During their working lives, people spend too freely, save too little, invest unwisely and generally make a hash of their finances. There are exceptions, of course, but study after study shows this to be true for millions.

If only they knew better, getting better financial training at school or work. Common sense says that will help, but how much of a difference does it really make?

“People who are financially literate tend to plan more, they save more, they invest smarter and the end up with more wealth at retirement,” says Wharton professor Olivia S. Mitchell, director of the Pension Research Council. “But that’s a correlation, not necessarily a causal analysis.”

In other words, does financial literacy produce more wealth, or do people with more wealth bone up on financial strategies to become more literate? Simply looking at the wealth and literacy of the well to do and the less so doesn’t reveal which comes first. Without a better understanding of the interplay between financial literacy and financial success, employers and other providers don’t know what kind of literacy program to offer, or which potential recipients would get the most out of it.

For decades, research on this subject has been limited by the difficulty in establishing control groups. Ideally, researchers would set up two or more identical groups, offer education to some and none to the controls, and then compare the outcomes over a lifetime. But that’s just too difficult. “Those [types of studies] are hard, they are expensive and they take a long time,” Mitchell observes.

So Mitchell and two collaborators, Annamaria Lusardi of the George Washington University School of Business and Pierre-Carl Michaud, Université du Québec, devised an alternative: use virtual, computer-generated people instead of real ones.

“A one-time financial education program may have little effect, as expected, but the long-term effects of a persistent financial education program can be sizable.”

“It’s very useful, because you can simulate them over their lifetimes and all it takes is a few hours on the computer instead of waiting 80 years,” Mitchell says.

The results are described in the paper “Using a Life Cycle Model to Evaluate Financial Literacy Program Effectiveness.” It shows that the most effective programs would use regular follow-ups, like booster shots, to reinforce the lessons given at the start. The research found that workers around age 40 are likely to get the most out of such programs, increasing their wealth at retirement by nearly 10%, because they are old enough to worry about retirement but still have enough time for better habits to pay off.

And it showed that real-world studies that try to gauge program effectiveness by comparing participants to non-participants can seriously overstate the programs’ value, since the two groups are not really comparable.

“Despite the widespread popularity of [financial literacy] programs in the U.S. and elsewhere, our recent literature review… argued that relatively little could be learned from most of the existing evaluations to date,” Mitchell and her colleagues write. “This is because analysts have typically not followed the protocol required by ‘gold standard’ randomized controlled trials, enabling researchers to extrapolate from observed results.”

Financial Complexity

Today, the value of financial literacy is widely assumed, but that wasn’t always so because needs weren’t as great as in the past, when people did not live as long. Now people must plan for a retirement that might last 30 years or more. Social Security is not enough to live on, and many pensions have been replaced by “defined contribution” plans like 401(k)s that shift the investment risk from the employer to the employee. So it’s now more important for individuals to save and invest for retirement.

At the same time, financial life has become more complex with the growth in different types of mutual funds, annuities, student loans, various kinds of mortgages and ever more convoluted tax laws. Retirees can face complex issues involving the best way to withdraw cash from the assets accumulated.

Traditionally, says Mitchell, employers who offered financial education “have properly concentrated on education around saving for retirement.” That’s largely because programs were designed to help employees make the most of workplace retirement plans like 401(k)s. But a good program, she adds, needs to go further, because the worker who makes poor decisions on spending, buying a car or choosing a mortgage will not have enough money left over to get the most out of the retirement plan. Here, financial education is critical, but the need to cover more topics makes it hard to perform the apples-to-apples comparisons to determine what works best.

For the study, Mitchell, Lusardi, and Michaud gathered extensive data on real people, including their levels of financial literacy and personal finances, and evidence of the financial decisions they had made over the years and the wealth they accumulated. The data was used to create 2,500 simulated people who mirrored that population. Using the simulated people, or sims, can supply a control group that is exactly like those who are offered various types of programs, the researchers write. By presenting the sims with typical challenges and opportunities from the real world, the researchers could judge how well real people would succeed financially under various conditions with different levels of financial education.

“Financial education can be very important, but you have to pick the group that is most likely to benefit from it, number one. In other words, probably not a 20-year-old.”

The work demonstrated that, in the real world, employees who decide not to participate in workplace literacy programs should not be used as a baseline to gauge the program’s benefits for participants. The non-participants choose not to take part because they tend to be younger, less educated and less affluent, and therefore feel they have little to gain by sitting through the course, which may cause them to lose time or money. A true control group would not be so different from the program participants.

“In fact, if one were to compare program participants and non-participants, one would erroneously conclude that the program had an enormous impact on the stock of financial knowledge, producing a 20 percentage point advantage for participants,” the researchers write. Effects turned out to be much more modest.

Those who take the course do show an initial gain in financial literacy and economic benefit, but this dwindles over time. The greatest effect of improved financial literacy was found with sims who received financial education at age 40 followed by regular booster shots. They had a 10% gain in wealth at retirement.

“In other words, a one-time financial education program may have little effect, as expected, but the long-term effects of a persistent financial education program can be sizable,” the researchers write.

But that’s not so for everyone. The researchers concluded, for example, that “the low-income and less-educated have less to gain from participating in such programs.” Any organization considering whether to offer financial education must therefore think about which potential participants are most likely to benefit.

Mitchell notes that it is not enough to simply offer a program on budgeting, saving and investing and assume the knowledge will linger until the participant needs it. An ideal program would be made available at times the recipients are “more teachable,” like when they settle down with a partner, have a child or start to feel retirement looming. “Those are moments when they are much more likely to be receptive to things like life insurance and disability insurance and saving for college,” she explains.

“Financial education can be very important, but you have to pick the group that is most likely to benefit from it, number one. In other words, probably not a 20-year-old. If you can provide some reinforcement, that would be most efficacious, because it would offset the depreciation [in knowledge] that would happen over time,” Mitchell says.