FREE NEWSLETTER
THE WIDOW TAX is sold to the wrong households. It gets pitched to affluent couples as the reason to convert to a Roth or buy life insurance. The pitch says that when one spouse dies, the survivor files single, lands in a higher bracket, and gets clobbered. I ran the numbers for three couples at three incomes, and at the comfortable end the widow tax often costs nothing, or even less than nothing.
The real cost lands lower down, on households nobody is selling anything to. What follows uses 2026 federal figures to show what the widow tax actually does in dollars, not in scary percentage points. The pattern runs opposite to the marketing. The higher your income, the smaller the hit, and the lower you go, the more it bites.
Three Things Move
When the first spouse dies, three things change, and they get bundled under one frightening label. One Social Security check stops, and a pension may shrink or end. That is lost income, and it is usually the largest of the three. It is not a tax.
Spending shifts as well. The survivor pays one Medicare premium instead of two, but the mortgage still comes out of the same account. The third change is the only one the tax code causes: narrower single brackets, a smaller standard deduction and lower thresholds for the Medicare surcharge. The widow tax is that third piece alone, and the three households below measure how big it really is.
The Affluent Couple
Both spouses are over 65. Their income is about $360,000: $80,000 from Social Security, $210,000 from a pension and required minimum distributions, plus $30,000 in qualified dividends and $40,000 in long-term gains. They sit in the 24% bracket, and both already pay the Medicare surcharge.
One spouse dies. The survivor keeps the larger Social Security check and most of the other income, landing at roughly $308,000. As a couple, they paid about $53,900 in federal income tax and a Medicare surcharge of about $9,240. As a single filer, the survivor pays about $55,000 in federal tax and a surcharge of about $6,355.
The effective rate rises from 15.7% to 18.7%, which is the headline people remember. But the federal tax barely moves, up only about $1,100, because the survivor has less income to tax. The Medicare surcharge actually falls about $2,900, because two enrollees in the couple’s joint tier cost more than one survivor a tier above. Net it all out, and the survivor pays about $1,800 a year less than the couple did.
The rate went up, and the dollars went down. At the high end, the widow tax measured in money hands back a small refund. This is why the Roth conversion pitch aimed at comfortable couples misfires. If converting to Roth would cut the survivor’s future taxable income without cutting the couple’s lifestyle, the converted money was surplus.
Surplus is money the survivor never needed to replace. Edward McQuarrie, a finance professor emeritus at Santa Clara University, made a version of this argument in a 2023 paper. He found the dollar hit inconsequential for affluent couples, and located any real cost lower down, where the Social Security torpedo bites.
Where It Actually Bites
Drop down to a couple with $180,000 of income: $60,000 from Social Security and $120,000 from a pension and required minimum distributions, all ordinary income. Both are over 65. The couple files jointly at $180,000; the survivor files single at about $150,000. As a couple, they paid about $17,148 in federal tax and no Medicare surcharge.
The survivor pays about $22,737 in federal tax and a new surcharge of about $2,885. The federal tax rises about $5,600, and a surprising chunk of that comes from the collapse of the new $6,000 senior deduction, which phases out against a lower income threshold for singles. Then the survivor crosses into a Medicare surcharge tier the couple never paid. The total cost of widowhood here is roughly $8,500 a year, set against a $30,000 income loss.
This is the one band where both the dollars and the rate move the wrong way. Now take a couple with $90,000 of income: $38,500 from Social Security, about the national average for two retired spouses, plus $51,500 from a pension and distributions. The survivor keeps the larger Social Security check and the full pension, landing at $73,500.
As a couple, they paid about $3,433 in federal tax, at an effective rate of 3.8%. The survivor pays about $5,278, at an effective rate of 7.2%. There’s no Medicare surcharge at this income, and there never will be. But the survivor’s effective rate nearly doubles, because more of the Social Security benefit becomes taxable, up to the 85% ceiling, when the single thresholds replace the joint ones.
The added tax is about $1,845, and the lost income is $16,500. Real money for a household least able to absorb it. It doubles the effective rate, and it hurts.
Read the three together and the widow tax points the wrong way from the marketing. At $360,000 it costs less than nothing, roughly a $1,800 annual saving. At $180,000 it costs about $8,500. At $90,000 it costs about $1,850, but the effective rate doubles.
The affluent couples being sold protection don’t need it. The middle couples who feel the sting are not being sold anything, and they’re the ones for whom a few thousand dollars a year actually constrains a life. The tax code has three different mechanisms, and which one finds you depends almost entirely on your income. At the top, the Medicare surcharge does the work, and it falls.
In the middle, the surcharge appears from zero. At the bottom, the Social Security torpedo raises the taxability of the benefit from 75% to 85%. The mechanism changes with the income, and so does the household’s ability to absorb the hit.
Plan Ahead
This is where the widow tax conversation usually stops, and where I think it should start. The moves most likely to leave a survivor better off are the ones you make years before, and they help whether or not the widow tax ever bites. Converting traditional IRA money to Roth over several years before retirement smooths your taxable income. It lowers the required minimum distributions that will later push a single filer into higher brackets.
It can also keep a survivor under a Medicare surcharge tier the couple never worried about. Drawing accounts in a sensible order, spending down the right buckets first, reduces the future tax base that a single bracket structure will tax more steeply. Managing required distributions as they grow, rather than letting them balloon off a rising balance, limits the single bracket exposure that builds across a long widowhood. A more tax-efficient bequest helps the people who inherit what is left.
None of that is widow rescue. It is good multi-year planning that happens to compound in the survivor’s favor. The widow tax is one input to that planning, not the reason for it. The reason is that the year your spouse dies is the worst possible year to be making financial decisions, and the more of those decisions you have already made, the fewer you hand to someone who is grieving.
You can do a rough version of this at your own kitchen table, and you should, ideally long before you need to. Estimate the survivor’s income after the smaller Social Security check and any pension change, then estimate the survivor’s spending from your actual budget, not a generic rule of thumb. Subtract. If reliable after-tax income still supports the life the survivor wants, with margin, the tax rate was never the thing to worry about.
Then project the survivor’s federal tax, the net investment income tax where it applies and the Medicare surcharge as a single filer, and compare it to the couple. Keep the tax separate from the lost income, so you can see what the tax code actually did. Sometimes, as the affluent couple shows, it does you a small favor.
Only then does a planning move earn a look, and only if you can answer what it costs today, what it might save later, who benefits and what has to come true for it to work. The widow tax is real, but it isn’t the catastrophe it’s sold as. At higher incomes it’s not a cost at all. Lower down it is a real cost, and the lower you go, the more it’s worth measuring, because the households it constrains have the least margin to spare.
The planning that matters most is the kind you do years ahead, smoothing income, holding down future distributions, managing the surcharge tiers and putting the estate in order. Do that, and you’ve done right by your survivor, for reasons that have little to do with fear and a great deal to do with care. You’ll have done it in the years when you still had the time, and the clarity, to do it well.
________________________________________________________________________________
John Urban is the founder of RetireSmartIRA, a retirement tax-planning app. Earlier, he founded GT Nexus, a supply-chain software company acquired by Infor in 2015. He lives in Northern California with his wife, Kathy, and enjoys time with family, travel, reading, Bay Area sports, and the occasional deep dive into the fine print of the tax code.
Surely the problem is ultimately in thinking of it as a “widow tax” at all? Isn’t it just the reversal of the extraordinary benefits that married couples have been lucky enough to enjoy for the duration of their partnership. Single people have never had the same perks (and likely many other things like higher per capita cost of housing, household bills etc etc)
For two earner couples our current tax system is fairly tax neutral – standard deduction and tax brackets are doubled up to the 35% tax bracket, at which point there is a marriage penalty starting about $768k in combined income.
For a single earner couple the system is beneficial.
This seems pretty fair and I wouldn’t consider it an extraordinary benefit as getting married shouldn’t come with a tax penalty.
The perks that you allude to are available to all who cohabitate – roommates, non-married significant others etc.
You are correct though – it is not so much a widow’s tax as a single person’s tax. Assuming the married couple were able to leverage their combined earnings into a larger net worth and retirement income than if they were single all their lives, then with proper planning they should be in good shape.
Thanks John for a great article. Very inciteful and counter to what many advocate.
I fall in the middle level. I was not smart enough to do Roth conversions before retiring or in my early retirement years. We had 90% of investments in TIRAs. Fortunately, I wised up and did enough Roth conversions to get our TIRA down to 12%, incurring IRMAA premiums and sometimes higher marginal tax rates.
Sadly, my wife passed last year. As a single taxpayer this year, I will pay the first tier of IRMAA. I will not complain but my SS and pension still cause me to pay the IRMAA premium. I use my QCDs to cover my RMDs, but there are no other options I know of to reduce the widow tax.
Even though I consider myself financially savvy, I did not appreciate the impact the “widow tax” has on the survivor. Hopefully, others can learn from your article.
John, what an excellent article!
This was so well written and so informative! And the comments it resulted in are amazing. It was so good to see that Rick Connor joined in too. I’ve always admired his writings and the writings of other long time contributors.
This is the kind of writing and commentary that I hunger for on Humble Dollar.
Thank you for all of the time and effort that you put into this article.
Lou,
Thank you. The comments are always the best part.
John
John, thanks for an excellent article, on a complex topic. You clearly show that running the numbers is key to understanding the impact to a surviving spouse. Your article also generated a number of insightful and useful comments. The post by achnk53 nine days ago inspired me to run some case studies and write up an article. I approached it a bit differently, so I think I will continue to work on it. Like a number of commenters, I have many years of experience with AARP’s TaxAide, and I resonate with their stories of working with lower income widows and widowers, trying to help them understand their new tax reality. I’ve also had similar experiences with divorced people. Many of the articles I’ve read about this focus on the year after the death of a spouse. I haven’t see anything that focuses on the continuing impacts, and how they may change. It’s not easy, too many variables, but it might be fun to try to look at it.
Rick,
Thanks, that means a lot given how much time you’ve spent in this exact material through TaxAide. Please do run with your own piece. I approached this as a snapshot comparison on purpose, so a different lens on the same problem is a good thing, not overlap to avoid.
Your last point is the one I keep turning over. Everything in my article treats the year after death as the whole story, but two mechanics keep pushing the picture past that year. RMDs force larger withdrawals every year the survivor ages, since the IRS divisor keeps shrinking, so the tax bill can climb for a decade even if the portfolio never grows. And IRMAA runs on a two-year MAGI lookback, so a new widow’s Medicare premium for the first year or two is still riding the couple’s old joint income, before the single-filer numbers show up. Year one and year three can look nothing alike, for reasons that have nothing to do with markets or spending.
I’m sure there’s a lot more to think about as you develop your article. I look forward to reading it when you’re done.
Best,
John
John, many thanks for a thorough analysis of different scenarios and an objective assessment on one of the several common tax-related myths.
A small observation about the middle couple, specifically about the new OBBBA additional deduction of up to $6000 per eligible person. The single filer loses this deduction completely and therefore there is an additional tax cost for the income.
I thought this deduction is temporary and will phase out in 3 years. Does your analysis assume that this would be extended further, potentially indefinitly? If not, then I think the cost of widowhood in this case will drop from ~$8500 to less than $6000.
Thanks again for the excellent piece.
Sanjib,
Yours is a great catch. My analysis assumed nothing about an extension. I ran 2026 law as it stands, a snapshot rather than a projection, and you correctly point to the limitation in that.
I reran the middle couple under post-2028 law rather than guess at the impact. You’re right about the direction, though the number lands a bit higher than your estimate: about $6,800, not quite under $6,000. The reason it doesn’t fall further is that the couple’s lost deduction sits in their 22% bracket while the survivor’s smaller lost deduction sits in her 24% bracket, so the rate difference eats into the savings the raw dollar amounts would suggest.
One piece is untouched either way: the $2,885 Medicare surcharge in that example. IRMAA runs on MAGI, not the deduction, so it doesn’t move when the provision sunsets.
Thank you for reading the article so closely. I always look forward to the additional insight that comes out in these comments!
Best,
John
Here is a link to Professor (emeritus) Edward F. McQuarrie’s December 2023 article titled “Widow Tax Hit Debunked” where he concluded “the problem (in thinking about the impact of the widow tax issue) is a myopic focus on tax rates to the exclusion of a sounder analysis that keeps both tax dollars paid and the postmortem dollar reduction in expenditure at the center of the frame.”
For our tax decisions, I like to focus on our current joint marginal tax rates rather than considering our joint current effective tax rates in tax actions I take, like Roth conversions, and also then consider the likely impact on future taxes on the survivor after either my wife or I die. As I cannot predict the future I often take a position in my tax decisions to do what is likely best for my spouse should I die a long time before she does.
I would add to the list of tax consideration matters to think about is that if the spouse who is the likely survivor widow(er) has health conditions that would then require paid assistance with activities of daily living that are potentially tax deductible expenses that could have a major future impact.
There is a recent Jason Zweig article in the WSJ in June 2026 titled “How to Build a Retirement That Actually Fits Your Life” about an upcoming book due out in March 2027 written by Professor McQuarrie and Dr. William Bernstein, to be titled “Retirement: How to Save Enough, Invest It Well, and Make Your Money Last” that I am looking forward to reading. I believe both gentlemen have been speakers at past Boglehead conferences.
Thanks for your article John.
Thanks for pulling that McQuarrie quote out. I’m a longtime admirer of his work, and that paper is what pushed me to run the dollars instead of the rates. It’s worth noting where his sentence indicts my own piece. I kept tax dollars at the center and left the postmortem reduction in expenditure as an aside rather than a number. Dan Smith made the same point elsewhere in this thread. Two readers landing on it in one morning tells me I underweighted it.
Your marginal versus effective distinction is the sharper version of what I was reaching for. Effective rate is a reporting statistic, useful for describing what happened. Marginal rate is the decision variable, the only one that tells you whether the next dollar converted is worth converting. My article argued against being frightened by a rising effective rate. You’re supplying the part that comes next: when you actually act, use the marginal.
The long-term care point is the best addition anyone has made here. Qualified long-term care costs are deductible medical expenses above the 7.5% AGI floor, so a survivor paying for help with activities of daily living can end up itemizing well past the standard deduction the couple used to take. That reverses the largest single piece of the harm I described, for the household least able to absorb it otherwise. I didn’t have it in the frame and I should have.
Your default of planning for whichever spouse is likely to be alone longest is a sound way to hold a question nobody can answer, and close to where I ended the piece. Thanks also for the Zweig article. McQuarrie and Bernstein together is reason enough to mark March 2027 down.
Great article John, and this comment is also one to all remember. “Effective rate is a reporting statistic, useful for describing what happened. Marginal rate is the decision variable, the only one that tells you whether the next dollar converted is worth converting.”
Great article, John! Seeing the examples and how each of the components affects the outcome really helps to see the actual effect. I will get busy doing some calculations following the procedure you outlined.
Thanks Dave, I really appreciate your comment.
This was very informative and I appreciate the work you put into this analysis.
A few years ago, I started a spreadsheet to compute a surviving spouse’s tax burden, in parallel with our current MFJ status. We fall slightly below the middle category for income in retirement. For determining the impact of a surviving spouse, I had calculations for Medicare expenses and IRMAA (if any), percent of SS benefits taxable, and the loss/availability of 0% long-term capital gains bracket. I also estimated the reduction in expenses versus the loss of the lower SS benefit. A key difference for our situation is the use of SPIA annuity income to further reduce RMDs computed from the end-of-year T-IRA balance (using Secure Act 2.0 provisions).
When I retired, we used about one-third of our IRA portfolio to purchase (joint-survivor) single-premium immediate annuities (SPIAs) and use the other two-thirds to invest in the market (70/30 asset allocation). We had undergone a six-year Roth conversion plan (2017 – 2022) that reduced our traditional IRA holdings from 100% to 60% (with the goal of having a 40% reduction in estimated RMDs). In parallel, we kept the effectively conversion tax rate at roughly 23% (half at 22% and half at 24%). We paid IRMAA to the first tier (which is not too bad). Because of the annuity income and SS benefits, we were comfortable using our RMD primarily for discretionary expenses (mostly travel) and can consider withdrawals to be variable (rather than required for daily expenses).
Due to the conversions, when I started RMDs, the withdrawals only applied to roughly 50% of our portfolio (the other half being in Roth). With the advent of Secure Act 2.0, which allows the excess annuity payout to count against the RMD as computed from the T-IRA end-of-year balance, our effective RMD dropped from roughly 4.38% to 1.78% (at age 76/79, weighted by account size). Additionally, since roughly half the portfolio is in Roth, the true required minimum distribution from the portfolio is currently 1%. Each year the RMD percentage increases but the fair market value (FMV) of the SPIA contracts also decreases. This makes the effective (excess) payout increase as well, which counteracts the increase in RMD percentage. By having a “low” RMD, it made making adjustments to our income to maximize tax savings, such as keeping our MAGI <$150K to max out the $12K senior deduction, for example.
The net effect for our situation is that the surviving spouse avoids any IRMAA surcharges and as well as any significant tax dollar increases. This assumes only withdrawing the RMD and using Roth assets for additional income. In this case, the lower income spouse took SS early while the higher income spouse claimed at age 70. Consequently, the loss (of the lower income) turned out to be less than the reduction in expenses due just for major identifiable expenses (insurance premiums, medical out-of-pocket, dental, eyeglasses, hearing aids, automotive expenses, retail, etc.).
Snak123 – You clearly have and use strong planning and analytic skills. Thank you for sharing your process and insight!!!
John, thanks for this most enlightening contribution. It sure is a great argument for a Roth 401k for working people earning modest incomes. Less tax is always better than more tax.
But all is not necessarily lost for we 70-something seniors with traditional IRA balances. Income taxes will surely rise for many of the newly single, but they will be offset for many, as some living expenses decrease. For example, one less medicare and supplement premium, one less car, and lower food expense.
Thanks Dan. The expense side is the leg most of these analyses skip, and your examples are the right ones: one Medicare and supplement premium, one car, a grocery bill that drops even if it doesn’t halve. That offset is real, and it’s why I wanted to keep lost income, spending and tax on separate lines rather than bundled under one label.
Helpful article. Of course, folks can quibble about the assumptions for the couples. For example, my pension will reduce to 50% benefit for my wife when I die. That lowers the income component significantly, but will also reduce the tax obligation – there’s less to tax.
Perhaps it’s more appropriate to discuss the widow/widower’s financial changes, not focus on the widow/widower tax.
Thanks. That’s the takeaway I most hoped would land, and you’ve said it more plainly than I did. The widow tax is rarely the problem. The lost income is the problem, and the tax code is a secondary effect that sometimes runs the other way.
Your 50% survivor pension is the clearest version of it. Half a pension gone is a far bigger number than any bracket change, and with less income to tax, the bill itself can fall even as the rate on paper rises. That’s my affluent couple arriving by a different route.
On the framing, you’re right and I hedged. Widow tax is the phrase in circulation, so I used it in order to argue with it. Financial change is the better label.
I am a volunteer tax preparer with AARP. A couple years ago I wrote about this topic for my low income clients: Taxing Situations – HumbleDollar.
The reason many of them turn to AARP is their lack of understanding of the tax code and the last thing they are thinking about when a spouse dies is the need to increase their withholding.
Howard, Your post from 2025 is really great, the real-world examples provide real insight. Your experience with recently widowed taxpayers particularly resonated with me. One thing that’s striking is that the higher tax bill rarely comes as a surprise because of a tax law change, it comes because almost no one realizes how much changes when the surviving spouse moves from the MFJ brackets and standard deduction to filing as Single. By the time they discover it, it’s often at tax-preparation time, when there’s little they can do except write a check and adjust withholding for the following year. Your counseling those taxpayers a year in advance may be one of the most valuable services you provide.
You idea of the middle couple probably is the middle for people who read financial advice on the internet, but is not the actual middle. Most retirees, married or single, would be thrilled to have an income of $180K or $150K.
You’re right, and it’s a fair correction. “Middle” there meant the middle of my three examples, not the middle of the country. Median household income for Americans 65 and older is somewhere in the mid-$50,000s, so even my $90,000 couple sits well above typical.
The households you’re describing are the ones I’d most like to have written about, and didn’t. At those incomes federal tax is often little or nothing to begin with, so the widow tax barely registers in dollars. What lands is the missing Social Security check. That’s a different article, and probably a very useful one. Also, Take a look at Howard Rohlender’s past Taxing Situations – HumbleDollar.
Yes, but when you get to $300K or $400K, you start to pay considerable tax, particularly if you are single.
John, that is very informative. Thanks for sharing your perspective on this. I was one of those not looking at this correctly but now I am.
Cheers
Thank you Chris&Steve!
Thank you for this article. Lots of learning happening this morning. Mostly stop preaching to my fat-walleted brother that he’s doing it all wrong. Hahaha.
Thanks Heidi – Your comment made me laugh out loud!