Genetics and insurance
From a public policy perspective, so-called 'adverse' selection may sometimes not be adverse at all. The 'adverse selection' argument for genetic (or gender) discrimination rests on a mis-measure of the benefit of insurance to the population as a whole. A fall in the number of people insured can be consistent with a higher number of losses compensated by insurance, if more of the right people that is, the higher risks buy insurance.
Short versions of the argument: for actuaries, for economists, for lawyers, for general readers.
Teaching materials:
Video example (5 mins); Youtube channel.
Most accessible paper (written for lawyers).
10 slides (of which 3 key) for lectures (can be freely re-used or adapted, no attribution necessary).
Longer presentation
A University of Kent micro-site.
Main publications:
BOOK: Loss coverage: why insurance works better with some adverse selection
Why insurers are wrong about adverse selection
Written for lawyers full text open access
Insurance loss coverage and social welfare
Reconciliation of loss coverage to utilitarian concepts of social welfare full text open access
Insurance loss coverage and demand elasticities.
(Insurance: Mathematics & Economics, 2018)
General results for any downward-sloping insurance demand function
full text open access
Insurance loss coverage under restricted risk classification: the case of iso-elastic demand
(ASTIN Bulletin, 2016)
Comprehensive analysis for this demand function
pdf (29 pages)
Loss coverage in insurance markets: why adverse selection is not always a bad thing
Short version for a conference pdf (17 pages)
This paper was awarded a prize in June 2015 by
the International Actuarial Association!
Loss coverage as a public policy objective for risk classification schemes
(Journal of Risk and Insurance, 2008)
The first full publication of the idea
pdf (21 pages)
Some novel perspectives on risk classification
(Geneva Papers on Risk & Insurance, 2007)
An inchoate version of loss coverage (in a broader paper, also covering psychological and legal perspectives)
pdf (28 pages)
Other related publications (mainly genetics and insurance):
Genetics and insurance in the United Kingdom 1995-2010: the rise and fall of scientific discrimination
Background article on genetics and insurance
(Encylopaedia of Applied Ethics, 2nd ed., Elsevier, 2011)
pdf (14 pages)
Genetics and insurance: an actuarial
perspective with a difference
(Reply to Human Genetics Commission consultation, 2001)
An even earlier version from 2001: "...self-selection is at first order a positive
phenomenom..." (pp4-5)
pdf (15pages)
Some history:
My first paper on genetics and insurance was published in 1997 (see 1). Over the next several years around the millenium, I consistently argued that insurance companies should not be allowed to ask questions about genetic test results ( to the HGC, see 2). I wrote to the Government Actuary (see 3, 4, 5, 6, 7, 8) and to The Actuary magazine (see 9) advocating that the Actuarial Profession change its policy on numerous occasions in 1999 and 2000.
I did all this because there appeared to be potentially serious problems if insurers were allowed to ask routinely about tests, for exampleThere also appeared to be no realistic prospect of any material harm to the financial stability of insurers if their access to tests was banned. The remote possibility that this might one day change could be addressed if and when it arose.
Most actuaries and many others disagreed with me at the time, I believe largely because actuarial ideology suggesting the necessity or desirability of genetic discrimination distorted their view of the evidence. I argued at the time that this ideology was malign (see eg 10, 11) ; after another decade of evidence, it now looks increasingly ridiculous.
The ideology that genetic discrimination was necessary and/or desirable in private insurance led to the UK government establising a Genetics and Insurance Committee (GAIC) along the lines originally suggested by the Association of British Insurers. GAIC was designed to legitimise genetic discrimination, by approving applications from insurers to use genetic tests for insurance purposes. I criticised this concept on many occasions (see eg 12, 13, 14).
The criteria for GAIC approval were very favourable to insurers: essentially a test could be approved if it could be reliably shown to be of some commercial value to insurers. However, because of the lack of credible evidence, GAIC did not progress as the insurers or government envisaged. Over a period of 10 years only one application (for Huntington's disease) received approval. In 2009, the government announced that GAIC would be disbanded.
Archive of this page:
My older notes on genetic discrimination in insurance are retained on a separate page.
Old page on genetic discrimination and insurance Various notes from 2000 to 2009