How virtual reality leads to more retirement savings

With a hunch that people would save more for retirement if they felt connected to their future selves, researchers devised a virtual reality world that introduced participants to their aged counterparts.

I spoke last week with Hal Ersner-Hershfield, a visiting assistant professor in the Kellogg School of Management at Northwestern University, about what the study found -- and what the results mean for our senior savings.

What prompted you to do this research?

It started with the knowledge of this problem that people tend to under-save and overspend. Examining the determinants and predictors of this, one that was just starting to be examined is this sense that one of the reasons we fail to save is that we don't feel connected to who we're going to be in the future. That idea isn't mine. That's something I drew on from economists and philosophers.

I started working on this in an empirical way -- looking at whether our sense of connection to our future selves actually was, in fact, a predictor of our saving behavior. I found that it was. People who feel more connected to their future selves, people who feel more like their future self is a direct extension of them rather than some foreigner they've never thought about, those are the people that saved more for retirement.

The natural extension of this was: how can we bridge the gap between who we are now and who we will be in the future. The idea that my colleagues and I came up with was, could we actually present research participants with images of what they'll look like, approximately, when they're in their retirement period?

Talk about the research process.

We used a combination of software that relies on an age-progression algorithm. [This] allows people to take a photo of somebody and run it through this program. The age-progression algorithm does all the things that happen with aging -- sagging the cheeks, the jowls, the eyes, graying the hair, adding some wrinkles. We conducted the study in a virtual reality environment, which allowed people to step into the body and the mind of someone we want to represent. We can have people go into this virtual world and look into a virtual mirror and interact with who they'll be in the future.

Half of our participants did that -- they saw themselves in the future. Half of our participants just saw a digital avatar of themselves now. We found that the participants who were exposed to their future selves allocated almost twice as much money to a hypothetical retirement account compared to the participants who were exposed to their current selves. We've since run follow-up studies that rule out alternative explanations.

Were you surprised by these findings?

We were surprised that it had this big of an effect and we were surprised that it seemed to work. I have to caution that this is still the hypothetical realm. We are now trying to examine this using actual saving decisions. We're confident that this will translate to real decision making.

What should be done with these results?

In one of the studies in the [as yet unpublished] paper, I take this out of the virtual reality world. We created a web-based platform where people make their retirement allocation decisions. Staring at them as they do this is either an image of themselves now or themselves in the future. Those images change expression as a function of the quality of the decision. If I allocate a lot toward retirement, my future self will actually smile at me. This is just a teaser for what we hope can be used on a mass scale for people who are making real decisions.

Could this type of virtual reality technology could be replicated for another decision-making situation?

I've been talking to other people about looking at this in terms or obesity or health or smoking. I don't think it needs to be confined to the financial domain by any means.

What's the next step for this work?

One step is developing the web-based application so people can use it.

Another big future direction is determining whether this actually impacts future saving behavior and looking at the temporal scale of those impacts. If I show you an image of your future self now, it might impact a decision that you make today. But will that impact your behavior over the next week, two weeks, month? How fleeting are the effects of these manipulations?

Beyond that -- and this is definitely in the distant future -- looking at the boundary conditions. For whom do these manipulations work best? For whom do they have no effect? Under what conditions and domains do they work best?

Another big one is: what's the mechanism here? We have an idea that what's happening is that these manipulations are causing people to be more empathetic toward their future selves, to better understand how their future selves will feel. But we're still searching for the data for that. That's something we need to uncover in the next couple years.

Photo, top: Hal Ersner-Hershfield, current

Photo, bottom: Ersner-Hershfield, at 68 / By Michelle Del Rosario

This post was originally published on Smartplanet.com