Parents, students, and policymakers are increasingly worried about the cost of college. Much of this anxiety is overplayed in the media. A college education remains, for most, a sound investment. Debt levels remain at manageable levels, especially for those with the highest amounts, since they also earn the most.
Richard V. Reeves
John C. and Nancy D. Whitehead Chair
Senior Fellow - Economic Studies
Senior Research Assistant
In addition to private investments in college education, there are considerable public investments, too, in the form of direct support for institutions, grants and subsidized loans to students, and tax expenditures intended to encourage saving towards the cost of college.
In this paper, we focus on the last of these, specifically on “529” college savings plans. These are tax-advantaged vehicles for saving towards educational spending. The main points of the paper are:
- Affluent families reap almost all of the benefits from the federal tax incentives. Currently, any capital gains or dividends produced by 529 savings plan investments are not taxed. This provision almost exclusively benefits the seven in ten families with 529 plans who have six-figure incomes, because lower-income families are not liable for taxes on capital gains and dividends. For example, married parents with two dependents pay capital gains or dividend taxes only if their combined earnings are over $100,000.
- We estimate the value of the federal tax breaks for married parents at various points in the income distribution with two dependents investing stock-based 529 college savings accounts with an average annual rate of return of 6 percent. We find that the only families seeing a federal subsidy from the capital gains tax exclusion of 529 savings plans are those earning six figures or more (given the standard deduction is $12,700, these families claim four dependent exemptions of $4,050 each, and 15 percent capital gains tax bracket starts at $75,900, that means that families with income less than $104,800 already may save tax free in stocks). The total tax break they receive ranges from about $7,076 on the low end to as much as $32,148 for those who can afford to save the most.
- 529 plans are growing in size and now hold around $275 billion. As 529s grow, so do the cost of associated tax advantages, which will cost the federal government almost $30 billion over the next decade.
- Most states offer additional incentives to save into 529 plans. In the 24 states that report spending on these tax breaks, the combined cost is around $265 million each year; the total for all states with tax benefits is almost certainly substantially higher (nine states do not report how much they are spending). Additionally, state income tax deductions are just as regressive as the federal benefit. In states that provide the necessary data (unfortunately, most do not), around three-quarters of the tax benefits go to households with annual incomes over $100,000.
- There is no evidence that the tax incentives attached to 529 plans boost savings rates significantly overall; savers simply switch over to the tax-advantaged vehicles. Furthermore, there is some suggestive evidence that the subsidies flowing into 529 plans may have been a contributory factor to cost inflation in higher education.
- There is a strong case for ending or limiting these tax expenditures. We outline options for reform, including the provision of matching funds for low-income families; elimination of the accelerated gift rule; and a rollback of state income tax breaks.
There are, of course, considerable political barriers to reform here, as recent history shows. In January 2015, President Obama attempted to redirect the tax expenditures flowing to upper middle class families through 529 plans into tax credits to help a broader range of American families. He was forced to retreat under pressure not only from Republicans, but from liberal members of his own party.
Obama was right, and they were wrong. But the proposal provoked the ire of what the Washington Post’s Paul Waldman described as “what may be the single most dangerous constituency to anger: the upper middle class—wealthy enough to have influence, and numerous enough to be a significant voting bloc.”1 (This class is the subject of my new book Dream Hoarders: How the American Upper Middle Class Is Leaving Everyone Else in the Dust, Why That Is a Problem, and What to Do About It.)
What are 529s and how do they work?
529 plans are state-operated savings accounts, named after Section 529 of the IRS Code that codified them in 1996. The original 1996 legislation deferred tax on undistributed earnings. The Economic Growth and Tax Relief Reconciliation Act of 2001, signed by George W. Bush, then made earnings growth completely tax-free (a provision that had been vetoed by President Clinton in the 1996 effort). The provision became permanent in 2006.
Since the mid-2000s, the growth in 529 plans has been considerable. By the end of 2016, total asset values in 529 plans reached $275 billion. But in the general population, 529 plans are little understood and little used; in one survey, 72 percent of respondents did not know what they were.
529 plans have several defining features:
- Funds must be used for qualified educational expenditures
- No income-eligibility threshold
- Waiver from the standard gift expensing rule allows families to ‘superfund’ accounts
- Grandparents (and in some cases others) are allowed to contribute
- Only a small fraction of 529 assets (5.64 percent) count in making financial aid calculations2
- Many states (33) offer income tax deductions for contributions
- No residency requirement (e.g. can live in one state and fund an account in another)
A few of these provisions are relatively unusual, so we provide a few more details here.
Qualified educational expenses
Among the expenses the IRS counts as qualified education spending are tuition and related fees, room and board, books and supplies, special services, and computers and related equipment. Room and board expenses are limited to the lesser of the room and board allowance in a school’s cost of attendance measure, or the actual amount paid to student housing. 529 funds cannot be used to pay for insurance, sports expenses or monthly health dues, electronics, smartphones and tablets, transportation and travel cost, and repayment of student loans.
Waiver on tax rules for gifts
A unique feature of 529 college savings plans is a waiver from the regular tax rules that apply to gifts. Parents can elect to combine five years’ worth of contributions (up to $140,000 for a jointly filing couple, per recipient) into a single year. The Obamas themselves took advantage of this provision, contributing $240,000 to their two daughters’ 529 plans in a single year. Higher upfront contributions obviously increase capital gains, and therefore the value of the capital gains tax break. The special rule for accelerated gifts solely benefits those who have already saved for college—it is simply a windfall for wealthy families, with no discernable policy justification.
Most states, and the District of Columbia, have their own 529 plans with varying contribution limits and restrictions. Maryland, for example, allows up to $2,500 (individual filer) or $5,000 (joint filing) in tax deductions per beneficiary per year and a total maximum contribution cap of $350,000. By saving $10,000 a year into 529 plans for two children, affluent married Maryland taxpayers can therefore cut $363 off their income tax bill.3
Table 1 summarizes 529 plan policies for each state. We report the following: (1) state tax benefits in the form of a credit or deduction; (2) residency restrictions or requirements; (3) any matching funding for low income families using 529 plans; and (4) the annual cost of the tax deductions and/or credits (for those states reporting it).
There is considerable variation between states. At one end of the spectrum, states like Colorado, New Mexico, and South Carolina have no annual limit on fully deductible contributions and allow a balance of up to $400,000 in each plan. Only one of these three states has a programs aimed at matching contributions for low- and middle-income families. A handful of states allow income tax deduction/credits for contributions to any state’s 529 plan, including Arizona, Kansas, Missouri, Montana, and Pennsylvania (which allows a deduction against state income tax of up to $28,000 a year).
At the other end of the spectrum, Nevada offers no income tax benefits for 529s, but does provide a grant of up to $300 annually for low- to middle-income families who save in such plans. The other states providing direct assistance to low income families are Arkansas, Colorado, Kansas, Louisiana, Maine, Maryland, New Jersey, North Dakota, Utah and Virginia.
1. 529s: A handout to the upper middle class
Who reaps the benefits of these tax breaks? The main beneficiaries are upper-middle-class families. The Government Accountability Office (GAO) found that in 2010, 47 percent of families with 529 plans had an annual income of over $150,000. (Note that only 11 percent of families with children made more than $150,000 that year.)4 The tax benefits associated with 529 plans also skew towards higher-income families, according to the GAO. The median tax savings for families making up to $100,000 was $561, compared to $1,958 for families making between $100,001 and $150,000, and $3,132 for families with incomes over $150,000.
2. Quantifying the 529 tax break
529 plans are relatively young, however, and both the costs to the federal government and the benefits to affluent families are likely to rise, assuming no change in policy. To estimate distributional impacts, we estimate the value of the federal subsidy for married parents with two children on different rungs of the income ladders. Specifically, we look at families living at the:
- Federal poverty line;
- National median income;
- Bottom of the top quintile (i.e., $120,000 annually);5 and
- Top of the distribution and maxing out 529 benefits.
Family at 100 percent of federal poverty line
In 2016, the federal poverty line for a family of four was $24,300 a year.6 Among the very small proportion (around 17 percent) of such families saving into 529 plans, the typical amount is just $495 a year, according to a survey conducted by Sallie Mae.7 Although in practice savings tend to increase with the child’s age, we assume a consistent annual contribution throughout the childhood (ages 0-18) of the beneficiaries. We also assume the investments are in a stock portfolio, with an annual return of 6 percent, a reasonable rate according to Vanguard data.8
Such a family could expect to see investment growth of about $7,315 by the time their children reach traditional college age. But if the family remains on a low income, they receive no benefit from the capital gains tax break, since they are not liable for the tax in any case. In fact, there may be disincentives to save in 529 plans, such as account minimums, plan fees, and interaction with aspects like tax credits, financial aid, and the EITC. For example, 529 college savings plan withdrawals and the American Opportunity Tax Credit (AOTC) cannot be used on the same qualified educational expense. The value of the AOTC is $2,500 annually for up to four years, which is just under three times what we assess to be the tax-free growth a low-income family could expect to see from a 529 investment.
Family making national median income
Median household income in the United States was $56,516 in 2015, according to the most recently available data from the Census Bureau.9 For our hypothetical family (married parents with two dependents) at this income level, we once again base our estimates on the Sallie Mae survey. The average annual contributions made by families making between $35,000-$100,000 who have 529 plans (about 30 percent) was around $1,592. After applying the same portfolio, annual return, and time assumptions as before, we find that this middle-class family could expect to see investment growth of about $23,500 by the time their children reach traditional college age. Once again, however, because of the capital gains taxation threshold, the forgone federal taxation on such growth is zero.
The “just upper-middle class” family
Next we turn to a family that above the threshold for entry to the top income quintile, with an income of $120,000.10 For this family, we assume annual contributions of $3,195 (consistent with Sallie Mae’s findings for those six-figure income families with 529s). Under the same assumptions as for the other two families, we find that this family would see investment growth of about $47,172. In this case, the capital gains exclusion (assuming a 15 percent rate) leads to a subsidy of about $7,076. To put these figures in context, the maximum annual Pell grant is $5,920 as of the 2017-18 award year.
The rich super-saver family
Lastly, we examine the benefits that accrue to a very affluent family that intentionally maximizes the tax advantages available to 529 holders. The Sallie Mae survey is of limited use here, since the highest income category was $100,000 or more. We have imagined the savings behavior of a married couple that in a typical year make around the 95th percentile of the income distribution (i.e., $215,000, with year-to-year fluctuations), with windfalls in between, and who are therefore able to “superfund” the 529 plans for each of their two children on their 5th and 10th birthdays, to the tune of $50,000 each time for each child (i.e., total contributions of $100,000 per child, adding up to $200,000).11 For the remaining years of the children’s lives, we assume that the family contributes at the same rate as other families with six figures who are saving ($3,195).
This kind of strategy is only available to those with very high incomes and/or wealth, at least in certain years. Very few families can afford to save hundreds of thousands of dollars in a single year. But for these rich, tax-efficient families, the benefits of 529 are very great. Following investment growth of $214,319, the tax break resulting from the interaction of the capital gains tax exemption and the “superfunding” waiver on gift rule amounts to $32,148.12
3. The cost of 529 plans is set to rise
As 529 plans mature and more families use them to fund college costs, the price tag to the U.S. Treasury will also rise, unless some reforms are undertaken. Over the next decade, the federal government is set to spend almost $30 billion on 529 tax expenditures, according to Treasury’s Office of Tax Analysis. Annual costs are projects to be just under $4 billion by 2026.
4. States are no better
The regressive nature of federal tax expenditures on 529 plans is clear, even under the most optimistic assumptions about savings levels. But what about state tax deductions? The story here is essentially the same: 529 plans are a handout to the affluent. There are data constraints here, since only 24 states report their tax expenditures on 529 plans. Combined total spending in those states is around $265 million each year:13
Even fewer states (just four) break down these expenditures by the reported income of the household claiming the deduction or credit. Again, six-figure households claim upwards of 75 percent of the state tax benefits.
5. Who really benefits from 529s: Parents or colleges?
A common defense of 529 tax advantages is that college is becoming an expensive proposition, even for relatively affluent families. As the liberal economist Jared Bernstein reflected on the 2015 reversal of the 529 reforms: “following that experience, and from the debate over free, or debt-free, college in the presidential campaigns, I’ve come to see the problem of college affordability as one that climbs very high up the income scale.”14
There is no denying that the price of a college education has increased, outpacing inflation for the past ten years. But rather than being the solution, regressive government subsidies could be part of the problem. Three main theories have been posited to explain rising prices:
It seems likely that all of these factors play a part, along with many others. Here, we focus on the last of the three: the contribution of public subsidies to tuition cost increases. According to the “Bennett hypothesis,” first introduced in a 1987 New York Times op-ed by then Secretary of Education William Bennett, subsidies for higher education lead to a reflexive rise in tuition.
The Bennett hypothesis has mostly been applied to financial aid subsidies for low-income students. While the literature is mixed, there is some suggestive research showing that it may hold in particular contexts, especially for private colleges. In one analysis, Larry Singell and Joe Stone found that the most generous form of federal subsidy, Pell grants, had no impact on fees at public universities but that at private universities, “increases in Pell grants appear to be matched nearly one for one by increases in list (and net) tuition.” A recent report from the New York Federal Reserve finds that each dollar spent on subsidized loans pushed up fees by 60 cents, and concluded that “the effect is most pronounced for more expensive degrees, those offered by private institutions, and for two-year or vocational programs.”
What about the 529 subsidy? It seems logical that government subsidies in the form of tax breaks for 529s might also drive up the very costs they are intended to help meet. After all, the children of parents with high incomes are the likeliest to attend and complete college. It seems at least possible that, at the margins, colleges might use the windfall provided by federal and state subsidization of 529s to increase ‘consumption’ spending in an effort to secure a competitive advantage: amenities and the creation of new services and programs, often requiring things like an expansion of administrative staff.
As far as we are aware, there is no research on the influence of 529 subsidies on college costs. But it is possible that by helping more affluent families pay for college, these plans increase the amount that they are willing to pay. In a new paper, “The Affordability Conundrum: Value, Price, and Choice in Higher Education,” Beth Akers, Kim Dancy, and Jason Delisle find that students from more affluent families (those most likely to benefit from 529 plans) will consider more expensive colleges than their less affluent peers. Full-time dependent students from households making more than $109,000 are, on average, willing to pay almost $8,000 more each year than their peers from households with incomes between $67,000 and $109,000.15 For a private (non-profit) college education, they are willing to pay $15,000 more. There is also some evidence for the persistence of college ‘taste preferences’ across generations,16 with the children of those who attended a certain kind of college (e.g., selective public) much more likely to follow suit than their peers, even controlling for other factors.
The cost of college has increased fastest at the types of institutions preferred by, and attended by, students from the households most likely to benefit from the 529 public subsidy. These institutions also typically have relatively low numbers of students on Pell grants. It is far too early to claim that 529 plans are a contributing factor behind rising college costs, but it is also too early to rule it out. We should be alert to the danger of a vicious circle here, as rising college costs increase political pressure to provide subsidies even to affluent families, which may in turn fuel price inflation.
6. 529s: Options for Reform
What is to be done? The case for the abolition of 529 plans (while honoring existing breaks) remains strong. The $30 billion that the federal government will spend over the next ten years could be better spent elsewhere. The same applies to the state-level tax expenditures. This is money that cash-strapped states could spend much more effectively on other programs. Here we propose four reforms: two aimed at federal policy, and two at state policies.
- Limit capital gains exclusion for higher-income families
One of the arguments used against President Obama’s 2015 reform was that it would hit “middle class” families. As we have shown in this paper, this is largely not true. Almost all the benefits go to those at the top. This concern could be met by retaining 529 benefits for less affluent families, but withdrawing them from higher-income households. As the new congress pursues tax reform and attempts to search for new revenues, one option is to repeal the 529 capital gains tax exclusion on households with a modified adjusted gross income of more than $220,000 (or $110,000 if filing singly), bringing 529s into line with the rules applied to Coverdell Savings Accounts. This could save around $17 billion over the next ten years.17
- Eliminate “accelerated gift” rule
As discussed earlier, the accelerated gift rule effectively allows parents (or others) to gift five years’ worth of money into a 529 account in a single year without triggering the gift rules in the federal tax code, up to a maximum of $140,000 for a married couple. Allowing families to front-load the plans in this way increases the capital gain and therefore the value of the capital gains break. By definition, only the most affluent families will be able to take advantage of this superfunding rule. It could be eliminated even if the other tax advantages of 529 plans were left in place.
- Limit state income tax benefits
As we reported above, in the 24 states who report spending on these tax breaks, the combined cost is around $265 million each year; the total for all states with tax benefits is almost certainly substantially more.18 This is not an insignificant amount of spending, especially in an environment where constrained state budgets has resulted in substantial cuts in funding for public colleges and universities.
States could also tighten rules on how long funds remain in 529 accounts, particularly as the rationale is to encourage long-term savings rather than short-term tax avoidance.19 In New York, a parent can put $10,000 (the maximum allowable amount) into a 529 plan, wait a week for the funds to clear, and then withdraw the money to pay for college fees immediately (or use for some other purpose, of course). This will bring tax savings of up to $645.20 In Montana, by contrast, funds withdrawn within three years of the establishment of the plan are subject to the usual state income tax.
If nothing else, there is too little transparency on what states have been spending on these tax breaks and the beneficiaries. Of the 33 states that offer such benefits, 24 report expenditures, and just 4 provide an income breakdown of recipients.
- Increase state match for low-income families
As well as reducing subsidies to higher-income families, states could increase them for poorer ones. Only 11 states currently offer subsidies to low-income families saving into 529 plans, mostly through matching grant programs. The amounts are low—typically around $400 a year in subsidy—and take-up rates are modest. If states were able to save money by limiting or removing the tax breaks for affluent families, perhaps these funds could be used to make more generous matching grants to poorer families.
States could look to the growing number of experiments designed to promote an asset-based financial model for meeting college costs. In San Francisco’s Kindergarten to College (K2C) program, for example, every child starting in kindergarten automatically has an account opened with an initial deposit of $50 from the City and County of San Francisco. Another program, the SEED for Oklahoma Kids experiment, randomly selected newborn children in the state for auto-enrollment in the state’s 529 college savings plan and seeded the accounts with $1,000. The program also offers a dollar-for-dollar savings match up to $250 a year to families making below $29,000 and 50 cents per dollar for those with incomes between $29,000 and $44,000. The evidence suggests that the impact on rates of savings by beneficiaries has been extremely modest (an average of $47 compared to $13 in the control group).
The limited research on these programs suggests that the biggest impact comes from encouraging a “college-going mindset” among children and in lower-income families. There is some evidence that simply having a savings account, controlling for other factors, results in increased college attendance.21 To the extent that college savings funds remain a feature of the policy landscape, widening the range of families and students who benefit is an important policy goal.22 But we should be realistic about the capacity of low-income families to save, even with additional incentives.
Will 529 plans remain part of the policy landscape? We hope not, but fear so. Once a powerful group starts to see benefits from a particular tax break, it is hard to take it away. Voters are human, and therefore have an aversion to loss. There is little in the current political climate to expect much movement at the federal level. However, short of abolition we have offered some concrete reform options, to limit the tax deduction to families who are not affluent, and end the “superfunding” exemption on gift rules. These are in the spirit of the Trump administration’s stated desire to simplify the tax code. Meanwhile, state policymakers could take some steps to reduce wasteful spending on regressive income tax deductions on 529 contributions and/or boost subsidies to lower-income families.
The chance to gain a college degree is seen by many as an important element of the American Dream, and a part of the “great equalizer” of education. Right now, however, higher education serves to reinforce class divisions more than to reduce them. 529 plans are one element of a system of financing that is both unfair and ineffective. 529 plans are of course just one example of the “upside down” tax expenditures in both federal and state tax codes. There are other, bigger fish to fry, notably the mortgage interest deduction. But the Bush-era handout to 529 holders is a strong contender for the prize of most absurd tax break of all: a boondoggle for the rich that does nothing to increase savings, wastes precious federal tax dollars, and may even fuel inflation in tuition fees.
In policy terms, the case for reform is unanswerable. Politics, of course, it is a different matter.
- Paul Waldman, “Don’t Mess with Government Giveaways to the Well-Off,” Washington Post Plum Line, January 28, 2015 (https://www.washingtonpost.com/blogs/plum-line/wp/2015/01/28/dont-mess-with-government-giveaways-to-the-well-off/).
- If a grandparent owns the account it is excluded entirely, but may negatively impact financial aid assessment in the subsequent year.
- Income tax reduction of $575 at top Maryland rate of 5.75, minus a $212 increase in federal income tax.
- Data come from analysis of Census’ American FactFinder “FAMILY TYPE BY PRESENCE OF OWN CHILDREN UNDER 18 YEARS BY FAMILY INCOME IN THE PAST 12 MONTHS (IN 2010 INFLATION-ADJUSTED DOLLARS)” Table C19131.
- The pre-tax household income at the bottom of the top quintile (i.e. 80th percentile) was $117,000 in 2016.
- “Federal Poverty Guidelines.” Families USA. http://familiesusa.org/product/federal-poverty-guidelines.
- These figures are based on average annual savings over the course of a three-year period (2013-2015). They are drawn from a series of Sallie Mae reports, How America Saves for College, issued between 2014-2016 outlining college savings patterns among American parents of different income levels. Projected savings amounts reflect average annual savings during this three-year period. For more information see https://salliemae.newshq.businesswire.com/sites/salliemae.newshq.businesswire.com/files/doc_library/file/HowAmericaSaves2014_FINAL_4.9.14.pdf, https://salliemae.newshq.businesswire.com/sites/salliemae.newshq.businesswire.com/files/doc_library/file/HowAmericaSaves2015_FINAL.pdf, and https://salliemae.newshq.businesswire.com/sites/salliemae.newshq.businesswire.com/files/doc_library/file/HowAmericaSaves2016_FINAL.pdf.
- While the Obama proposal to eliminate the tax-advantage of 529 savings plans would have taxed withdrawals at ordinary income tax rates, a case has been made that in such a world those families that make primary use of such plans would likely shift their college saving funds to other investment accounts subject to capital gains taxation. Thus, we assumed a 100 percent stock portfolio and constructed our estimate of the federal tax benefit to families as the foregone capital gains tax that would have been assessed on 529 savings plan investment growth. For more information see https://www.washingtonpost.com/opinions/tax-free-college-savings-plans-should-have-a-means-tested-component/2015/01/25/7008d93c-a34c-11e4-903f-9f2faf7cd9fe_story.html?utm_term=.224201d64498; https://www.edvisors.com/blog/tax-college-savings-plans-01-2015/
- The minimum threshold for the top quintile of household income in 2015 was $112,262. For more detail see http://www.taxpolicycenter.org/statistics/household-income-quintiles.
- The 95th percentile of the income distribution in 2015.
- The $15,000 in household income would result in a capital gains rate 15 percent rate applied to the $214,319 of the investment growth.
- Note that these estimates have been generated by collecting reported figures from the most recent state-level expenditure reports (as such, the data used for state-level estimates span the period of 2009-2016). Some states, like Colorado, rely on models to estimate the magnitude of specific tax spending provisions in their state expenditure reports.
- See comments in response to “A college is cutting tuition for 529 savers. It shouldn’t.” 3rd October 2016. https://www.brookings.edu/blog/social-mobility-memos/2016/10/03/a-college-is-cutting-tuition-for-529-savers-it-shouldnt/.
- In order to generate this figure the authors of the study used NPSAS data to tabulate the degree to which the sticker price charged at the institution a student actually attended compared to the average sticker price for the sector the institution was classified in (i.e., two-year public, four-year public, private nonprofit, public two-year, and private for-profit).
- The report concluded that “having a parent who graduated from a selective university is associated with a three-fold increase in the likelihood that a child will attend a selective university, even when we control for a variety of family advantages.”
- This calculation was done by taking the total Coverdell and 529 savings plan federal tax expenditure estimate for 2017-2026 from Turner (2016) and multiplying by the share of total 529/Coverdel account balances owned by the those in the 95-100th percentile of household income according to a Treasury analysis conducted in 2009 (https://www.treasury.gov/resource-center/economic-policy/Documents/09092009TreasuryReportSection529.pdf).
- Nine states, including several populous ones like Illinois, offer deductions/credits for 529 savings plan contributions, but do not report associated expenditure figures.
- “Compare by Features.” Savingforcollege.com. Available at http://www.savingforcollege.com/compare_529_plans/?plan_question_ids=302&page=compare_plan_questions.
- “529 Plan State Calculator.” Merrill Edge. Available at https://www.merrilledge.com/college-savings/529-state-tax-calculator.
- Elliott, W., & Beverly, S. (2010a). Staying on course: The effects of savings and assets on the college progress of young adults (CSD Working Paper 10-12). St. Louis, MO: Washington University, Center for Social Development.
- It should be noted that the cost of such universal programs tend to be expensive. According to San Francisco city officials the K2C program costs about $650,000 per year.