In the spring of 2014, Ed FitzGerald visited a Cleveland elementary school to announce a new and highly ambitious education program: free college savings accounts with cash subsidies for every kindergartner in Cuyahoga County, Ohio. At the time, FitzGerald was the county executive and the Democratic nominee for governor. A similar savings account program had been introduced a few years earlier in San Francisco, but that was an experiment limited to public schools. FitzGerald’s initiative included children in private and parochial schools as well. With 10,500 accounts in its first year, it promised to be the largest effort of its kind in the country.
At a press conference, FitzGerald cited the increasing cost of higher education and low college attendance rates as reasons for the new program. Past efforts by the state and federal governments to encourage saving for college, through so-called 529 investment vehicles, had largely failed. Research from the Board of Governors of the Federal Reserve System found that in 2013, these 529s achieved a meager 2.5 percent participation rate among eligible families. For households in the bottom half of the income distribution, participation was below 1 percent.
The savings plans, named for the 529 section of the federal tax code created by Congress in 1996, built upon an idea pioneered by Michigan in the late 1980s. The idea was that government could help parents pay for college by offering them tax breaks on savings accounts for higher education. The money parents withdrew from the accounts to pay for education expenses and the interest generated by the invested principal were tax exempt.
Over time, the number of 529 accounts and the amount of money invested in them had grown. But participation in the plans skewed toward wealthy parents who had college degrees, were already likely to send their children to college, and stood to benefit the most from tax breaks on their investments. Because the early state-run 529 plans required parents to be proactive in opening accounts, they tended to exclude parents who weren’t aware of the plans. In 2012, the U.S. Government Accountability Office found that even among parents who were saving for college, as many as half didn’t know about 529 plans.
The program in Cuyahoga County was designed to be different. Participation would be automatic, so parents wouldn’t have to deal with the hassle of filling out application forms. Every family would be mailed a bank card with an initial $100 deposit from the county. In time, county officials expected foundations and local businesses to chip in with grants to match parents’ own deposits. The money would keep growing for 13 years until the child was ready for college. “It’s not a panacea. It’s not a complete solution,” FitzGerald acknowledged, but “we are -- in a small way, anyway -- making the cost of college easier to afford.”
FitzGerald said local officials had a responsibility to give children “a real sense of hope” that they could attend college and compete in the global economy. “This is part of our mission now in county government,” he declared.
The mission didn’t last long. Barely a year later, FitzGerald had finished his term as executive, lost his campaign for governor and receded from public life. In his absence, the county council voted unanimously to repeal the program, citing high administrative costs and low parent engagement. Almost $3 million, most of it from accounts that parents never claimed, went back to the county’s general fund.
Despite the well-documented failure in Cuyahoga County, many other states and localities are testing the same basic idea. At least 42 programs in 29 states promote long-term savings for post-secondary education, according to a recent review by the Corporation for Enterprise Development, or CFED, a group that promotes college savings accounts. The majority of these are run by nonprofits, but the model is increasingly being adopted by elected officials.
The early demise of Cuyahoga County’s program raises questions about why governments should offer these accounts, and what they should expect to get back from them. If they’re meant to trigger frequent, regular saving by most account holders, early indications suggest that the results will be disappointing. Making a significant dent in college affordability, a purpose so central to the accounts that it’s part of their name, seems improbable. Still, organizers of these programs say there are other reasons to offer the accounts. They can be used as a carrot to improve student performance and help encourage low-income families to establish savings accounts. They teach children about saving money for long-term goals. All of that is true. But is it enough?
Former Cuyahoga County Executive Ed Fitzgerald launched the largest savings program in the country.
Nevada launched a similarly ambitious kindergarten-to-college savings initiative the same year FitzGerald unveiled his in Cleveland. Using fees collected from financial institutions, the Nevada treasurer’s office seeded $50 per kindergarten pupil into a collective 529 investment vehicle run by the state. Following the examples of San Francisco and Cuyahoga County, Nevada made its program automatic, targeting all kindergartners in public school, regardless of income, and automatically providing seed money without any initial action by the parent. Families that followed up by opening their own personal 529 accounts were eligible for up to $1,500 in dollar-for-dollar matching deposits over a five-year period.
As in Cuyahoga County, the Nevada program had trouble engaging parents. In the first two years, the state opened about 66,000 “College Kick Start” accounts, but only 1,400 families started personal accounts. Without such accounts, parents couldn’t make future deposits or take advantage of outside matches. On its own, without additional deposits, the $50 in seed money would barely grow to $90 in 13 years at 5 percent interest. Dan Schwartz, who took over as state treasurer in 2015, took a look at the discouraging numbers and started casting about for a way to fix the problem. He didn’t want to discontinue the program. He wanted to save it.
Last fall, Schwartz began offering an additional $200 matching deposit for every kindergartner whose parents opened their own 529 account. It was a one-time demonstration to see if the extra incentive would attract new account holders. The idea was that data from the demonstration would win over lawmakers who balked at a similar proposal last year because of low parent engagement. “We will have a different story to tell at this legislative session,” says Grant Hewitt, Schwartz’s chief of staff. “It’s going to be a positive one that leads to the expansion of College Kick Start.”
Other programs around the country are trying different kinds of tweaks. In St. Louis, City Treasurer Tishaura Jones has defined parent participation as more than making personal deposits. Like some of the other programs, the one in St. Louis provides $50 in seed money and up to $100 in dollar-for-dollar matches in the first year. But nearly 90 percent of St. Louis public school students are eligible for free or reduced lunch, and many families don’t have the money to spare for a savings account. So the St. Louis program rewards other forms of participation. For every week of perfect school attendance, students earn $1 for their accounts. If parents complete a financial education course, they can earn another $50.
In the first year, as expected, most St. Louis families did not put up their own money. The city opened 3,143 accounts, and fewer than 2 percent of the eligible children had received deposits from a parent or guardian. But the other behavioral incentives were more successful. Overall, about 23 percent of account holders earned deposits for some kind of incentivized activity. Jones is now partnering with the Common Cents Lab at Duke University, a behavioral economics think tank, to find more ways of increasing parent participation. This year, her office is offering financial incentives for parents to attend credit score counseling and make deposits during tax season.
Asked if she’s disappointed that so few families put their own money into the accounts, Jones bristles. “This program is only two years old,” she says. “We shoot ourselves in the foot with instant gratification.” That’s the viewpoint of CFED as well. The group is a partner in the St. Louis program and has a goal of enrolling 1.4 million children in savings accounts around the country, up from an estimated 313,000 in 2016. This year, it’s supporting new programs in Boston, Louisville, New York City and Oakland, Calif.
St. Louis Treasurer Tishaura Jones believes parent participation can be defined as more than making personal deposits to accounts.
Jones may be right that time will improve performance. The discouraging data from Nevada, Cuyahoga County and St. Louis reflect results from very young programs. In San Francisco, which started offering college savings accounts to kindergartners in 2011, the participation rate has climbed to about 16.6 percent, up from 8.1 percent in the first year. That’s well above the national average for 529 plans, but still represents a small fraction of the accounts provided to families each year. “I want us to get higher,” says San Francisco Treasurer José Cisneros, who oversees the accounts. “I would love to see us get to 50 percent of families making deposits.”
By opening accounts automatically, Cisneros and other treasurers have bypassed the major problem with the older state-run 529 plans, which is that most parents don’t take the initial step of establishing a savings account for their children. But officials have not solved the related problem of parents not knowing about the accounts, or not making use of them once they’re open, Cisneros says. “It’s a real struggle for us to grow awareness that these accounts exist.”
While officials study the reasons for low participation, and seek ways to increase it, they’re starting to market the programs as multipurpose initiatives that do more than just address college affordability. St. Louis, for example, is trying to use its program to lower the incidence of households without a checking or savings account. The idea is to help low-income parents avoid high interest rates and fees from payday lenders and check cashers. After one year, 114 families in the program had opened accounts at a bank or credit union. That’s a small benefit, but it’s one that Jones can highlight now; it will be years before her office knows the college attendance rates of account holders, or how much parents ultimately saved.
“We don’t measure the success of a program by how many families make deposits,” says Shira Markoff, a research associate at CFED. “They face day-to-day constraints that they need to spend their income on.” Instead, CFED emphasizes the importance of opening the accounts and offering financial incentives to build upon those initial savings. According to CFED, whether the money comes from parents, the city, or outside funders matters less than whether the accounts ultimately make children more likely to attend college.
In fact, there is some research supporting the idea that simply having a college savings account is more important than how much money gets deposited. William Elliott, a professor of social welfare at the University of Kansas, has found that even children with less than $500 in a college savings account are more likely to enroll in and graduate from college than children with no savings account at all. His research also suggests that the purpose of the account matters: Even after controlling for other predictive factors like parents’ income and education, children with basic savings accounts that are earmarked for post-secondary education are more likely to attend and graduate from college than those who put savings into an account with no specific asset in mind.
Elliott sees the effects as mostly psychological. “I don’t think the end goal of the savings accounts is saving itself,” he says. “It’s more about building expectations.” Even when the accounts don’t accumulate much money, they can give students the sense that paying for college is possible. “If you’re growing up in a low-income area, you really do doubt that you’ll ever reach college,” Elliott says. “You might aspire to it, but you’ll doubt that it could actually be a reality.” Having the accounts, even with small amounts of money, may erase some of that doubt.
San Francisco Treasurer José Cisneros: "I would love to see us get to 50 percent of families making deposits."
Much of what went wrong in Cuyahoga County can serve as a lesson to other jurisdictions in how not to proceed. Cuyahoga’s program was enacted on a partisan vote: The three Republicans on the 11-person county council all opposed it. FitzGerald’s decision to run for governor left the nascent program to a divided county government. “I don’t think this program ever will be stopped, quite frankly,” FitzGerald predicted in announcing its debut. A year later, it was gone. (FitzGerald could not be reached to comment for this story.)
“Our hope was that launching this program would create dynamic effects,” says Dale Miller, a Democrat on the county council who voted for the proposal but later approved its repeal. “The county would put up this money and then local foundations, community organizations, churches and family members would put in additional resources. Unfortunately, almost none of that happened.”
Jack Schron, a Republican on the council, says the program was rolled out before it was ready. Had it been linked to existing 529 plans, money that was deposited could have accrued tax-free interest. That wasn’t done. “If it took another year to get right, who cares?” Schron says. “No one’s against kids. No one’s against trying to help individuals get ahead. But unfortunately, [the county didn’t take] the time to think about how you do it right.”
The newer savings account programs in other places are starting small and scaling up over time. Last year, Durham, N.C., launched a pilot at a single elementary school. This year, Boston will open accounts for children in five schools with plans to expand to all city public and charter schools in 2020. In New York City, a Queens school district will test the universal savings account model with an option to make the program available across the city after a three-year demonstration.
Markoff says other jurisdictions have learned from Cuyahoga County that they need to line up private funders and a broad coalition of support before starting a program. Elliott, the savings account researcher, says officials should take a broad and realistic view of what the programs can achieve. “If we think that low-income people are going to save large amounts of money and they’re going to save monthly, we’re grossly mistaken,” he says. “So we shouldn’t set those kinds of expectations for them.”
For Miller in Cuyahoga County, the lesson is quite different. He chalks up the repealed program as a casualty of the policy innovation process. When trying a new idea, sometimes you fail. “I think it was a grand vision and a little bit beyond our capabilities in retrospect, but not by enough to be ridiculous,” Miller says. “I support an experimental approach in government and, if you take that approach, not everything you try is going to work.”
*CORRECTION: A previous version of this story incorrectly named Grant Hewitt as Greg Hewitt.