By Prof. Syon Bhanot (Swarthmore College)
To help promote retirement security for individuals and families, governments around the world offer policies to encourage retirement savings—in the United States, these take the form of 401(k)s, IRAs, and other tax-preferred savings plans. These programs are expensive—for those U.S. programs, the tax subsidies amount to over $100 billion each year. Are they worth it? The answer depends on how well these programs are designed from a behavioral perspective: how do real individuals and families respond to tax incentives? And, if they don’t respond as expected, can behavioral approaches do better?
To find the answers to these questions, researchers have looked overseas, to countries with similar savings programs to the U.S., and better sources of data. Active vs. Passive Decisions and Crowd-out in Retirement Savings Accounts: Evidence from Denmark, by Raj Chetty, John Friedman, Søren Leth-Petersen, Torben Heien Nielsen, and Tore Olsen, published in 2014 in the Quarterly Journal of Economics, tackles these questions and provides some surprising answers.
In 1999, the government of Denmark reduced a subsidy it offered to individuals contributing to capital pension accounts by 14 cents per kroner – but only for people in the top income tax bracket. The figure below shows the effect of this policy change on rates of capital pension account contribution, with the dotted line representing the income level at which the subsidy-reduction policy kicked in.
At first glance, this is pretty compelling evidence that the policy changed behavior. It seems as though average capital pension contributions decreased a lot for those affected by the policy in the years after the policy was implemented (1999-2001). One might look at this graph and conclude something like, “government savings incentives play a major role in helping people save in Denmark.”
There are two main problems with this conclusion, as the authors cleverly explain. First, the reduction in contributions in the above graph is entirely driven by a small segment of the study population – 19% to be exact – who significantly reduced their capital pension contributions after the policy change (in fact, most of these folks seemed to stop contributing to capital pensions entirely). In other words, over 80% of the study population did not adjust their capital pension contributions at all, in response to a pretty significant decrease in government financial incentives to contribute!
The second problem concerns the behavior of the 19% of people who did respond to the policy. What did they do with their money after the new policy went into effect? Here again, the authors have some fascinating answers. The quick summary: people moved their money from capital pension accounts… to other pension and savings accounts. The graphs below show that while the policy might have affected people’s contribution to their capital pension accounts specifically, it did not really decrease the overall savings rate.
Why does this matter? Well, it means that the subsidies for saving from the government were not really encouraging people to save – instead, they were basically encouraging people to save in a particular way, taking their money from one form of retirement savings and putting it into another form of retirement savings. In the end, the authors estimate that every kroner the government gave out in savings subsidies only increased overall savings by… one cent. Not a great return.
Fortunately, the authors present some pretty compelling evidence that there are better ways to motivate people to save more for retirement. Specifically, they turn to one of the most powerful behavioral tools in public policy: the default. In the paper, they look at Danish workers who switch from working at firms with a lower default employer pension contribution rate to firms with a higher default employer pension contribution rate. What they find can be seen at right.
What is this figure showing us? It demonstrates that when a worker moved to a firm that used a higher default employer pension contribution rate, the worker tended to stick with that higher rate – and not only that, they actually didn’t change their other saving behaviors (taxable saving or individual pensions) very much.
To understand why, think back to your first day at a new job. More likely than not, a few things happen on Day 1. Someone brings you a bunch of paperwork. You look at it. You sigh. You check Facebook. Then you sign the papers. And you don’t really think too much about it. What this study finds, basically, is that when that paperwork includes (as a default) a higher employer pension contribution than you had at your previous job, new workers don’t really seem to notice – or at least, they don’t behave as if they notice.
What can we conclude? Well, financial incentives to save may not work as well as we like to think. And defaults seem to matter… a lot. Even when it comes to major financial issues like saving for retirement, people seem to be pretty irrational, misbehaving in ways we may not have expected.