She’s really into My Little Pony and has formed a 1st grade gang, called the Pony Sisters, which I believe is non-violent, though possibly hostile to fairy lovers. Anyway, a few months ago she decided she really wanted a large, $20 talking Princess Cadance figure, which my wife and I said we weren’t going to buy for her. She gets $1.50 a week in allowance tied to chores. (I know some people don’t do that, but this paper suggests it’s the way to go, so there.) So in my head, I was like, Good luck, kid.

Last week, she pulled out $20 and asked to be taken to Target. There appears to be no foul play by the hands of the Pony Sisters. I checked.

MORE: Will the housing rebound crush the job market?

A paper out this week from the National Bureau of Economic Research (NBER) suggests that I shouldn’t have been so surprised. The conclusion: The best way to get people to save is to treat them like children.

The paper is called Do Lottery Payments Induce Savings Behavior? and is written by a bunch of economics professors, including three from the University of Maryland, one from Northwestern, and another from Maastricht in the Netherlands. And it’s important, because, as you may have heard, we Americans are pretty bad savers. So people have been looking for ways to tackle this.

The professors asked 100 students at the University of Maryland to choose from a variety of savings accounts. Some of them were similar to traditional savings accounts in that they paid a set consistent interest rate over time, a rate that would yield $15 over three weeks in this case. Others worked on more of a lottery system, which the professors called prize-linked accounts. Those accounts paid a lower rate of interest — $15 over five weeks instead of three — but 5% of the time the accounts would pay an additional $200 in interest over the same five weeks.

What did the professors find? Introduce a prize, and people are much more interested in saving and willing to save for longer. Overwhelmingly the students chose the accounts with the potential for big payouts, even though in most instances the guaranteed accounts ended paying more in the same time. The bigger the potential prize, even if the chances of winning shrunk, the more likely people would be to save.

And in a way this is what my daughter is doing as well. Having a large prize in mind was a powerful motivator. She actually hasn’t saved much money since. Without a goal, saving isn’t that interesting for her. Some economists have said allowing people to name their savings accounts, like first-house account or Christmas account, will promote more savings.

MORE: Yes, you can make money in Europe. Here’s how

There aren’t any prize-linked accounts like this in the U.S., but the authors say they are available in other countries. And it might not just be applicable for savings accounts. You could image applying this to 401(k) plans where employers pay out prizes. That might be a more effective and cheaper way for employers to encourage retirement savings.

So would this really work, and is it a good idea? It’s not clear. These are college kids taking the test who don’t have a family or house that they need to save for and may be less risk averse than the general population. What’s more, there may be an inequality problem. Yes, if we could get people to save more, we might be collectively better off, though perhaps not now when we need spending to boost the economy. But under this system, the rewards of the additional saving might not be evenly spread. So while you might have more aggregate savings, some individuals might end up worse off.

Still, this is pretty neat idea for a startup bank, and it might actually be a bit of financial innovation that makes sense.