# What could account for me owing exactly \$6000 in taxes?

In preparing my taxes, my on-line tax software says I owe exactly \$6000.

My wife and I make a combined income of around \$140,000. We work for the same employer. On her W4, we didn't adjust taxes for children, but on my W4, declared our 2 children, per the IRS's W4 instructions.

I find it strange that the tax calculator is calculating that I owe precisely \$6000. That is such a round number suggests to me some error somewhere.

Was there a law that changed that would account for this difference between my employer's paycheck calculation and my software? For instance, maybe before I could get \$3000/per child and that is no longer the case?

• “That is such a round number.” Yeah, our pattern-matching brains don’t expect to see patterns. It’s just that a number ending with 000 has the same 0.1% chance of happening as one that ends with 001. Or 666. You’d probably notice owing \$6666, too. Commented Feb 11 at 4:22
• Increase your income by 50 dollars as an experiment and see what the software gives. Or 1 dollar. Commented Feb 11 at 9:34
• I am reminded of this Richard Feynman quote. "You know, the most amazing thing happened to me tonight. I was coming here, on the way to the lecture, and I came in through the parking lot. And you won’t believe what happened. I saw a car with the license plate ARW 357. Can you imagine? Of all the millions of license plates in the state, what was the chance that I would see that particular one tonight? Amazing!"
– d_b
Commented Feb 11 at 19:26
• @RonJohn : and out of 330 million people, over 30000 would end up with that by chance. Even if we only include those who have to fill taxes, it would be well over 10000 people every year who end up with that by chance.
– vsz
Commented Feb 12 at 6:00
• There are many many "fines" (in the broad sense) the US / IRS applies, and they are all round numbers. As Nugg says, check child credit. Just another one is that bullshit that was applied for some years where in short if you didn't have health insurance you had to pay a "fine" of I believe it was \$2000 per person. Commented Feb 12 at 16:06

Humans tend to see patterns when there aren't any.

US (IRS) income taxes are rounded to the nearest dollar. Which means that:

• 10% of all taxpayers will receive a refund or have a tax bill to pay that ends in a 0.
• 1% of all taxpayers will receive a refund or have a tax bill to pay that ends in a double-0 - i.e., a multiple of \$100.
• 0.1% (1/10 of 1%, 1 in 1,000) of all taxpayers will receive a refund or have a tax bill to pay that ends in a triple-0 - i.e., a multiple of \$1,000.

A quick Google search finds an IRS report that there were over 162 million returns filed in 2023. Which means approximately 162,000 returns have a multiple of \$1,000 refund or tax bill. That's a lot of people, and you are apparently one of them.

• That assume randomness with normal distribution, which I find questionable. Neither the rules nor people income are truly random. Commented Feb 12 at 2:13
• Last digits have more or less flat distribution. First digits are pretty much known to favor 1s. Commented Feb 12 at 2:16
• @AndrewSavinykh: Firstly, the answer here is not focusing on the fairness of the distribution, just that OP's outcome is not unlikely by any stretch. Secondly, there are usually too many modifications/calculations happening for a bias in salary to reflect one-to-one to a bias in tax form result. For sufficiently large numbers, the trailing digits can effectively be considered to be random. To put it differently, Benford's law does not apply to trailing digits, only leading digits. Commented Feb 12 at 2:38
• @Fattie The IRS allows - and most tax preparation programs implement - rounding to the nearest dollar. As long as it is done up/down 0.50 - i.e., 0.50 - 0.99 -> 1.00, 0.01 - 0.49 -> 0.00 and done consistently (i.e., not just when you think it will favor the taxpayer) it is perfectly fine. Commented Feb 12 at 16:06
• @Fattie As I say, since the IRS is fine with rounding, most tax programs round. They don't include cents in any of the final numbers (refund or payment). Commented Feb 12 at 16:09

Pure coincidence. The likelihood of owing exactly \$0 at the end of the year is extremely small - people almost always owe something or get a refund. Is it possible that there was an error in your return? Absolutely, but the round number \$6,000 has nothing to do with that.

Tax calculations are complex, and without tax prep software it's simply impossible to calculate your tax liability with accuracy unless you have an extremely simple situation. Here are some common reasons why I've seen large balances owed, but there are many more possible reasons:

• Incorrect W-4
• Underwithholding on bonus income, including stock option exercises and RSU vesting
• Realized capital gains and other investment income
• No longer qualifying for tax credits such as the Child Tax Credit (must be under 17 years old to qualify)
• Self-employment income
• Other income (e.g. rental income)
• Receiving unemployment payments

There was a temporary expansion to the child tax credit, but that affected 2021 returns and went back down to "normal" for 2022 and 2023.

The first question I always ask when someone has a big underpayment is: what changed compared to last year? Did your withholdings, income, deductions and credits, or tax filing status change? Is the amount you owe substantially different from prior years?

• The question is not why the amount owed is so large; the question is why it ends in so many zeroes. Commented Feb 11 at 15:18
• the chance of owing or getting back exactly 0 isn't that small, considering there is a pretty high cut off below which nothing is owed or returned. Commented Feb 11 at 15:43
• One year, the amount of my State refund was the exact amount of the check I had to write the IRS. My accountant suggested playing those numbers on a lottery ticket. Commented Feb 11 at 19:00
• Interestingly enough, the first occurrence of a pure thousands (ending in 000) in the 2023 tax tables is for line 15 values between \$75,850 and \$75,900 when the tax is \$12,000 (single and married filing separate). The next is for \$83150-\$83200 for head of household. And no other 000 value in the table up to \$100000 (so 3 total mentions, 8248 tax entries in the table - less than 1 in 1000). The next point is that the tax withheld impacts the final tax owed, so it is not just a distribution in the tax tables. (And above \$100000 the numbers don't come out of the table.) Commented Feb 12 at 22:17

Based on the numbers you included above (\$140k total before tax income for you and your spouse), and two children under 18, you should expect to owe about \$11k-\$12k in taxes if you have no other deductions or income, based on the IRS calculator. Check your withholding against that; you can in fact use that same calculator and include exact details including withholding, and see if it's correct. This is just an estimate - your specific circumstances vary a lot - but it is probably a good ballpark number. If you're way above this, it may indicate an issue; if you're withholding way below this, then also that may be an issue.

If you paid \$6k in taxes during the year, you probably owe about \$6k. Don't worry about the "exact" part; nothing here will ever be that exact (unfortunately). The child tax credit is \$2000 per child (source), so it doesn't explain the entire difference. If your employer is only withholding \$6000 between the two of you, then consider asking them to increase that withholding.

This assumes you have only the two jobs and nothing else. If there are other sources of income, then you may need to consider those in your withholding, or paying estimated taxes throughout the year.

Statistics can be handled very naively by folks who don't make sports book software.

Yes, nominally there's a 1:100,000 chance of getting 000:00 on the end. (Or 1:1,000 if you forget cents.)

So that applies to the pool of IRS returns.

However.

Every return is subject to a number of positive or negative amounts by the regulations. There are probably on the order of ten of these. All of those are round numbers. Every return is subject to these. This completely changes the naive bookmaking approach.

I am reminded of this Richard Feynman quote. "You know, the most amazing thing happened to me tonight. I was coming here, on the way to the lecture, and I came in through the parking lot. And you won’t believe what happened. I saw a car with the license plate ARW 357. Can you imagine? Of all the millions of license plates in the state, what was the chance that I would see that particular one tonight? Amazing!"

Whilst that is a wonderful Feymnanism, if Feynman was here he'd instantly point out the mistake.

If (particularly if I'm in Kansas) I saw a plate "CHIEFS15" or "PAT 015" it would be very naive to then say "As in Feynman's great explanation, humans naively see patterns where none exist."

"PAT 015" could be a sheer coincidence (and you'd be correct to use "1:1,000" type logic there) but there are other gross and blatant factors involved, which statistically drastically outweigh that approach.

It's almost certainly the case that

• some sort of levy/credit is involved

• very simply OP or the software made a digit mistake somewhere