January 24, 2014 | Research Feature

Save or Be Selfish?

A new model can help economists understand how spouses make financial decisions — and why many couples struggle to save.


In the last four decades, the way Americans save for retirement has shifted dramatically. Where once defined-benefit, employer-based plans made saving for retirement a matter of fact, today 401(k) and similar plans leave key decisions about how much and when to contribute up to individual households.

While that freedom of choice is alluring, there is overwhelming evidence that people today don’t save enough for retirement. Experts typically recommend that people save roughly 10 percent of their current income. Yet of those whose employers offer a 401(k), when polled, about one-third report not contributing anything to it. Of those who do contribute, the average savings rate is about 6 percent — just over half the standard recommendation. Furthermore, people appear to understand they have a savings problem: about two-thirds of people say they should be saving more for retirement.

Why do people undersave even when saving is in their own best interest? The question has puzzled economists, who, when examining how people make choices about financial management, often frame their work around rational self-interest. This work invariably focuses on individual decision making, and that might contribute to the puzzle. “We should take seriously the fact that there are multiple people in many households,” Professor Andrew Hertzberg explains. “Spouses, parents, children, and in some cases, extended family are all financially interrelated. Spouses share wealth, they may give money to their children or may contribute to their own parents or siblings.”

Economists have shied away from thinking about multi-person financial decision making because of the complexities of modeling more than one person’s behavior in a household. “For simplicity, economists normally assume that everyone in the household is either selfless or has exactly the same objectives,” says Hertzberg, who reexamined that assumption. His new model makes it easier to study household financial behavior in a way that more accurately reflects how household partners interact and make decisions. In the model, Hertzberg considers savings as the outcome of a trade-off between each household member’s desire to spend on her own wants and her concern for saving for the future. This trade-off is affected by the fact that even in the most well-adapted relationships, people are usually a little bit selfish. This matters because consumption is a way to spend on oneself, whereas savings is shared with a partner. “A couple may want and agree to save ten percent of their income each month,” Hertzberg says. “But when one of them has the opportunity to spend a few more dollars on themselves, many people do so. And the result is that they undersave.”

The severity of the undersaving problem is affected by factors such as what percentage of income is spent on shared public consumption — things like children, rent, or food that are essentially shared — or private consumption, like sporting goods or clothing. In general, a household that has a higher fraction of shared expenses will undersave less because the interests of the household members will be more closely aligned.

Hertzberg shows that one factor that can make the undersaving problem even more dramatic is spending on private durable goods, such as a sports cars or jewelry. His model predicts that household members will overspend on these items. Why is this? “Household members undersave not because of impatience, but because selfishness results in their undervaluing shared savings,” he explains. “So the ability to spend a lot of money on a sports car that will provide me with pleasure for many years is the perfect way to redirect shared resources to myself.”

The model may explain why the savings rate went up for households immediately affected by the passage of the Retirement Equity Act (REA) in 1984. Among making many other sweeping changes related to pensions, the law introduced a mandatory requirement that retirement account holders obtain spousal permission to withdraw or borrow against retirement savings or to opt out of employer sponsored life insurance programs. Until then, it was possible for one spouse to tap those savings or sign away benefits without seeking the permission of — or even informing — their partner. In including this feature in the REA, the government recognized that when a household is made up of more than one adult, financial decision making is — or should be — fundamentally a joint decision.

The savings rate increase after the passage of the REA suggests that one tactic households can use to increase their savings rate is to save in the form of assets that require joint approval to make withdrawals or borrow against. That’s one reason that jointly owned housing can be a successful form of savings — one spouse typically cannot take a loan against the house without the partner’s consent. Hertzberg’s model could provide a basis for answering a related financial services question: how much would people pay to solve their undersaving problem with other types of savings commitment technologies, such as mandatory joint authorization?

Finally, says Hertzberg, the model reinforces the idea that members of households would be more successful at building savings and assets if they negotiated how much to save, rather than simply allowing residual funds — whatever they don’t spend — to accrue.

One of Hertzberg’s next projects builds on this research by considering how households set up their financial accounts and how that impacts subsequent financial decisions about savings, spending, and debt. This research might help tackle another long-standing puzzle in economics: why do people who have money in the bank — savings — also have debt on their credit cards on which they are paying interest? “For a single person, the rational choice is to pay off the credit card and save the cost of interest,” Hertzberg notes. “But it might make sense in a household where members have individual credit cards. Credit card limits will act as a restraint on how much individual members will consume. A couple may choose to delay paying off the credit card because they know each family member would just go out and spend more.”

Andrew Hertzberg is assistant professor of finance and economics at Columbia Business School.

Read the Research

Andrew Hertzberg

"Exponential Individuals, Hyperbolic Households"

View abstract/citation  Download PDF  

Have You Read Columbia Business?

Columbia Business School’s alumni magazine connects alumni with each other and the School; celebrates alumni milestones and accomplishments; and chronicles the impact of Columbia Business School alumni, faculty members, and students on the global business landscape.

Read Columbia Business>