The marginal rate is the one that decides whether you work the extra shift

Let me tell you about the single most misunderstood sentence in American life: "If I take that overtime, it'll push me into a higher bracket and I'll take home less." Folks say it at break room tables and kitchen counters every autumn, with total sincerity, and it is wrong in the one way that matters and right in a different way that nobody explains. Today we sort out which is which, because the difference is the difference between understanding your paycheck and being quietly fleeced by a misunderstanding of it.

Two numbers describe your tax rate, and people use them as if they were one. The first is your average rate: total tax divided by total income. The second is your marginal rate: the tax on the next dollar you earn. The average is a report card on the past. The marginal rate is a price tag on the future. And just like every other price, it changes what people do.

Myth first: a raise cannot cost you money through brackets

Start with the rumor, because it is the one that makes people turn down work. Under a bracket system, each rate applies only to the dollars inside its band. Crossing into a higher bracket does not re-tax everything you already earned at the higher rate. It taxes only the dollars above the line.

Here is a labelled hypothetical. These are made-up brackets chosen for easy arithmetic, not the real schedule, so do not use them to file anything.

• 10 percent on taxable income up to $20,000.
• 20 percent on taxable income from $20,000 to $60,000.
• 30 percent on taxable income above $60,000.

Take a worker with $59,000 of taxable income. The tax is 10 percent of $20,000, which is $2,000, plus 20 percent of the next $39,000, which is $7,800. Total: $9,800. The average rate is 9,800 divided by 59,000, about 16.6 percent.

Now she picks up a big overtime check and her taxable income goes to $65,000. Same schedule. The first $20,000 costs $2,000. The next $40,000, up to $60,000, costs 20 percent, or $8,000. The final $5,000 sits in the 30 percent bracket and costs $1,500. Total: $11,500. She paid $1,700 more in tax on $6,000 more income. She kept $4,300 of the extra $6,000.

Notice what did not happen. The overtime did not make her entire income taxable at 30 percent. If it had, she would owe $19,500 and be worse off. The federal bracket system does not work that way. A raise through brackets alone never lowers your take-home pay. Tell your cousin.

The truth underneath the myth

Now the part the cousin is half right about. The marginal rate on that last $5,000 was 30 percent, three times the rate on her first dollars. She kept seventy cents of each dollar, not ninety. And that is what matters for the decision to work the Saturday shift.

Economists have a tidy way to say it. An hourly wage of $30 is a pay rate. The government's marginal rate turns it into an after-tax pay rate. If you are in the 30 percent band, the $30 hour is a $21 hour. If you are in the 10 percent band, it is a $27 hour. The employer pays the same $30 either way. The worker gets $6 more or less for exactly the same effort, and has every reason to notice.

This is not an exotic theory. It is the same logic you apply when a store raises the price of something you buy. Raise the price of leisure's rival, which is work, by taxing the extra hour harder, and you get a little less of it. Not none. A little less. Multiply a little less across the whole workforce and you have a fair-sized economy shaving its own output.

I should be straight with you about the other side, because the honest argument is stronger than the cartoon. Economists call two forces the substitution effect and the income effect. A higher marginal rate makes leisure relatively cheaper, so people may work less: the substitution effect. But a higher tax also leaves you poorer, so some people work more to keep their standard of living: the income effect. Which one wins depends on who you are. A second earner in a household deciding whether a job is worth the child care tends to be responsive to the rate. A single owner working sixty hours a week to keep a business afloat may barely change. My argument is that responses are likely modest for many prime-age workers and larger at the margins: second earners, people near retirement, and anyone who can change when and how income is reported.

That last group is where the action is. At the top, a high marginal rate does not usually make anyone stop working. It makes them rearrange: take compensation as stock rather than salary, defer income, or shift it to a form taxed more lightly. A marginal rate is a tax on a decision, and clever people respond cleverly.

The numbered argument

Let me lay the argument out in order, so you can take it to the break room.

1. Brackets tax slices, not totals. The rate on the slice is the marginal rate. Everything below it is taxed at the lower rates, always.
2. Decisions happen at the margin. Nobody decides whether to work all year in one stroke. They decide on this shift, this project, this second job, this hire. The rate that applies to that choice is the marginal rate, not the average.
3. The marginal rate stacks. Federal income tax is one layer. Add the employee share of payroll tax, 7.65 percent for most wage earners on the first layers of income, and then any state income tax. A worker in a state with an income tax who lands in the 20 percent federal band can have a combined rate well above 30 percent on the next dollar. A self-employed worker pays both halves of the payroll tax, 15.3 percent, with a partial deduction for the employer half. The combined number is the real price tag on the next hour.
4. Rates matter more for the choices that are easiest to change. Working a second job, expanding a store, hiring a person, selling a stock, retiring a year early. Those are the margins on which a high rate does its damage.
5. Therefore a tax system that wants more work should look at the top of each bracket, not the average rate. Lowering the rate on the next dollar is not the same as handing out a refund. It changes behavior going forward, and that is the entire logic of cutting marginal rates.

The steelman, fairly stated

The best liberal answer goes like this. Marginal rates matter, sure, but the empirical response of most workers is small. People work because they have bills. A tax cut that costs the Treasury a great deal in lost revenue for a tiny change in hours is a bad trade. And the top rates are low by historical standards, when the economy grew well with rates far higher.

Let me answer that point by point, and with respect, because it is a real argument.

First, the response of hours is small for many workers, yes. But hours are not the only thing that responds. Reported income, the form income takes, investment, the choice to incorporate, and the decision to start or expand a business all respond. Economists who study taxable income treat shifting, timing and avoidance as margins of response alongside hours worked, and those are exactly the margins where a rate cut or increase changes revenue. We look at when that logic pays for itself, and when it honestly does not, in where tax cuts pay for themselves and where they honestly do not.

Second, "top rates were much higher and the economy grew." True of some years, and the fine print is the point. When the top federal rate was very high, the code was stuffed with shelters, deductions and special forms of income, so that few people paid the statutory top rate. The rate on paper and the rate paid were different animals. The 1986 reform traded lower rates for fewer shelters, and we cover how the code filled back up afterward in tax complexity is a subsidy for lobbyists, and the code shows how it works. A history of the schedule itself is in a century of top income tax rates, from 7 percent in 1913 to the 2018 law.

Third, a rate cut that does not pay for itself can still be worth it. I would not tell you every tax cut pays for itself, and any commentator who does is selling something. The case is that a lower marginal rate buys more work, saving and risk than the static arithmetic captures, and that the choice about how big the government should be belongs to voters, not to a tax scheme designed to hide the bill.

Where the marginal rate hides: the cliffs

A bracket schedule is the visible part of the marginal rate. The invisible part is the benefit phase-out. When a family's income rises and the earned income credit shrinks, food assistance falls, a health subsidy disappears and a housing voucher costs more, each of those losses acts exactly like a tax on the next dollar. Add them to the bracket and the payroll tax and the true marginal rate for some low-income workers can exceed what a high earner faces. That is not a rumor, it is arithmetic, and we work an example in benefit cliffs can make a raise cost a single mother real money. If you want to know who is carrying the load in the first place, see Who actually pays the federal income tax, read straight from the IRS tables.

I bring it up to make a point about fairness that conservatives are uniquely placed to make. We are told we do not care about working families. Then explain to me which party's tax and benefit system punishes a single mother of two for taking a promotion. Whatever the intentions, the effect is that the system charges the climber more than the one who stays put. We should say so, loudly, and offer a design that does not.

A Founding anchor

Hamilton argued in Federalist No. 21 that taxes on consumption have a particular virtue: the citizen has some control over how much of them he pays, by his own choice of what to buy. That is a paraphrase, so read the essay for his exact words. The idea underneath it is old and sound: a tax is least harmful where the citizen can see it, understand it and respond to it. A system in which almost nobody knows their marginal rate, because the schedule is stacked with credits and phase-outs and an employer's withholding system hides the whole thing, is the opposite of a transparent tax. A citizen cannot hold his government to account for a price he cannot see.

There is a constitutional history here too. The Sixteenth Amendment says Congress may lay and collect taxes on incomes "from whatever source derived." It says nothing about rates. Congress has always chosen them, and so voters have always been able to change them. That makes the marginal rate a political choice, which is how it should be. The mistake is treating it as a law of nature.

What a sane system would do

Short version, because you did not come for a seminar.

• Publish the combined marginal rate. A one-page chart showing federal, payroll, state and benefit phase-out rates at each income level. Taxpayers have a right to know the price of the next dollar.
• Smooth the phase-outs. A benefit that falls by a dime for every extra dollar earned is a 10 percent tax. A benefit that disappears all at once is something much worse.
• Prefer fewer, flatter rates. Not because any particular flat rate is sacred, but because the fewer the steps, the fewer the places where the next dollar is penalized.
• Do not confuse a cut in the marginal rate with a gift to the rich. The person who works the extra shift, expands the shop or takes the second job is the one who gets the benefit of a lower rate on the next dollar. That person is exactly who a growing economy needs.

Sit down and do your own math

You can do the exercise at home. Take last year's return. Find your taxable income and your bracket. The Internal Revenue Service publishes the tax rate schedules each year, and the instructions to Form 1040 print the tax table. Add 7.65 percent if you are an employee, or 15.3 percent if you are self-employed, and add your state's top rate on your income level. That sum is the marginal rate on your next dollar. Now ask whether you would work harder or smarter if it were ten points lower. If the answer is yes, you have just discovered why this fight matters. If the answer is no, you have discovered that you are the kind of person for whom the income effect wins, and the economy will need your steadiness too.

The case in brief

1. Brackets tax slices of income, so moving into a higher bracket never lowers take-home pay on its own.
2. Decisions are made at the margin, so the marginal rate, not the average rate, shapes whether people work an extra shift, save, hire or expand.
3. Federal income tax, payroll tax, state tax and benefit phase-outs stack into one combined marginal rate, and that combined rate is the true price of the next dollar.
4. Responses are modest for many workers and larger for second earners, people near retirement and anyone who can change the form or timing of income.
5. Voters, not nature, set marginal rates under the Sixteenth Amendment, so they can demand a code that shows the price and keeps it low.

What you can do: take your last Form 1040, add your payroll and state rates to your federal bracket, and write down your combined marginal rate before the next debate about taxes.