Dr. Yue Li, an Assistant Professor from the Department of Economics was interviewed at the 2017 American Economic Association (AEA) Annual Meeting on how people think about savings for retirement on January 7th 2017 at Chicago, IL. The AEA annual meeting is a three-day meeting each January that attracts over 13,000 of the best minds in economics to network and celebrate new achievements in economic research.
The presented paper is a joint work with John Duffy, a Professor of Economics at UC Irvine. In this paper, Yue and John design a laboratory experiment and recruit human subjects to explore how variations in Social Security replacement rates—the rate of Social Security benefits to pre-retirement after-tax average earnings—might aid or hinder individuals' ability to make good lifecycle consumption and savings plans, an assessment that is difficult to conduct in the field. The answer to this question would be useful for the government to better understand the behavioral responses towards future Social Security reforms.
There are two key main findings. First, for every income profile, they find subjects on average over-consume in the early periods of life and under-consume in later periods of life relative to the conditionally optimal level, which is the amount subjects should consume to maximize utility based on their current economic resources and expectations of future income. Second, they find any sudden drop in income or endowment will reduce lifetime utility. To explain the observed consumption patterns, they conduct a specification search, and find that a two-type model with one type being rational agents who consume the conditionally optimal level and the other type being hand-to-mouth consumers who consume non-capital income each period best fits the experimental data.
To put the findings into a broader picture, one could think of the two primary tools to resolve the long-term solvency issue of the Social security trust fund: raise taxes or cut benefits. If it turns out the government start to cut benefits, a basic life-cycle model with rational agents, which is the current workhorse model in economics, predicts that households will save more for their retirement. This additional savings will help them partially offset the effect of benefit cut on consumption after retirement. From examining consumption and saving decisions in this experiment, this paper finds that not every one is rational. A non-trivial fraction of subjects use a rule-of-thumb to decide their consumption amount, and the most common used rule-of-thumb is to consume the non-capital income each period. If this was the case in the real world, in response the benefits reduction, some people will not increase their private savings and so they will experience a very large fall in their consumption at retirement. This type of welfare loss is not captured by a traditional life-cycle models with rational agents, and is a serious issue that requires more attention.