One of the most important financial decisions people make is when to retire. It’s also one of the worst decisions many people make.

Specifically, they retire too early, resulting in serious financial shortfalls in old age.

The good news is that, according to a new study by Philipp Schreiber and Martin Weber at the University of Mannheim in Germany, there’s a simple two-question quiz that can help predict whether you’ll regret the timing of your own retirement. Here are the questions:

Question 1: You just learned that you are due a tax refund. If you’d like, you can get the $1,000 refund right away. Alternatively, you can get a $1,100 refund in 10 months. Which do you prefer?

Question 2: You just learned that you are due a tax refund. If you’d like, you can get a $1,000 refund in 18 months. Alternatively, you can get a $1,100 refund in 28 months? Which do you prefer?

The two questions are nearly identical. Each poses the same kind of choice. But the second question postpones the two options for delivery by 18 months, while the first offers an option for immediate delivery.

‍

Time will tell

The point of the exercise is to measure the consistency of a person’s time preferences. Someone with consistent time preferences should answer both questions the same way—choosing the early option both times, or the delayed option both times. Such consistency is a requirement for making financial plans that you stick with.

Some people, however, choose inconsistently. They will take the larger tax refund if both refunds require a delay, as in the second question. But they choose to accept the smaller amount when it is available immediately, as it is in the first question. For these people, the prospect of getting an immediate reward is simply too difficult to resist.

The respondents with inconsistent answers exhibit a tendency known as present bias, or hyperbolic discounting. They strongly prefer rewards that arrive right away.

While previous research has linked present bias to a lack of retirement savings, this new study, which tallied results from more than 3,000 Germans, shows that present bias can also lead people to retire before they are financially ready.

The researchers found that people who give the most inconsistent answers tend to retire significantly earlier (about 2.2 years on average) than those with consistent preferences. This leads to a roughly 13% reduction in their monthly benefits. Over time, these people are also far more likely to say they regretted the timing of their retirement.

‍

Help steering

This research helps us better understand why people choose to retire early. It can also help us find ways to stop people from retiring too early.

For instance, many workers now benefit from automatic savings programs that rely on default payroll deductions to help them save for retirement. These programs generally use a one-size-fits-all approach, recommending the same savings rate for all workers.

While these defaults have boosted savings for many Americans, they could be even more effective if they were personalized according to the results of the two-question quiz. Consider a person who exhibits a strong bias for receiving rewards in the present. Given the likelihood that she’ll be tempted by an early retirement, she might want to be defaulted to a higher savings rate during her working years. This will help her avoid future regret over the timing of her retirement decision, since she will have sufficient savings.

We can also redesign the Social Security enrollment process to minimize the possibility of regret. The program is structured so that you can start receiving payments at any time after the age of 62. However, the monthly payments will be larger for every month you delay signing up to receive benefits, at least until you turn 70. For instance, a person who can expect to receive $1,000 per month if they retire at 62 will see his benefits increase to approximately $1,750 per month if he can wait until he’s 70 before collecting.

A number of economists have argued that waiting for the larger payment is usually a much better deal. Nevertheless, most people aren’t willing to wait. According to an analysis by Alicia Munnell and Anqi Chen at Boston College, the most popular age, by far, to start Social Security is 62.

‍

Fewer regrets

With the new research from Germany, we can come up with strategies to encourage better decisions. One approach is to ask people to estimate their preferred retirement age when they are still working. Because retirement benefits are far off, they probably won’t be tempted by the smaller/sooner amount and will likely predict an age well past 62. While this estimated age isn’t binding, it will allow Social Security to personalize communication with that person in a way that might reduce their future regret. (The monetary amount itself won’t change, just the way the options are presented.)

Let’s say, for instance, that a person stated a preferred retirement age of 70 during his working years. When presenting his retirement options, Social Security could describe the benefits he’d collect at 62 as a relative loss of approximately $750 per month, at least compared with the benefits he’d receive if he waited until his preferred retirement age. Such framing could make the possibility of starting benefits right away far less appealing.

Nobody wants to regret his or her retirement choices; many of these decisions cannot be undone. By identifying those workers who are planning for a late retirement but are likely to succumb to the temptation of an early one, we can take steps to prevent mistakes before they occur.