The government's current policy of allowing state pensions to decline, while increasing reliance on occupational and other private pensions, will perpetuate the disadvantage that women face in providing adequately for their retirement. That is the central conclusion of new ESRC-funded research by Dr Jay Ginn, Co-director of the Centre for Research on Ageing and Gender at the University of Surrey:

'Women's typical sequence of paid and unpaid roles over the course of their lives limits their ability to accumulate private pensions. At the same time, the levelling effect of state pensions is being eroded.'

Dr Ginn has used British survey data to examine how women's patterns of paid and unpaid work influence their chances of building an adequate private pension of their own and how changing patterns of partnership and parenthood influence outcomes. The research shows that:

For the majority of women, their private pension disadvantage and hence their risk of personal poverty in later life is unlikely to diminish in the future. At the same time, new patterns of pension disadvantage are emerging, influenced by partnership status, parenthood, class and ethnicity.

For example, older women who are single enjoy a private pension advantage compared with women who were ever married. Compared with married women aged over 65, the chances that a single woman will receive any private pension income are nearly 7 times higher. The chances for widows are 4 times higher but for divorced women, they are only 1.5 times higher.

This 'single premium' is greatly reduced when comparing private pension coverage among mid-life women. It almost disappears in the youngest generation.

The effect of motherhood in reducing full-time employment, earnings and private pension coverage is weakest for the fifth of women with a university degree. Nevertheless, having children still substantially reduces their pension-building opportunities. The earnings of women graduates with a pre-school child are on average only 44% of the earnings of childless women graduates which has consequences for pension contributions.

The biggest negative impact of motherhood on private pension prospects occurs for mid-skilled women, those educated to GCSE level. These comprise about one third of working age women and the proportion is rising.

Among younger women, marriage and motherhood are diverging, with rising childlessness and childrearing outside marriage. The adverse effect that raising a family has on women's private pension coverage is magnified for lone mothers. Among women aged 20-39, divorced women's chance of having private pension coverage is less than half that of married women.

Divorced women begin to 'catch up' on lost employment, earnings and pension-building only when their children are independent, but they remain at high risk of poverty in later life. On average, children are 4-5 years old at the time of divorce, restricting their mothers' opportunities for earning and pension-building for a number of years. For this and other reasons, legislation allowing pension-sharing at divorce is unlikely to end divorced women's pension disadvantage compared with other women and divorced men.

The research prompts questions about the value of derived pensions versus independent entitlements. It suggests that the financial security of those who have devoted time to caring for their family is better protected by some form of carer credits than by rights based on legal marital status. But such protection can only be achieved through state pensions.

International comparisons reveal that the UK has a particularly harsh pension system for women, even among 'liberal', English-speaking welfare states, as the women-friendly features of state pensions introduced by Barbara Castle have been dismantled or rendered ineffective.

Women's disadvantage in accumulating private pensions is acknowledged in the government's Green Paper, Simplicity, Security and Choice: Working and Saving for Retirement, published last December. But no change in pension policy is proposed to remedy the situation.

Dr Ginn concludes that:

'Despite women's increased participation in employment, most cannot rely on private pensions to provide an adequate personal income in later life. Only improved state pensions with protection of caring periods, or alternatively a universal citizen's pension, can ensure that women's unpaid family care work does not lead to poverty and dependency in later life