If the average American went in for a financial checkup, he or she might get rushed to the emergency room. Forty-four percent of us can’t cover a $400 out-of-pocket expense, and 52% of American households have no retirement savings. We seem to be chronically poor at making financial decisions. We commit costly mistakes across all areas of personal finance including decisions about savings, investing, budgeting and borrowing.

Diagnosing the problem is easy: We make mistakes because financial decision making is hard, and we lack an understanding of the decisions we face. Finding a cure is much more difficult. To many, the obvious treatment is financial education, but recent research suggests that financial education is not effective. Some promising new ideas such as “just-in-time” education and “nudges” to help us make better decisions are starting to emerge as alternative approaches.

Ineffective programs

Governments around the world have invested huge resources in initiatives aimed at improving financial literacy, such as the Federal Deposit Insurance Corp.’s My Smart program, which helps low-income individuals develop financial skills.

Unfortunately, research into the effectiveness of these programs paints a grim picture. A recent meta-analysis looked at every known study examining whether a financial-education intervention — such as training sessions, classes or one-on-one counseling — improves positive financial behaviors and financial health. Across all those studies, there was almost no benefit. Those participating in financial education were essentially no better off. If we want to come up with better solutions, we need to understand the psychological barriers to financial education. Why is it so hard for people to learn?

Challenges to learning

The first reason is relevance. The mind is not like a computer that can store arbitrary amounts of information. Instead, we tend to retain only what is useful for navigating our current circumstances. For example, if you learn the abstruse mathematics of car leases to prepare for a negotiation at the dealership, you might be surprised to find how little you remember when renewing the lease years later. Financial education tends not to stick if it is not useful right away. Many curricula are flawed in precisely this way, teaching high schoolers about mortgages or debt management.

The second challenge is that the financial domain is a particularly complex one. You could call it a perfect storm. Many of the most important financial principles are highly counterintuitive. Take compound interest. Compounding is a nonlinear function. The more you save, the more the savings accelerate, because the interest accrued in one period earns additional interest in the next period. In one study, participants were asked to guess how much money they would have after 40 years if they put $400 a month into a savings account that made a 10% annual return. The median guess was $223,000. The correct answer: $2.5 million.

Here’s another example: The concept of diversification is a bedrock principle of investing wherein, by combining different assets in a portfolio, an investor can mitigate risk. When one asset does poorly, another is likely to do well. Those things tend to balance out, and, overall, the performance of the portfolio is more predictable than if one were to hold fewer assets. In one study, about half of participants believed precisely the opposite, that a diversified portfolio is less predictable. They seem to reason that putting all of those uncertain assets into a portfolio must make the performance of the whole portfolio highly uncertain.

Financial decisions almost always involve numbers, and many people have great difficulty with mathematics and anxiety about their numeracy. Not only does math anxiety lead people to avoid math, but it actually undercuts one’s ability to successfully complete tasks with a numeric component when they are encountered.

Another difficulty is that financial decisions often require projecting into the future. How much money will I need in retirement? Can I afford this house given what my income and expenses are likely to be in 10 years?

People make a lot of mistakes when projecting into the future. For instance, when thinking about how much spare money they will have in the future, people overweight increases in their earnings, and neglect how their expenses will grow, leading to overly rosy predictions. People also tend to be short-sighted, placing a lot of value on having money immediately, even if they would be better off delaying. This is one reason for the prevalence of exceedingly high-interest financial products like payday loans.

A third challenge for financial education is the mismatch between what people know and what they think they know. We tend to be poor at evaluating how well we understand things. Sometimes this leads to underconfidence, like when a prospective retiree avoids retirement-savings decisions because the landscape just seems so daunting and overwhelmingly complex. More often, the opposite is true. We are overconfident, thinking we understand things better than we do.

In a recent study, participants in a financial-literacy course were tested six months after completion of the course. Their objective financial knowledge was no higher than before the course; the concepts had faded. But their subjective knowledge, their sense of how much they knew, remained consistently higher. This kind of mismatch is problematic for financial education. It suggests that people often don’t know when they need help and are unlikely to be open to learning.

Promising ideas

For all those reasons, traditional financial education fails. What can we do instead? We need solutions that sidestep those psychological barriers.

Here are a few of the most promising ideas:

One common strategy discussed in the context of the overall failure of traditional approaches to financial education is the idea of just-in-time education. This concept builds on the idea that some form of financial education must be effective, and that a central failure of financial education stems from the fact that people engage in financial-education classes at a time when the lessons are not relevant for them. Consequently, the lessons seem unnecessarily abstract and are unlikely to be remembered when the time to use them comes along.

Just-in-time education aims to more tightly couple the timing of educational content with its use. For example, DC DOES and Bank on DC together provide financial education concurrent with access to a checking and savings account to participants of the Summer Youth Employment Program (SYEP).

Although just-in-time education has promise, a central challenge is that it is hard to implement. For instance, it is not easy to identify who will be searching for a home early enough to reach them at the moment when it is close enough to the purchase that they are receptive to new information but early enough in the process that they have room to alter their decision-making process in response.

A nudge toward better behavior

In contrast to aiming to educate consumers, an alternate path is to leverage nudges and “choice architecture” to encourage consumers to make the right decisions without additional effort on their part. Although there are many different ways to implement nudges, one of the most common and powerful strategies is to switch defaults.

For example, many employers have shifted the default setting on contributions to 401(k) retirement-savings plans from opt-in to opt-out. In this case, if the employee does nothing, a predefined amount (say, 3%) of her paycheck will automatically be directed into a retirement-savings account. Of course, if the employee wants to make a larger or smaller contribution, she can change the setting accordingly. Employers in the U.K. are required to enroll employees in a default pension plan, with an option to withdraw from the program. Analysis of this policy finds that it increased the rate of participation in these pensions by 37 percentage points, reaching 88% of employees.

Research has consistently shown that defaults are sticky and that an employee is likely to keep the default setting in place.

While nudges can be effective at encouraging people to reach their goals, they also come with challenges.

First, they often offer a one-size-fits-all solution when individuals vary greatly in their financial situations and needs. Recent work has proposed the concept of “smart defaults” and “adaptive defaults” that respond to specific consumer characteristics and tailor defaults accordingly.

However, an additional concern over paternalism remains, with fears of institutions overreaching their power and persuading consumers to take actions (even those that may be helpful to them) without their deliberate awareness.

Computers’ advice

A third approach, that of algorithmic recommendations, falls somewhere between explicit provision of knowledge and passive persuasion. In this case, computers can incorporate knowledge about a given consumer’s characteristics to present recommendations. Consumers could enter information about themselves, or the algorithm could gather information from other sources. In many cases, the recommendations could be most relevant and tailored to the individual. For example, Betterment is an online financial adviser that tailors its algorithmic advice based on individual investor preferences.

Still, two main obstacles stand in the way of clear success for this approach. First, the more informed these recommendations become, the more consumers may be wary of their privacy. And, second, research shows that people are averse to algorithms and will be less forgiving of them than of a human counterpart, particularly in a case in which algorithms make an error.

The first step toward better health is often a realization that what we’ve been trying is not working. None of these solutions is a panacea, and all of them have their challenges in terms of implementation. But they are new ideas and represent the efforts of a vigorous and creative research community dedicated to promoting better financial decision making.

Read the rest of Philip Fernbach and Abigail Sussman at MarketWatch