YOLO: Mortality Beliefs and Household Finance Puzzles

• Author: RAPHAEL S. SCHOENLE,RAWLEY Z. HEIMER,KRISTIAN OVE R. MYRSETH
• DOI: http://doi.org/10.1111/jofi.12828
• Date: 01 December 2019
• Published date: 01 December 2019

THE JOURNAL OF FINANCE •VOL. LXXIV, NO. 6 •DECEMBER 2019

YOLO: Mortality Beliefs and Household Finance

Puzzles

RAWLEY Z. HEIMER, KRISTIAN OVE R. MYRSETH,

and RAPHAEL S. SCHOENLE∗

ABSTRACT

We study the effect of subjective mortality beliefs on life-cycle behavior. With new

survey evidence, we document that survival is underestimated (overestimated) by

the young (old). We calibrate a canonical life-cycle model to elicited beliefs. Rela-

tive to calibrations using actuarial probabilities, the young undersave by 26%, and

retirees draw down their assets 27% slower, while the model’s fit to consumption

data improves by 88%. Cross-sectional regressions support the model’s predictions:

Distorted mortality beliefs correlate with savings behavior while controlling for risk

preferences, cognitive, and socioeconomic factors. Overweighting the likelihood of rare

events contributes to mortality belief distortions.

EXPECTATIONS ABOUT FUTURE EVENTS ARE CRUCIAL to intertemporal decision-

making. The economic literature reveals, however, that consumer expectations

often fail to align with the correct probabilities. One way in which individuals

err in the formation of expectations is by systematically overweighting rare

events with salient attributes (Bordalo, Gennaioli, and Shleifer (2012,2016)).

Motivated by this observation, we consider how the formation of mortality

beliefs influences household financial decision-making. Low probability events

have the potential to distort individuals’ mortality beliefs and consequently

affect economic behavior. For example, although sharks kill only one person

per year on average in the United States (cows kill 20 people per year), many

YOLO: “You only live once (expressing the view that one should make the most of the present

moment without worrying about the future, and often used as a rationale for impulsive or reckless

behaviour)” – Oxford English Dictionary.

∗Heimer is with Boston College and Carroll School of Management. Myrseth is with TrinityCollege

Dublin and Trinity Business School. Schoenle is with Department of Economics and International

Business School, Brandeis University.We are grateful for the assistance of Kyeongah Lee, Michael

Munsell, and Timothy Stehulak. We thank Eric French, David Love, and Alex Michaelides for

generously providing us with their codes, and the Qualtrics Research Suite, Diane Mogren, and

Kathy Popovich for help obtaining the survey data. We acknowledge the support of Boston College,

Brandeis University,the Federal Reserve Bank of Cleveland, University of St Andrews, and Trinity

Business School for purchasing the survey data used in this paper. The data were collected with

Institutional Review Board (“IRB”) approval of University of St Andrews. Wehave read the Journal

of Finance’s disclosure policy and have no conflicts of interest to disclose.

DOI: 10.1111/jofi.12828

2957

2958 The Journal of Finance R

people are afraid of going to the beach during the summer.1Similarly, during

Super Bowl 47, Prudential Financial Inc. aired a commercial in which 400

individuals were asked to post a blue sticker on a billboard to indicate the age

of the oldest person they know. The resulting distribution skewed toward the

age of 100 years, giving viewers the distinct impression that they would live to

be very old and prompting potential customers to contact Prudential. Indeed,

financial practitioners often use mortality beliefs in long-run financial planning

(e.g., Internet Appendix Figure A.1). 2In this paper, we present new data on

subjective mortality beliefs for approximately 4,500 respondents. This survey

was administered by a professional survey provider. We find that individuals’

beliefs about survival go from pessimistic among individuals who are young to

optimistic as individuals approach retirement age.3We attribute this flip from

pessimism to optimism to salient, cohort-specific stereotypes of cause-of-death.

Our survey asks respondents to assess the weight placed on different mortality

risk factors in considering their own survival likelihood. Young people tend to

overweight near-term rare events (e.g., plane crashes or animal aggression),

with higher weights on such events associated with greater survival pessimism.

As people age, they place less weight on these rare events and more on health

and the “natural course of aging.”

We next consider the effect of mortality beliefs on lifetime savings by devel-

oping a canonical life-cycle model with pre-cautionary savings. In this model,

mortality beliefs shave a direct bearing on the intertemporal trade-off between

consumption today and the discounted present value of future consumption

streams,

V∗

t(·)=max

Ct

{u(Ct)+βst+1Et[V∗

t+1(·)]}.(1)

Consumers decide throughout the life cycle whether to save or consume. Thus,

even small financial mistakes, incurred by modest short-run mortality belief

distortions, can accumulate by retirement into large financial shortfalls. We

calibrate the model using a survival belief function estimated with elicited

transition rates from our new survey. This simple, age-dependent survival

function captures changes in survival optimism over the life cycle.

Our model simulations show that subjective mortality beliefs cause the aver-

age retiree to undersave by approximately 26% relative to a benchmark model,

calibrated to actuarial data from the Social Security Administration (SSA).

Owing to retirees’ overestimation of long-run survival rates, individuals con-

sume about 27% less during retirement. These findings suggest that a life-cycle

model calibrated to subjective mortality beliefs gets closer to understanding

two notable puzzles in the household finance literature: Many young people

1According to a headline in the August 29, 2015 edition of the L.A. Times, ”La Jolla beach closed

after shark sighting.”

2Internet Appendix is available in the online version of the article on the Journal of Finance web-

site

3We can replicate this finding in canonical household surveys, namely, the Survey of Consumer

Finances and the Health and Retirement Survey.

YOLO: Mortality Beliefs and Household Finance Puzzles 2959

undersave toward retirement (Skinner (2007)), and the dissavings rate is too

low for many retirees (Poterba, Venti, and Wise (2011,2013)).4Additionally,

there is a sharp drop in consumption following retirement (as seen in the data,

for example, Bernheim, Skinner, and Weinberg (2001)), while consumption in

the benchmark model with actuarial rates continues to rise for several years

postretirement—further evidence that subjective mortality beliefs improve the

realism of the model.5

Moreover, we use simulated method of moments (SMM) to test whether

our life-cycle model can match actual data series on consumption and net

worth over the life cycle. Calibrating the model to subjective as opposed to

actuarial mortality beliefs improves the model’s fit by between 50% and 90%.

Finally, when we augment the model by adding realistic circumstances that

affect household decisions—namely, family formation and uncertain out-of-

pocket medical expenditures—we find that subjective mortality beliefs have

independent effects on savings behavior. These effects are qualitatively large

relative to the alternative mechanisms above.

Our model predictions are supported by cross-sectional regressions drawing

on our survey data, which include answers to off-the-shelf questions from com-

monly cited household surveys (e.g., the Survey of Consumer Finances (SCF)):

pessimistic survival beliefs are associated with a greater propensity not to save

or to even rely excessively on credit cards on a month-to-month basis. Similarly,

more (less) pessimistic beliefs are associated with an increase (decrease) in the

likelihood of spending savings more quickly, after accounting for respondents’

age and risk tolerance. To gauge economic significance, we find that the mag-

nitude of the relationship between survival beliefs and savings behavior is at

least as large as that between savings behavior and financial literacy—and as

that between intended bequests and savings behavior.

Though we are cautious to interpret causality from the reduced-form regres-

sion results, we provide evidence to limit concern over omitted variable bias.

The most intuitive candidate alternative explanation is that these subjective

mortality beliefs proxy for individual-level sophistication. Contrary to this ex-

planation, a strong correlation between mortality beliefs and savings persists

even after controlling for financial and numerical literacy, income, and demo-

graphic and educational differences. We also statistically estimate the scope for

omitted variable bias by applying the coefficient bounding procedure recently

introduced by Altonji, Elder, and Taber (2005) and Oster (2017). According to

this procedure, the estimated coefficients are robust and stable to informative

control variables, thereby strengthening our interpretation of the results.

4For example, just 52% of households had assets in personal retirement accounts as of 2008

(Poterba, Venti, and Wise (2011)), and there was a 4.3% annual increase in wealth among 70 to

75 year olds between 1998 and 2006, even after controlling for personal incomes (Love, Palumbo,

and Smith (2009)).

5The literature contains several compelling explanations for the drop in consumption at retire-

ment, including the nonseparability between consumption and leisure, and home production (see

Hurst (2008) for an overview).