One of the most frustrating things that comes up almost immediately when you start talking to people about money is the way people treat certain purely regulatory or legislative figures as having some kind of moral or persuasive value.
The example that has driven me crazy for years is the advice to “contribute enough to your 401(k) to maximize your employer match.” In most cases, that works out to something in the range of 3-6% of gross income. But what is so special about your employer’s matching contribution policy? Even if you believed an employer was acting in completely good faith to set their matching policy purely for the benefit of employees, you’d still want to know what value they were optimizing for. An employer that matches 100% of the first 3% and one that matches 50% of the first 6% of gross income may both be acting purely for the benefit of their employees (i.e. they’re not trying to minimize the cost of the program), while at the same time the 100% match might be optimized for lower-paid employees who can afford to contribute less, and are likely to otherwise rely on Social Security, while the 50% match is better for higher-paid employees who can afford to contribute more and whose Social Security old age benefit replaces a lower portion of their lifetime income.
In other words, “maximize your employer match” is only good advice if you fall in the group the program was optimized for! Otherwise you run the risk of contributing too much (and having to borrow money at high interest rates to meet living expenses) or too little (and seeing a sharp drop in income in retirement).
Once you know what to look for, you see these problems all over.
Required Minimum Distributions and the 5% Rule
There are two things it is absolutely essential to know about required minimum distributions from pre-tax retirement accounts, like traditional IRA’s, 401(k)’s, and 403(b)’s. First, tax rates on retirees are laughably low. Second, this is the first time in their lives wealthy retirees have ever been forced to pay taxes, and they hate it.
A 59 1/2-year-old retiree beginning to make penalty-free withdrawals from a traditional IRA as their only source of income (for example to take advantage of the Social Security Magic Trick), can withdraw $53,075 (filing singly) or $106,150 (filing jointly) and pay an average federal income tax rate of 8.8% on contributions, interest, and capital gains that have never been taxed (this is the average of a 0% rate on the first $12,400 or $24,800, a 10% rate on the next $9,875 or $19,750, and a 12% rate on the remaining $30,800 or $61,600).
After age 70 the situation becomes slightly more complicated due to the taxability of Social Security benefits for very-high-income beneficiaries, but remember the maximum portion of your Social Security old age benefits that is ever taxed is 85%, so Social Security benefits are never taxed more heavily than pre-tax retirement distributions.
What makes required minimum distributions so painful is that it’s the first time in their lives wealthy retirees don’t get to choose whether or not to pay taxes. If you spent your whole life making pre-tax contributions and retired a millionaire, you might consider yourself lucky. But in reality, that’s not how it works: folks who spend their whole lives avoiding taxes don’t enter retirement relieved that they won the game of life. They enter retirement furious that there’s nothing their accountants can do to keep them from paying taxes on their fortune.
Of course, I’m exaggerating in one sense: accountants absolutely do have ways of avoiding taxes in retirement, most popularly through “qualified charitable distributions,” which allow distributions from pre-tax accounts to “skip” the retiree’s tax return and thus potentially avoid triggering Social Security benefit taxability.
My point is much simpler: if you have a high balance in a pre-tax retirement account, and are charitably inclined, why would you allow the IRS to determine how much you contribute from the account in any given year? This is a classic example of a regulatory minimum being transformed into an anchor: at age 72 the IRS says you’re required to distribute at least 3.9% of your balance, so you distribute exactly 3.9% of your balance. But what does the IRS know about your financial situation or the needs of the charitable causes you care about?
A related problem manifests in the rule that most private foundations and endowments are required to distribute 5% of their assets each year. There are, again, lots of ways to game this rule, but regardless of your foundation’s strategy or goals, the 5% rule is almost universally treated as an anchor: “the IRS says we have to distribute at least 5% of the fund each year, so we will distribute exactly 5% of the fund each year.”
Phase-ins, phase-outs, and income anchors
What makes these policies particularly strange is when the threshold dollar values are assigned explicitly moral or ethical values. For example, line 10b of the year 2020 Form 1040 allows you to deduct “Charitable contributions if you take the standard deduction” up to $300. But what’s so special about $300? Is that the right amount to contribute? Is it too high, or too low? How was the number arrived at?
Likewise, in the 2020 tax year, the earned income credit phases in over the first $7,000 in adjusted gross income for single childless filers, and phases out between $8,800 and $15,800 in adjusted gross income. The tax preference for capital gains over earned income is strange enough, but in the case of the EIC you have the genuinely bizarre situation where people who work pay almost twice the marginal tax rate on their income even than people who realize short-term capital gains or receive interest income: between $12,400 and $15,800 in AGI, $100 in increased earnings results in an additional $10 tax liability and an $8 reduction in EIC, for a marginal tax rate of 18%, while an additional $100 in unearned income is only taxed at 10%.
You might call this a statistical artifact, and you’d be right. After all, the amounts of money involved here are so low you’re unlikely to find many people who stop working after earning $8,800 each year, simply because there are few parts of the country where $8,800 is enough to live on. The more realistic case is people stop reporting their income, or move from formal work to informal work, after reaching the EIC maximum.
If you think comparing unearned to earned income is unfair, then let’s make a more straightforward comparison, between a married couple with and without children. When a childless married couple goes from earning $25,200 to $25,300, they owe an additional $10 in taxes (their EIC having long since phased out). When the married couple with a child earns the same additional $100, the $17 reduction in the EIC represents a 17% marginal tax rate (the non-refundable portion of the Child Tax Credit would eliminate the additional $10 in income tax liability). Between $30,800 and $30,900 in earned income, the marginal rate jumps to 26%, as they exhaust their non-refundable CTC and owe $10 in income tax while losing $16 in EIC, while the childless couple pays just 10% on their additional income.
This example has none of the “economic” nuances of the capital gains example. You may really deeply believe that capital income needs to be incentivized through the tax code to encourage savings, to encourage investment, to Finance The Industries Of The Future! We can have an honest disagreement about that (you’re wrong). But what possible reason could there be to tax each additional dollar earned by parents more heavily than that of non-parents?
In the case of childless single adults, the EIC phase-out results in underreported income, and I don’t think that’s a good result but it’s not exactly a crisis simply because the amounts involved are so low. In the case of the phase-out for parents, the under-reporting of income continues (one spouse may earn unreported income while the other works in the formal economy), and it also carries the real risk of serving as an anchor in precisely the same way the charitable contribution, required minimum distribution, and 5% rule do: when you tell people “you can claim the maximum credit when your income is up to $25,200,” at least some of them are going to seek to earn $25,200.
This argument is sometimes made in the context of explicitly pro-natalist politics (“we should pay people to have children to preserve our civilization”), but I don’t care about that. However much you think the EIC should be (double it! halve it! eliminate it!), however much you think the CTC should be, whatever you think the income tax brackets should be (flat tax! confiscatory tax!), everyone should be able to agree that parents shouldn’t be paying higher marginal tax rates than non-parents who are, in every other way, identical.
For good anchors, against bad anchors
One thing I hope comes across from this post is that anchoring is a totally normal and inevitable part of policy-making. Anchors will arise every time you write a number into a law. If you provide free public education from age 4 to age 18, most people will enroll at age 4 and most people will finish at age 18. That’s partly because people like free child care, partly because people like free education, but it’s mainly because those are the numbers in the law. They’re normal, and since most people are normal, they do the normal thing.
What we should strive to do is write good numbers into the laws, conscious of the effect they will have on society’s anchors. Most of the anchors I described were written for arcane budgetary reasons or administrative reasons that make no sense today, and could be easily eliminated.
For example, the government doesn’t actually care how much money people spend in retirement — RMD’s are required in order to trigger the taxation of as-yet-untaxed balances. That may have made sense when IRA’s were created in 1974, but there’s no reason today the government should be in any hurry. If people don’t want to spend their money in retirement, let them keep it, then levy a flat 40% or 50% tax on balances at death. What’s the rush?
Likewise, eliminating the phase-out of the earned income and child tax credits would be expensive. Fortunately, both programs are already administered through the tax code, where everybody writes down their income and pays taxes on it. “Paying for” the elimination of phase-outs is as simple as adjusting the tax brackets so the amount of money raised equaled the additional expense. There are huge banks of computers in Washington dedicated to making precisely such calculations.
Finally, this is not a way of “de-politicizing” issues. In some cases, it will re-politicize issues: if non-parents aren’t willing to see their marginal tax rates rise in order to pay for a uniform EIC and CTC, then they should say so and make the case so we can decide who’s right, instead of hiding under the bureaucracy of the welfare state.