There's a book out recently by a “financial strategist” whose background is in insurance that advocates trying to avoid paying any taxes in retirement. This is actually possible to do. You basically only have enough money in traditional IRAs/401(k)s that it can be withdrawn in the 0% tax bracket (up to around $20K a year this year for a married couple taking the standard deduction). With little other income, Social Security is tax-free. Withdrawals from Roth IRAs are, of course, tax free.If you're in the 15% bracket or lower, long term capital gains and dividends from taxable accounts are also tax-free. Muni bond yields are also tax-free. None of this is earned income, so you don't pay any payroll taxes on any of it.Those who push strategies like this also tend to sell cash value life insurance, so they like to point out that loans from the cash value of your life insurance policies are also tax-free (although they don't mention as often that the loans aren't interest free.) So the strategy is minimal tax-deferred money, very tax-efficient taxable accounts, cash value life insurance, and a lot of Roth money through contributions to Roth IRAs, Roth 401Ks, and Roth conversions. I think the strategy is ill-advised for several reasons.
Tax Fear Mongering
The first problem I have with people advocating for this is that they do a lot of fear mongering. They like to say things like “Tax rates have never been this low so they're sure to go up in the future.” Well, tax rates aren't actually the lowest they have ever been. You may have noticed the increase in tax rates the last 5 years or so under the Obama Administration, but if you haven't, check out this link. The chart demonstrates that our lowest tax bracket has been both lower and higher than it currently is, that our highest tax bracket has been both lower and higher than it currently is, and that our average effective tax rate has been both lower and higher than it currently is (and remarkably stable, to boot). In addition, capital gains/dividend tax rates have been both higher and lower than they are now. There is absolutely nothing that is particularly different about tax rates now vs what they have been in the past. Tax rates do change with the political winds, but if all you're looking at is the historical data, it would seem just as likely that tax rates are going to go down as up. That argument is pretty easy to refute.
The next argument advocates of this strategy generally pull out is that “since our US debt is out of control and the Fed is printing money like crazy then tax rates are sure to go up in the future.” Several articles/books I've seen suggest tax rates are likely to double. While I won't deny that it is quite possible that tax rates will go up in the future, I don't buy that our national debt is particularly out of control by historical standards. Take a look.
This chart from The Atlantic shows that while our debt as a percentage of GDP is high, it certainly isn't anything particularly different from what we've had in the past. It was a whole lot higher in the 1940s, and we saw plenty of prosperity in the 1950s and 1960s (and if you go back and look at the other graph, the effective tax rate didn't go up during those decades.) I suspect that most people making this argument are either uninformed about history, or selling something (like books, ads, or more likely, whole life insurance.)
Lack of Understanding of Filling the Low Brackets
There are a lot of people who don't get this. It turns out that maximum tax rates can go way up between when you contribute to a 401K and when you withdraw your money and you can still come out ahead. Not only do you get decades of tax-protected growth, but you are also likely to have far less taxable income in retirement than when you are working. For example, you might have enough income while working to be in the 6th bracket, but since you need/have far less income in retirement, you might only be in the 2nd or 3rd bracket. So even if each bracket went up 5% or 10%, you STILL have a lower marginal rate in retirement.
You also may get to withdraw a significant amount of that income at less than your marginal rate. A typical doctor might save money at 33% by contributing to his 401K, and then withdraw a good chunk of it at 0%, 10%, and 15%, such that his effective rate on withdrawing is 10-20%. Obviously, saving taxes at 33% and paying them at 15% is a winning strategy. Who benefits from talking you out of maximizing your 401K contributions? Those who are selling an “alternative retirement account” i.e. cash value life insurance. I don't know if insurance salesmen are ignorant or conniving, but either way you probably don't want their advice on this question.
Lack of Understanding of the Cost of Roth Contributions/Conversions
Regular readers know I am a huge fan of tax diversification in retirement. By having money in tax-deferred accounts, Roth accounts, and taxable accounts, you can minimize your tax bill in retirement, which, all things being equal, is a good thing. I contribute to Backdoor Roth IRAs each year and when I was in the lowest tax brackets (residency and military service) I preferentially put money into Roth accounts. I may also do some Roth conversions in low income years and early retirement years.
Mixing Insurance and Investing
Perhaps the biggest reason I dislike this idea of going for a zero percent tax bracket in retirement is it causes people to “invest” in cash value life insurance policies that they wouldn't otherwise buy. Remember that I said that all things being equal, a low tax rate in retirement is great. However, all things aren't equal if you're paying taxes at high marginal tax rates in your peak earnings when you don't have to and they certainly aren't equal if you're earning the low returns available in whole life insurance instead of the higher returns available with more traditional investments like stocks, bonds, and real estate. Remember you buy life insurance with after-tax dollars. So you can put $17,500 into your 401K, or you can pay a life insurance premium of $11,725. If the 401K investment grows at 8% and then is withdrawn at a 15% tax rate, and the life insurance cash value grows at 4% and then borrowed tax-free (but not interest free) 30 years later, the difference is $149,682 vs $38,029. Which would you rather have? I don't particularly think that cash value insurance compares favorably with a taxable account, but I can understand why some conservative, highly taxed investors might find it attractive. However, when you compare life insurance to a 401K or Roth IRA, the insurance nearly always comes out looking terrible.
Trying to get into the 0% tax bracket in retirement eliminates the benefits of spreading your income out over many years (and thus fully utilizing the low, non-zero brackets). It can also cause you to make mistakes in the Roth vs Tax-deferred decision and in the Investments vs Insurance (masquerading as an investment) decision. The goal isn't to pay as little in taxes as possible in retirement. The goal is to have as much after-tax, after-expense money to spend as possible in retirement, at least when adjusted for risk taken. Don't fall for the Zero Percent Bracket argument made by insurance agents when selling their wares. Tax diversification is good, but only at the right price. Going for the Zero Percent Bracket will probably cause you to pay too high a price.
What do you think? Do you plan to pay zero taxes in retirement? Why or why not? Have you heard this argument from insurance agents? Sound off below!