On 21 December 2012, a ban on the use of sex as a criterion in pricing insurance came into force throughout the European Union. In a forthcoming article in the journal Regulation and Governance (pre-publication version available now), I have examined how the ban came about and what its significance is. The ban was originally proposed by the European Commission in 2003, but the Commission was persuaded to drop the proposal. Instead, insurers were required to publish the statistical basis for discriminating between men and women, to show that differences in premiums were justified by differences in risk. This compromise was successfully challenged in the Court of Justice of the EU by the Belgian consumer association, Test-Achats.
At the root of the debate about sex discrimination in insurance was the question of whether discrimination was necessary for the market to function efficiently. Insurers do not use every relevant piece of information when pricing their products. Some information is too expensive to acquire, some is already subject to legal prohibitions (such as information about ethnicity or religion) and some is subject to voluntary restraint (such as genetic information). Furthermore, long-established practices of risk classification become entrenched in insurance. Long use yields long data series that enable insurers to check that their assessment of risk is robust, and consumers get accustomed to conventional discriminatory practices.
What drew the European institutions to disrupt the convention that sex is used in pricing some forms of insurance? One answer is that the Commission was concerned about the rise of defined contribution pensions, where individuals buy an annuity with their accumulated pension savings. Women get smaller annuities than men because they live longer, on average. However, on a straightforward cost-benefit analysis, this did not make a strong case for banning discrimination, as women are likely to pay more now for car insurance. Furthermore, insurers can find other indicators for long life expectancy (such as occupation and family history), and they will find new ways of identifying safer drivers too.
A stronger answer is that discrimination was already prohibited in some states of the EU, so different countries had different rules despite the supposed existence of a single market in insurance. The Court of Justice held that there should be a common rule across Europe. Since non-discrimination is a fundamental right, it held that this should be the basis for regulating the single market.
While insurers resisted this strongly, arguing that it would disrupt the market, their reaction now that the matter is decided is rather muted. On the Today programme on 21 December 2012, the spokeswoman for the Association of British Insurers emphasised that the effects on premiums were unpredictable, that the market remains highly competitive, and that consumers should shop around. The fact is that insurers do not want an extended public debate about discrimination. They prefer the legitimation provided by market competition, which leaves insurers autonomy in determining their pricing strategies, subject to regulatory constraints. Public scrutiny is uncomfortable because it is prone to reveal that insurers’ practices have a weaker technical basis than they like to imply. During the debate on sex discrimination, statistical studies were done which suggested that premium differentials were not always explained by underlying differences in risk, and that some potential alternative predictors were ignored. This should not really surprise us. While insurers have competitive incentives to search for good predictors, they also have marketing reasons to set prices to the advantage or disadvantage of particular groups.
It is tempting to see the ban on sex discrimination as a step towards a more ‘social’ Europe, going beyond the creation of a free trade area and regulating market transactions for social purposes. However, it is not clear what the social purpose really is: the Court of Justice has upheld a principle rather than pursuing a goal. The case highlights that markets are based on conventions which come under scrutiny when they are exposed to cross-national comparison. The ban reconstitutes the European insurance market and adjusts the ‘playing field’ for competition, rather than addressing a social policy problem or promoting equality of outcomes.
(Woman next to car image via Shutterstock.)