The Gender Directive

1. The Gender Directive Implementation of data publication requirements in the United Kingdom Presentation to EC Forum on the implementation of Article 5 of Directive 2004/113/EC Heather Miller9 September 2009

2. Background • Gender Directive was implemented in the UK by means of the Sex Discrimination (Amendment of Legislation) Regulations 2008 amending the Sex Discrimination Act 1975. • From 6 April 2008, these regulations specify the conditions under which insurance companies are allowed to differentiate based on gender differences when pricing insurance policies.  • One condition is that the use of gender as a factor in the assessment of risk is based on relevant and accurate statistical data.  • A second condition is that this data must be compiled, published and regularly updated in accordance with guidance issued by HM Treasury.

3. Treasury guidance on data publication • Published March 2008 • Available on HM Treasury website • Produced on the basis of discussion with industry • Data is published by the Association of British Insurers and the Continuous Mortality Investigation organised by the UK actuarial profession

4. Data Publication Requirements • Only apply to products where gender is a factor in the calculation of premiums and benefits • Data must be published in a form that is intelligible to a non-expert • The publication must identify the source of the data and the period to which it relates. • Technical terms must be explained • Data requirements are set out in more detail in guidance for each main policy type

5. Products for which data publication is required

6. Nature of data publication • In single year age points or five year age bands • Longer age bands are permitted where populations are small: e.g. critical illness policyholders <30 years old or motor insurance policyholders > 80 years old • In some cases publish ratio of male to female outcomes (e.g. mortality rates), in some cases publish ratio of average costs per policy • Data should be reviewed and, if necessary, updated at intervals not exceeding a set period (3-4 years for most products)

7. Example 1: Motor insurance

8. Example 2: Private medical insurance • Average cost of a claim per policy for women as a percentage of average cost of a claim per policy for men for Private Medical Insurance, 2005- 2007 • This data is based on an average cost of a claim per policy provided by four insurance companies. To protect the confidentiality of the data, this is unweighted.

9. Considerations re publication schemes • Who is the audience for the data? • How will data be made publicly available? • How will competition be protected? • Who will collate the data? • How often will the data be updated and who by? • Are the cost and resources required worthwhile?

10. More information: Guidance: http://www.hm-treasury.gov.uk/d/consult_insurance070308.pdf Published data: http://www.abi.org.uk/Facts_and_Figures/Data_by_Age_and_Gender.aspx

11. ANNEX – Publication requirements for different insurance products

12. Life insurance and Annuities • Publication may be by way of a table or chart illustrating the ratio of male to female mortality for insured risks in the United Kingdom or an appropriate region. The published data must illustrate recent differences in mortality by gender and by age. It may be based on graduated data using single year age points or raw data by age ranges not exceeding five years up to age 75 and 10 year age ranges thereafter. • The published data may aggregate different forms of annuity and life assurances. It should be reviewed and, if necessary, updated at intervals not exceeding four years.

13. Critical illness insurance • Publication may be by way of a table or chart illustrating the ratio of male to female critical illness rates for insured risks in the United Kingdom or an appropriate region. The published data must illustrate recent differences in the incidence of critical illness by gender and by age. It may be based on graduated data using single year age points or raw data by age ranges not exceeding ten years for ages up to 30, five years up to age 80 and a single age range thereafter. • The published data may aggregate different forms of critical illness insurance. It should be reviewed and, if necessary, updated at intervals not exceeding four years.

14. Income protection insurance • Publication may be by way of a table or chart illustrating the ratio of male to female sickness, disability and unemployment for insured risks in the United Kingdom or an appropriate region. The published data must illustrate recent differences in the incidence and cost of claims by gender and by age. It may be based on graduated data using single year age points or raw data by age ranges not exceeding five years up to age 80 and a single age range thereafter. • The published data may aggregate different forms of income protection insurance. It should be reviewed and, if necessary, updated at intervals not exceeding four years.

15. Motor insurance • Publication may be by way of a table or chart illustrating the ratio of male to female average costs per policy. The data should be broken down by age ranges not exceeding five years, with a single age range for ages 80 and above. • Data may aggregate all forms of cover and for all relevant motor vehicles. It should indicate the accident years reported. It should be reviewed and, if necessary, updated at intervals not exceeding three years.

16. Private medical insurance • Publication may be way of a table or chart illustrating the ratio of male to female average claims costs in 5 year age bands, with all short term (up to five years) medical insurance products combined. • The published data should be reviewed and, if necessary, updated at intervals not exceeding three years.

17. Other types of policies • For other types of policy not covered by the categories above, and for new types of policy outside the categories above, publication may be by way of a table or chart illustrating the proportionate differences in risk between males and females. Where relevant, the data may reflect evidence from non-UK sources on which it is reasonable to rely. • The published data should be reviewed and, if necessary, updated at intervals not exceeding two years.