Private Information in Life Insurance, Annuity, and Health Insurance Markets

• Date: 01 October 2017
• Published date: 01 October 2017
• DOI: http://doi.org/10.1111/sjoe.12189
• Author: Amelie C. Wuppermann

Private Information in Life Insurance,

Annuity, and Health Insurance Markets∗

Amelie C. Wuppermann

University of Munich (LMU), DE-80539 Munich, Germany

amelie.wuppermann@econ.lmu.de

Abstract

Economic theory predicts that private information on risks in insurance markets leads to

adverse selection. To counterbalance private information, insurers collect and use information

on applicants to assess their risk and to calculate premiums in an underwriting process.

Using data from the English Longitudinal Study of Ageing (ELSA), this paper documents

that differences in the information used in underwriting across life insurance, annuity, and

health insurance markets attenuate private information to different extents. The results are in

line with – and might help to reconcile – the mixed empirical evidence on adverse selection

across these markets.

Keywords: Biomarkers; ELSA; health-related risks; insurance underwriting

JEL classification:D82; I13

I. Introduction

Information asymmetries about risk types in competitive insurance markets

are known to induce inefficient outcomes due to adverse selection. In

particular, the leading model by Rothschild and Stiglitz (1976) and various

extensions robustly predict that if individuals have private information on

their risks, higher risk individuals buy more insurance coverage than lower

risk individuals (Chiappori et al., 2006). However, the empirical evidence

of adverse selection is mixed (see Cohen and Siegelman, 2010, for an

overview), leading to doubts on the existence of private information – or at

least putting into question whether individuals use their private information

in their decision to buy insurance coverage.

To counterbalance individual information, insurers collect and use in-

formation on insurance applicants to assess their risks in the process of

insurance underwriting. Whether there is private information on risks that

∗The author would like to thank Sebastian Bauhoff, Luc Bissonnette, Florian Heiss, Hen-

drik J¨

urges, Helmut Farbmacher, Daniel McFadden, Erik Meijer, Nicolas Sauter, Beatrice

Scheubel, and Joachim Winter for insightful comments, and Deutsche Forschungsgemein-

schaft through GRK 801 for financial support. The analyzed data were made available

through the UK Data Archive.

©The editors of The Scandinavian Journal of Economics 2016.

Scand. J. of Economics 119(4), 855–881, 2017

DOI: 10.1111/sjoe.12189

are insured in different insurance markets thus depends not only on how

much individuals know about their risks but also on how much insurers

know through underwriting. In this paper, I test empirically whether in-

dividuals have information on their own health risks, and whether this

information remains private after underwriting in different health-related

insurance markets. That is, I test whether individuals have information on

their risks that exceeds the information that insurers collect and use for

underwriting. I then conduct an additional analysis to test whether indi-

viduals use their residual private information in the decision to purchase

insurance coverage.

I focus on information on risks insured in health insurance, life

insurance, and annuity markets. The rationale for choosing these specific

markets is that life insurance and annuity markets, on the one hand, and

different health insurance markets, on the other hand, insure similar risks,

but the evidence of adverse selection is mixed across these markets: there

is strong evidence of adverse selection in annuity markets (Finkelstein and

Poterba, 2002, 2004, 2014; Einav et al., 2010; McCarthy and Mitchell,

2010) but only limited evidence for life insurance markets (Cawley and

Philipson, 1999; Hendel and Lizzeri, 2003; McCarthy and Mitchell, 2010).1

There is evidence of adverse selection in group health insurance markets,

such as employer-sponsored insurance in the US, but only little evidence

for individual health insurance markets; see Cutler and Zeckhauser (2000)

for an early summary of the literature.

The differences in adverse selection across life insurance and annuity

markets can be explained by multiple dimensions of private information.

Private information on risk type and on risk preferences constitutes a promi-

nent example (Finkelstein and McGarry, 2006). Risk-averse individuals tend

to buy insurance coverage and to try to prevent risk, and thus they might

live longer. This counteracts adverse selection in life insurance but exacer-

bates adverse selection in annuity markets (Cutler et al., 2008). It is harder

to argue that multiple dimensions of private information explain differences

in adverse selection across health insurance markets that insure exactly the

same risk. Cohen and Siegelman (2010) suggest that the absence of useful

private information in some but not in other markets, or individuals’ in-

ability or failure to act upon private information might explain differences

in adverse selection across markets or market segments. In this paper I

contribute to this literature by documenting that underwriting differences

across health insurance markets indeed attenuate private information to dif-

ferent extents. Furthermore, I present evidence that individuals act upon

their private information if they can.

1However, He (2009) presents evidence of adverse selection in life insurance.

856 Private information in health-related insurance markets

©The editors of The Scandinavian Journal of Economics 2016.