the Big Bad of this article: the marginal effective tax rate. You can probably guess what this is from everything we’ve already discussed. This is the amount of additional tax paid on your next dollar, plus the amount of money you stop getting from the government now that you make more money.

The government gives the poor and the middle class a bunch of tax benefits that aren’t available to the rich. These programs “phase out” as you get wealthier. For example, the Child Tax Credit begins to “phase out” at $400,000. If you make $400,001, the Child Tax Credit is only worth $1,950/child. If you make $439,999, the Child Tax Credit is worth only $50/child. After that, the credit zeroes out completely.

This is called “income-testing,” a form of “means-testing.” It is supposed to be a way for the government to fulfill its obligations to the poor without giving away free money to the rich. After all, the rich don’t need the money, and paying them would cost the government a lot of money. The practice is very old.⁹

However, this turns into a one-two punch for the non-wealthy taxpayer climbing the ladder. As the taxpayer earns more income, his tax bracket goes up (because the richer you are, the more you pay). However, at the same time, the taxpayer is also losing income because government programs he used to fully qualify for are only paying a fraction of what they used to. His marginal effective tax rate can therefore be very high.

How high? Let’s find out.

Our Canadian Cousins

In the spirit of international comity,¹⁰ we will now switch from American to Canadian taxes. Follow the footnotes, though, because we’ll use them to point out the equivalent tax concepts in American law.

Consider two Canadian friends of ours:¹¹

• Cindy is a married taxpayer with three children, ages 3, 4, and 10, living in British Columbia. Cindy’s salary is $500,000.¹² (Her husband stays home with the kids.)

• Thomas is also a married B.C. taxpayer with three children of the same ages. His spouse also stays at home. However, Thomas’s salary is only $45,000.¹³

Cindy is—to use a bit of fiscal jargon—“filthy rich.” She qualifies for no means-tested tax credits at all, and she is far above the “caps” on social insurance payments.¹⁴ So, she simply pays her taxes. Based on her tax bracket, 20.5 cents of the next dollar she earns will go to her provincial government in Victoria, and 33 cents to the federal government in Ottawa. Cindy’s marginal tax rate is therefore 53.5%. Since she qualifies for no income-based tax credits, Cindy doesn’t lose out on any government subsidies by earning more, so her effective marginal tax rate is the same: 53.5%.

Things are not so simple for Thomas. His tax bracket means that he will owe 15 cents of his next dollar to the federal government and 5.06 cents¹⁵ to his provincial government. However, Thomas is also below the caps on social insurance payments to the Canada Pension Plan and Canadian Employment Insurance.¹⁶ Those will cost him an extra 6.06 cents. Thomas’s marginal tax rate, then, is 26.1%.

That’s way below Cindy’s, which is exactly how it’s supposed to work: Cindy is rich, Thomas is poor, so Cindy pays a greater share of her income to support the nation.

But hang on.

Thomas is also benefitting from a number of means-tested Canadian tax credits. As his income rises, he gradually loses access to those credits.

First, Thomas benefits from the Goods & Services Tax Credit (GST).¹⁷ The full amount of the credit would be $1,217. At his current income of $45,000, Thomas is in the means-testing “phase out” range, so his credit is smaller but still hefty: $1,203.95. However, if Thomas’s income rises to $45,001, his GST tax credit falls to $1,203.90. When he earns that dollar, then, Thomas has to pay 26 cents of that dollar to Canada, and he loses 5 cents Canada was previously paying him. In other words, Thomas only gets to keep 69 cents of his shiny new dollar. If his earnings rise to $45,002, same thing: he pays 26 more cents to Canada and he loses 5 more cents from the credit.

Still, Cindy’s only keeping 46-and-a-half cents on every new dollar she earns. Surely, then, Thomas is still doing better than Cindy.

Thomas receives money from the Canada Child Benefit. The full amount would be $22,144 for his three children, but Thomas is in the means-testing “phase out” zone, so he actually gets $20,608.¹⁸ Then, for every additional dollar he earns, the benefit falls by 19 cents.

Thomas also receives $202.35 from the Canada Workers Benefit,¹⁹ although the full credit would be much larger ($2,616) if his income were half what it is. Alas, when Thomas earns his next dollar, his Canada Workers Benefit will shrink 15 cents to $202.20.

But keep going, buddy; we’re not out of this yet. Thomas also has means-tested benefits from the provincial government of British Columbia. Most of them either don’t apply to him or he’s below the phase-out threshold, but he does get the British Columbia Family Benefit. The full amount of the credit is $4,688, and Thomas still gets $4,341. However, every time Thomas earns another dollar, that credit shrinks by 4 cents.

Let’s add this all up. When Thomas earns his next dollar, 26 cents of it go to Canada for income/payroll tax plus Thomas loses a total of (5+19+15+4=) 43 cents in tax benefits. His marginal effective tax rate is a whopping 69%²⁰—more than 15 points higher than Cindy’s! When he earns his next dollar, he’ll be only 31 cents richer! Thomas is lower-middle class, yet, somehow, he’s getting taken to the cleaners!

There’s one more, pernicious twist to all this.

Suppose, next year, Cindy and Thomas have another child.

Because Cindy is rich, Cindy’s tax bill is unaffected.

From one perspective, Thomas benefits a lot: his annual Canada Child Benefit expands from $20,608 to $28,072! Of course, his children are future taxpayers whom Thomas has produced for the betterment of all Canada, and, even though Canada’s child benefits are far more generous than America’s,²¹ this credit still probably does not cover Thomas’s expenses! Nevertheless, for parents, more help is always welcome.

From a different perspective, though, Thomas takes it on the chin: his larger Canada Child Benefit is means-tested much more strictly, which means it phases out way faster. Whenever Thomas earned an extra dollar last year, he lost 19 cents from the Canada Child Benefit. Now that he has another kid, he will lose 23 cents from the Canada Child Benefit. Having more kids increases Thomas’s marginal effective tax rate to 73%. For every extra dollar he earns, he’ll only take home 27 cents.

Why This Number Matters

You might instinctively recoil from the idea that anything bad is happening to Thomas here. Yes, he gets to keep less of his next dollar than Cindy… but that’s because he’s been getting a bunch of money from the government dole, and Cindy hasn’t! It’s fair that, as Thomas becomes more capable of supporting himself, he should get off welfare in order to make room for others more needy.

However, economics is all about incentives. People do what the law encourages them to do. One of the things people really like is making more money. Growing, lower-middle-class families like Thomas’s especially benefit from being able to move out of an apartment into a house, buy a bigger car (to comply with often-burdensome car seat laws!). They benefit from earning more money, because they are able to get the things they need. We benefit from them earning more money because they become more responsible for their own lives, they avoid becoming dependent on government, and they (generally) perform more socially valuable work.

All things considered, then, it would probably be a good thing for Thomas’s kids, and for the nation as a whole, if Thomas switched from his current $45,000/year job to a $75,000/year job, even if he enjoyed the new job less and found it less fulfilling. That extra $30,000 a year could be transformative.

…but how likely is he to switch to a job he hates for the sake of his family if $30,000 extra dollars “on paper” turns out to be only $11,000 in terms of extra money he can actually spend? That’s what a 73% marginal effective tax rate²² is doing to Thomas: it is making his extra dollars worth so little that it’s unclear that he should even bother trying. Maybe he’ll be desperate enough to go for it anyway! Or maybe he won’t: he could pass up the better job opportunity and a chance to become more responsible for his financial future, because it would be 100% of the work for only 27%²³ of the benefit.

Or he might pick Door #3: he might just decide not to have another kid at all, because of the drag it would put on his attempts to grow his income and provide a better life for the kids he already has. So we see that high marginal effective tax rates can have contraceptive and abortifacient effects.

In the end, things will probably work out okay for Thomas’s family, just like millions of other lower-class families muddle through every day in both the U.S. and Canada. Nevertheless, even taking into account the fact that Thomas’s high tax rate is partly because of the government’s generosity to him, it still seems unfair. The notion that a low-income person like Thomas might have a higher tax rate than a rich person like Cindy—for any sensible definition of "tax rate"—seems to run counter to the premise of a graduated income tax system. As Adam Smith wrote in The Wealth of Nations:

The necessaries of life occasion the great expense of the poor. They find it difficult to get food, and the greater part of their little revenue is spent in getting it. The luxuries and vanities of life occasion the principal expense of the rich; and a magnificent house embellishes and sets off to the best advantage all the other luxuries and vanities which they possess…. It is not very unreasonable that the rich should contribute to the public expense, not only in proportion to their revenue, but something more than in that proportion.

Even setting aside this sense of unfairness,²⁴ it still seems likely that such a high marginal effective tax rate is a big drag on Thomas’s family and a teensy-tiny drag on his nation as a whole.

Multiply this teensy-tiny drag by every single middle-class family in the country, however, and pretty soon you’re talking about real money.

And real babies.

You Need Babies

It’s no secret that the fertility rate in developed countries is below the replacement rate: 2.1 children per woman.²⁵ In fact, every high-income country has a fertility rate well below replacement. France, at 1.8, is the best the developed world is managing right now. South Korea, at 0.9, will soon enter a population contraction the likes of which are historically associated with the Black Death or total war.²⁶ In response, some nations have turned to immigration to shore up population numbers; others have tried various financial incentives to have more babies. ²⁷

However, if the tax code itself creates disincentives to births, that will blunt the effectiveness of any other population-sustaining policy a government might try to adopt. Financial incentives won’t work if they’re counteracted by financial disincentives, and immigration won’t work if immigrants quickly experience the same disincentives and suppress their fertility accordingly.

In particular, what we see here are disincentives to birth that especially impact the middle-class and lower-middle class, while the wealthy and the very poorest²⁸ largely avoid it. This risks hollowing out the middle class, the political community which Aristotle hailed as “most perfect,” warning that a state without a thriving middle class “is soon at an end.”

How to Fix It

We are not the first people to describe this problem with marginal effective tax rates. Anyone who was around for the U.S. welfare reform debates in the 1990s probably knows more about it than we do. (Well, more than James does, anyway.) We know that some De Civ readers have gotten this far and are going, “Yeah? So? Governments are doing their best.” It is, in fact, quite well known in tax policy circles that any means-tested program disincentivizes work. Tax legislation has made huge strides in the past few decades to make things a lot better than they used to be.²⁹

Yet the problem persists, because it is inherent in any system of means-testing. If the government gives you a benefit because you are poor, it has to take that benefit away when you stop being poor. It can do that slowly or it can do it quickly, but it has to do it, and that spells upward pressure on marginal effective tax rates for people who aren’t rich. You can design a tax regime to minimize this effect,³⁰ but you can’t escape it.

So… maybe we should scrap all income testing and caps.

If the government thinks families should be subsidized to the tune of $7,000 per child for raising children, it could simply give $7,000 to every parent raising a child, regardless of income level. If the government thinks poor people should receive half their retirement savings back as a tax credit, give the same cash to rich people.³¹ This would solve the marginal effective tax rate problem. We will illustrate:

Suppose Canada abolishes the income phase-outs. Everyone who qualifies gets the full value of the Child Benefit ($22,144 for three kids), the Workers’ Benefit ($2,616), the B.C. Family Benefit ($4,688), and the Goods & Services Sales Tax Credit ($1,217), while also ending income caps on Canada’s Pension Plan and Employment Insurance. In short, Thomas now receives considerably more money from the government ($32,500 instead of the $27,500 he used to get). How does this affect his tax rates?

• Thomas’s Simple Tax Rate is unchanged: he still pays 20.06% in state and provincial income tax, plus 6.06% in social insurance. After exemptions,³² that still adds up to $8,500 in taxes, for a simple tax rate of 19%.

• Thomas’s Marginal Tax Rate is also unchanged: his next dollar will be formally taxed at the full rate of 26.1%, just like before.

• Thomas’s Effective Tax Rate falls, because he is receiving more benefits from the government without paying more to the government. Specifically, it tumbles from negative-42% to negative-53%.³³ Under neither circumstance can Thomas accuse the government of being miserly toward him!

• Thomas’s Effective Marginal Tax Rate, however, falls drastically: since he no longer loses government money just for earning more money of his own, his effective marginal tax rate tumbles from 69% all the way down to 26.1%—the same as his marginal tax rate! (Just like Cindy!) Thomas’s disincentive to earn money has just been cut by nearly two-thirds!

Of course, Thomas would not be the only beneficiary of this. If means-testing is abolished, the glamorous and wealthy Cindy gets all the government subsidies as well, from the Workers’ Benefit to the Child Benefit—the same $32,500 Thomas now receives. Let’s see what this does to her tax rates:

• Cindy’s Simple Tax Rate rises. She’s paying the same income tax rate as before,³⁴ but she now also has to pay payroll tax on her entire salary (not just her first $65,000 or so). This sends her total tax bill up from $224,000 to $248,000, and her rate therefore rises from 44.8% to 49.7%.

• Cindy’s Marginal Tax Rate also rises, because she is now paying social insurance taxes even at the top of the income spectrum. Her next dollar will be formally taxed at 53.5% in income taxes (just like before), but payroll tax now gets a bite of her dollar, too. Her marginal tax rate becomes 59.3%.

• Cindy’s Effective Tax Rate both rises (because her taxes go up) and falls (because she is now—for the first time—receiving government subsidies). In the end, the benefits win out! Her effective tax falls from 44.8% to 43.2%.

• Cindy’s Marginal Effective Tax Rate matches her marginal tax rate, just like it used to: they’re both 59.3%.

By abolishing means-testing and income caps, we seem to have brought our tax system back into alignment with the graduated ideal: people with more pay a larger share.

However, we’ve also just blown a hole in the budget! Thomas is now receiving an extra $5,000 or so from the government and not paying any additional taxes. Cindy is also now receiving nearly $32,000 from the government that she never received before. She is paying additional taxes, to the tune of nearly $25,000—but those new taxes aren’t enough to cover the extra benefits that she and Thomas are receiving! The government ends up over $12,000³⁵ short under our “end means-testing” plan. Multiply that revenue gap across 40 million Canadians (or 350 million Americans, if you do this in the States) and, hoo boy, you’ve got a fiscal crisis on your hands!

This fiscal crisis can be solved, though: just raise overall income taxes to pay for the expanded subsidies. If the top Canadian federal tax rate rises from 33% to 38%, that raises enough extra revenue from Cindy to fill the fiscal hole opened up by her own benefits.

Alternatively, you might point out that Cindy is already facing fairly huge disincentives to work and we shouldn’t hike her marginal tax rate any further. Fair enough! You could also close the budget deficit by hiking the bottom tax rate from 15% to… uh…

*checks notes*

“Seriously?”

“That’s what it says! Tell them!”

What you have to do to close the deficit this way is raise the bottom federal tax rate all the way to 58%. Even the richest Canadians, like Cindy, are only taxed at 33% federally on their top dollars. This would tax everyone’s first income at nearly double that rate. It would be nuts to adopt it. Turns out, it’s really hard to raise large amounts of new revenue from poor people!

Your other option is to do what a graduated tax system usually does: share the burden. Raising the bottom rate by 3.5% (to 18.5%) ensures Thomas has some skin in the game.³⁶ Raising the top rate by 4.5% (to 37.5%) closes the rest of the revenue gap while retaining the graduated ideal.³⁷

Obviously, this example is drastically simplified. There are more than two taxpayers in all of Canada, and those taxpayers are in all kinds of different tax situations at different income levels with different benefits available to them. We are unable to calculate exactly how big of a budget deficit Canada would create if it eliminated means-testing for all tax credits, nor can we say exactly how much Canada would need to hike general income taxes in order to make up the difference. This is to say nothing of America, the country James actually cares about!

Our point is simply this: you can pay for the elimination of means-testing by raising general income tax rates. The necessary tax hikes may even turn out to be surprisingly modest. Doing this completely eliminates the perversity of a “graduated” tax system where people in the lower classes have a higher marginal effective tax rate—and therefore stronger disincentives to work—than wealthy people.

This brings us back to Daniel’s startling claim at the beginning: since governments could fix this by ending means-testing and progressively raising income tax rates, the fact that they choose not to do so means all means-tested programs are, in effect, tax cuts that primarily benefit the rich.

Means-testing looks superficially generous: the taxpayers are giving money only to the actually needy, and keeping it out of the hands of rich people who don’t. This feels virtuous, and it is overwhelmingly popular. Taxpayers support means-testing even though it has a host of other problems, from creating a vast, wasteful means-testing bureaucracy to requiring poor people to navigate the vast, wasteful means-testing bureaucracy (with all that spare time working single dads famously have). Yet we see the bureaucracy as necessary. How else can we make sure that we don’t accidentally give rich people a bunch more money?

As Daniel has shown, though, there’s a case to be made that this is exactly what we should do: for the sake of the poor, we should give the rich more money!³⁸

---

1. This includes FICA (payroll tax). It excludes state taxes and federal taxes not based on income, such as consumption taxes or tariffs. For simplicity, the analysis excludes capital gains and dividends, which receive favorable tax treatment in the United States (and Canada, and most of the rest of the world, for that matter).

2. Federal Form 1040, line 9 (married filing jointly). His adjusted gross income (line 11) was identical: $95,905. His gross income (before pre-tax 401k deductions and such) is a little higher than that, in the neighborhood of $110,000.

3. James here: This puts my family neatly in the upper-middle class.

By income, we’re in the Bottom 65 percent (or the Top 35 percent, as I prefer to say). Obviously I used my 2022 pre-tax income for that calculator ($106k and change), so we’d be a little higher now—maybe the top 33 percent.

I tend to define the middle class by income quintile, at least when we are talking about economic class and not social class:

• The “middle middle class” has a household income between the 40th and 60th percentile (roughly $58,000 - $94,000).

• The upper-middle class lives between the 60th and 80th percentiles (up to roughly $150,000).

• The lower-middle class lives from the 20th to 40th percentiles (down to $30,000).

• Anyone with less than $30,000 of household income is poor—sorry, “working-class” is the preferred corpospeak of the day.

• Any household bringing in more than $150,000/year is wealthy.

4. Form 1040, line 7 ($7,705) plus their payroll taxes from W-2 boxes 3 and 4 ($7,576).

We are not including the equal portion of the payroll tax that is assessed to employers, even though the tax incidence of that likely falls heavily on the employee.

5. …for reasons other than services rendered, that is. Soldier salary is not a welfare program; it’s earned income!

6. This number has bounced around a lot since 2018, and is likely to do so again when the Trump expansion of the child tax cut expires next year. Both parties want to renew it, but neither knows where to find the money.

7. Actually, James also got $21 from the “foreign tax credit,” whatever that is, so he really only paid $11,260 in income taxes.

8. Medicare gets 2 cents, Social Security gets 6 cents, and the income tax takes 22 cents, based on the current tax bracket for a married couple’s 95,905th dollar.

9. While modern welfare states only began to emerge in the late 19th century and did not reach their current form until after the Second World War, government programs providing support to the poor are, of course, nothing new. By the late Roman Republic, the grain dole aspect of the Cura Annonae had become a permanent fixture of the state, providing subsidised, or later free, grain to poor people in Rome. In the early Empire, Tiberius noted that it was his duty as Emperor to provide it, and that ending it would result in "the utter ruin of the state". The program persisted in some form, whether in Rome or in Constantinople, for well over half a millennium afterward.

10. …that is, because Daniel was way more willing to do the research / math.

11. …that is, we made them up.

12. Her 2023 net income and taxable income were therefore both $499,369, if you really want to get into the weeds of federal lines 23600/56 and 26000/68. This is slightly less than her salary because a portion of her Canada Pension Plan contributions (made on her first $66,600 of income) were deductible.

13. His 2023 net income and taxable income are therefore $44,585, for the same reason as Cindy’s in the previous footnote: Canada Pension Plan reductions.

14. Just like America’s payroll taxes, Canadian payroll taxes (the Canada Pension Plan plus Employment Insurance premiums) are “capped,” so you don’t have to pay them anymore once you’ve earned a certain amount of income. In 2024, the Canada Pension Plan taxes are capped at $68,500 (with the employee paying a maximum of $3,867.50), and the Canadian Employment Insurance tax is capped at $63,200 of income (with the employee paying a maximum of $1,049.12).

Footnote to the footnote: in the Canada Pension Plan, the first $3,500 of income is also exempt from the tax, so you’re only paying the CPP tax on $65,000 of your hard-earned Canadian dollars. This is completely irrelevant to our argument, but it might throw you off if you are trying to check our figures and don’t know about it.

Second footnote to the footnote: Canada presents EI and CPP payments as “insurance premiums,” not “tax.” However, Daniel and James agree that a insurance program you are automatically enrolled in and which the government mandates you to pay for is a tax, at least for our purposes today.

15. Yeah, fractions of a cent are fun. Government, right?

16. The Canada Pension Plan is basically equivalent to U.S. Social Security. Like Social Security, it’s collected from a payroll tax, the tax rate is similar (5.95% in Canada versus 6.2% in America), and both only tax a portion of income ($65,000 CAD to a maximum of $3,867 versus $168,600 USD to a maximum of $10,453.20)

Canadian Employment Insurance is basically equivalent to U.S. Unemployment Insurance, but there are some differences: Canadian E.I. is a federal program. America’s U.I. is a mandatory federal program managed by the states, with a lot of state-by-state variation. We will use Minnesota, since James lives there.

Nearly half of Canadian E.I. comes directly out of an employee’s paycheck; the remainder from the employer. In Minnesota, U.I. is technically assessed exclusively to the employer, which means it never shows up in employee income in the first place. Canadian E.I. is assessed at a flat rate: 1.66% in 2024. Minnesota U.I. is assessed at a base rate of 0.1% for employers with extremely low unemployment, but employers with lots of unemployment pay much higher rates—up to 9%—with additional assessments in years when the state’s U.I. trust fund sinks below a certain threshold.

James legitimately has no idea what rate his employer pays for U.I., which is one reason why it was easier to do this analysis with Daniel’s Canadian figures.

17. This is a tax credit for paying sales tax, at least in theory. (In practice, you just get the credit, without having to prove you actually bought anything.)

The United States doesn’t have a national value-added sales tax like Canada does, so they have no exact equivalent for this. The State and Local Tax Deduction (SALT) is probably the closest: SALT gives you a deduction (not a credit) for paying non-federal taxes, which often includes sales taxes. However, you only use the SALT deduction if you itemize, which almost nobody does, because almost nobody is wealthy enough, or doing enough weird deductible stuff, for itemizing to make sense.

18. This is similar to the U.S. Child Tax Credit, but much larger and much more harshly means-tested. For example, James receives the full U.S. Child Tax Credit, and is unlikely to enter the phase-out zone at any point in his future career.

But Thomas, who makes less than half of what James brings home (even less if you convert USD to CAD), is already well into the phase-out zone for the Canada Child Benefit.

19. The U.S. equivalent of the Canada Workers Benefit is the Earned Income Tax Credit. It’s designed—perhaps a little ironically—as a program to incentivize work. Our understanding is that the U.S. EITC phase-out is somewhat more carefully crafted to avoid penalizing people for earning additional dollars. The Canada CWB appears to be known best for its overly complicated form, and, uh… yep, after filling one out for Thomas, we agree!

20. nice

21. America currently provides a (partly refundable) tax credit of $2,000 per child. However, this higher allowance was only established recently, by the Trump tax cuts, and is scheduled to revert to $1,000 per child in 2025. On the other hand, America’s child tax credit is far more broadly available, because its phase-out kicks in at a much higher threshold. On the other hand, because the credit is partially non-refundable, it doesn’t reach many of the poorest families that need it most.

James nevertheless admits to some jealousy: after 2025, the U.S. Child Tax Credit will pay him $2,000/year. In Canada, even with the much stricter means-testing, he would be entitled to $5,800/year, which would be pretty handy.

He comforts himself by reminding himself that this would be in “worthless” Canadian dollars, therefore no great loss. (He’s wrong. $5,800 CAD is still $4,300 USD.)

Both U.S. political parties have agreed that they want to maintain (or grow) the expanded child tax credit, but doing so in the shadow of multiple looming fiscal crises may prove challenging.

Meanwhile, several U.S. states offer their own child tax credits, of various sizes. Many have much lower phase-out thresholds than the federal credit.

22. “Wait just a cotton-pickin’ minute!“ you say. “A 73% marginal effective tax rate wouldn’t turn $30,000 into $11,000! $30,000 * (1 - .73) = $8,100! That’s the number these guys should be using!”

Not quite. When Thomas’s income jumps by $30,000, the first few thousand dollars he earns are indeed taxed at a marginal effective rate of 73%. However, as some of his benefits finish completely phasing out, and as he rises into a new income tax bracket, this changes, steadily. The $29,999th dollar Thomas earns will only be taxed at a marginal effective rate of 57.2%—still higher than Cindy, but now just a few points higher.

Daniel did the math. He used 2023 tax rates (“if for no other reason than that all the forms are much more readily available”), he assumed Thomas went ahead and had that fourth child, and, for the purposes of working out the true marginal income tax rate (METR) at each level, he also factored in other credits that we had excluded from our analysis (the British Columbia Climate Action Credit, the Canada Employment Amount, and similar). Overall, he finds that the average METR on Thomas’s $30,000 raise is 63.8%, which works out to an effective take-home pay of $10,860.

This next section is all Daniel:

To begin with, put together, at $45,000, the total simple tax liability is $8,525.84 (18.9%). At $75,000, it is $17,699.85 (23.6%).

With four children, the maximum GST credit for a married couple is $1,396. This vanishes entirely at $72,875, but, at an income of $45,000 (which is treated as $44,585 for the purposes of these benefits), the benefit is $1,382.95.

The maximum Canada Child Benefit available with four children as above is $29,931. At $500,000—that is, Cindy's income—no benefit is received, of course, but at $45,000 the benefit is $28,071.91 and at $75,000 the benefit is $21,221.59.

For the BC climate action credit, for a married couple with four children, at $45,000 the full credit of $1,260 is received, and at $75,000 a reduced credit of $918.38 is received.

For the CWB, at $45,000, the benefit is $202.35.

For the BC climate action credit, at $45,000 the full credit of $1,260 is received, and at $75,000 a reduced credit of $918.38 is received.

The maximum BC family benefit is $5,813; it comes with the proviso that it cannot drop below $3,718 for incomes below $114,887. At $45,000, this benefit is $5,465.69; at $75,000 it is $4,274.32. (It would require an income of $88,908 to be reduced to the floor.)

In all, if I've done my math right, at $45,000, Thomas would receive a net benefit of $27,857.06, while at $75,000 his family would receive a net benefit of $8,714.44.

That means that to earn an extra $30,000, he has lost $19,142.62 between extra taxes and reduced benefits, effectively being taxed at 63.8% on that $30,000.

His METR at $75,000, meanwhile, is 57.2%, still higher than Cindy's, but not as high as his METR at $45,000, where it was about 73.12%. It jumps around as he becomes ineligible for certain credits, moves into the phaseout zone for others, goes up a tax bracket provincially and federally, and hits caps on CPP and EI taxes.

Were he to earn $85,000, meanwhile, his METR would plummet, entirely due to the drop in the phaseout rate for the CCB, going to 43.7%. By $90,000 it would drop further to 39.7%. The rate would keep jumping around after this as he moves up tax brackets and becomes ineligible for income-tested benefits, or, for the BC family benefit, has the phaseout kick back in.

23. Technically, 36.2% of the benefit, due to varying tax rates on various chunks of his $30,000. See previous footnote.

24. …which you may or may not share. We are not going to try to turn this article into an analysis of the moral underpinnings of our intuitions of fairness, justice, and taxation.

25. …at least in high-income countries where infant mortality is low and famines rare.

26. The causes of this decline are complex, disputed, and beyond the scope of this article.

27. The reasonableness of these policies is beyond the scope of this article.

The success or failure of these policies so far is—you guessed it!—beyond the scope of this article.

28. One effect of linking so many benefits to income is that it requires people to file a tax return in order to access those benefits. Unfortunately, it is disproportionately the very poorest, who stand to gain most from these benefits (in addition to receiving any regular refund to which they may be entitled) who do not file.

However, the reasons for, and consequences of, this effect are—say it with us—beyond the scope of this article.

29. There was a time, not that long ago, when you might receive a $5,000 welfare benefit from the government as long as your income stayed below, say, $20,000 a year. However, if you earned $20,001 per year, you lost the entire benefit, instantly. The effective marginal tax rate on that dollar, then, was not a piddling seventy-two percent, but instead well over forty-nine thousand percent. Obviously (obviously!), nobody with any sense would ever get off the dole unless they either won the lottery or had an unusual amount of determination, intelligence, and basic pride, because they couldn’t make the leap from $20,001 to $25,001 (or whatever the actual numbers were) in a single day and couldn’t afford not to.

This is why we invented income phase-outs for means-tested programs! Even though it has its problems, the current system is a lot better than the old system!

30. The Earned Income Tax Credit in the United States was explicitly drafted to ease disincentives to work. Nevertheless, the sheer size of the credit, the rapidity of its phase-in, and the need to phase all that money out before you reach the middle class add up to a situation where the EITC nearly has a stronger effect on marginal effective tax rate for a father of three than the Canada Workers Benefit. There’s still plenty of other uncomfortable distortions in the States (scroll to the section about current marginal tax rates). You really can’t avoid it in a means-tested tax system.

31. Theoretically, the same logic applies to credits like the American Opportunity Tax Credit (which gives students a credit for attending college) and the Affordable Care Act Premium Tax Credit (which gives families cash for buying medical insurance on the Obamacare exchanges): just give the same subsidy to everyone!

You might reply, “Hey, if we give everyone $2,500 for going to college, won’t colleges just raise tuition by $2,500 and absorb all the benefits themselves?” Yes, of course! Colleges already did that. Cheap federal subsidies for the college students are a clear cause of the radical, much-faster-than-inflation increase in the cost of college. (Not the only cause, but a clear one.)

Tax credits that simply put taxpayer money in the pocket of powerful corporations shouldn’t be expanded; they should be abolished, because they don’t help anyone. But, if, for whatever, reason we decide to keep these irrational credits on the books, naturally we would make them available to everyone in order to avoid hiking the marginal effective tax rates of poor people.

32. You’ve already had to learn enough refundable Canadian tax credits for one day, dear reader; we aren’t going to make you learn about the non-refundables, too. See our calculation spreadsheet for details.

33. Originally, Thomas paid $8,525.84, but received $27,489.21. Since he received more government money than he paid, Thomas’s effective tax rate was negative. Specifically, it was ($8,525.84-$27,489.21)/$45000 = -42.14%.

If he receives the full value of all benefits, without income phase-outs, he starts receiving $32,470.00, which you plug into the same equation to get -53.21%.

The Canada Child Benefit and B.C. Family Benefits are the vast majority of the government largesse he receives. If you exclude these pro-natal benefits from our analysis, Tom only received $2,540.30 from the government under means-testing, and $4,967.00 with means-testing eliminated. His effective tax rate remains positive in both cases: he pays an effective tax rate of 13.3% with means-testing phase-outs and 6.4% without them.

34. Her income tax bill actually goes down slightly, because her Canada Pension Plan contributions, which are now uncapped, rack up a larger deduction. Her provincial + federal income tax bill was $217,576 when everything was capped, but now becomes $215,268.

The increased CPP contributions more than make up for this increased deduction, unfortunately for Cindy.

35. Precisely $12,793.96. Thomas’s new benefit is $4,980.79, Cindy’s is $32,470, and Cindy’s new taxes generate $24,656.83 in revenue. Add the first two, subtract the third, boom, there’s your budget deficit.

36. Note, however, that you wouldn’t want to do exactly this. You want Thomas and other people who are benefitting from the end of benefit caps to have to pay for some of it. You don’t want the truly poor people who are already receiving the full benefit amounts (because they are so poor) to face an income tax hike to help pay for it. That’s all pain for them. However, since raising the bottom rate necessarily affects all those people in the bottom tax bracket, that’s exactly what you’d be doing.

If you took this general approach to sharing the tax burden, what you would probably want to do instead is introduce a new second-lowest tax bracket, consisting of people (like Thomas) who are in the means-testing range, would benefit from the end of means-testing, and attach a small tax increase to those people.

However, this whole footnote risks sending us deep into questions of tax incidence and who ought to bear which burdens of taxation in a free and democratic society. All of that is outside the scope of this post.

Our point here is simply that, if you end means-testing, you will have to raise income taxes to make up the revenue shortfall, and there are several different ways of sharing the burden among your citizens, depending on how your voters view the fairness question.

37. Here are the results of this tax increase on the effective tax rate for both Thomas and Cindy:

• Under the original, means-tested system, Thomas’s effective tax rate was -42.1% and Cindy’s was 44.8%.

• Under the uncapped system, Thomas’s effective tax rate became -53.2% and Cindy’s 43.2%—lower than it had been under means-testing!

• Now that we’ve raised taxes to pay for it, though, Thomas’s effective tax rate rises a touch to -51.2%, and Cindy’s are back up to 45.6%. Cindy’s effective tax rate is now higher than it was under means-testing. However, she could get her effective rate back down below where she started again by having a fourth child.

38. …then tax it away again.