5. Policy options for funded retirement savings arrangements to tackle the gender gap

The gender pension gap is the result of many interconnected and complicated factors relating to society, employment, childcare, education, and individual bias. Therefore, fixing the gender pension gap will necessarily require measures to address these other areas, in particular those relating to the labour market and its inequalities with respect to participation, pay, and the cost of care.

While the pension system itself cannot correct for all of these factors, its design should at least not increase inequalities, and at best should reduce the impact that existing inequalities can have on the retirement benefits that women will receive. Public pensions can incorporate various progressive formulas and subsidies in their design to offset some of these inequalities. However, the importance of retirement savings arrangements is growing around the world, so their potential contribution to the gender pension gap – currently at 26% in OECD countries – can be expected to increase going forward. Policy makers must therefore consider how these arrangements may contribute to the gap both today and in the future, and ensure that their design does not increase the gap, especially given the close relation of the retirement income from these arrangements to employment and income patterns.

This chapter first summarises the various drivers that directly contribute to the gender pension gap within retirement savings arrangements that previous chapters have identified, and how each of these drivers affects the retirement income that women can expect to receive. It then looks at various policy options to address each of these impacts. It concludes with a reflection on the current challenges facing policy makers to close the gender pension gap.

The gender pension gap is reinforced at every stage of preparing financially for retirement:

  • Women have less access to retirement savings arrangements.

  • Women less often participate in retirement savings arrangements.

  • Women contribute lower amounts and less frequently.

  • Women earn lower investment returns.

  • Women have lower benefit entitlements of their own.

  • Women have lower retirement income and need it to last longer.

There are two biological drivers that impact the amount of retirement income that a women can expect to receive. First, women give birth to children, and take time off work around the birth of their child, during which they may not have income to contribute to their retirement savings plans. Second, women on average tend to live longer than men, and so their retirement savings must finance a longer period.

However, the majority of characteristics of being a woman that contribute to the gender pension gap relate to social constructs and institutional factors. These have a large impact on how women participate in the labour market and their patterns of employment and pay. In addition, they also have other subtle influences on the choices that women themselves make with respect to the types of careers they pursue, their financial education, and attitudes towards saving and investing. All of these factors play a role in determining the amount of retirement income that women will be able to obtain.

Women’s role as caretaker of children and/or family members is often the source of the determinants of the gender pension gap that relate to the labour market. Women’s higher involvement in carrying out this unpaid work compared to men may lead to lower participation in the labour market, more breaks in their employment history, higher rates of part-time work, and lower salaries. This means that they are less likely to have access to a retirement savings arrangement, are less likely to contribute even if they do have access, and are likely to have lower and less frequent contributions. Furthermore, even if their spouse accumulates sufficient retirement entitlements to finance both partners’ retirement, women may still have fewer entitlements of their own, leaving them in a more vulnerable position in case of divorce or the death of their spouse.

Career patterns are a major factor in the gender pension gap. Women in the OECD have a career length just two-thirds that of men. Women also make up 70% of part-time workers in the OECD (OECD, 2019[1]). Career patterns are the largest contributor to the gender pension gap in the United Kingdom (Jethwa, 2019[2]). The OECD analysis of Germany and the United States indicates the likely influence of the caretaking role and career breaks on the gender pension gap, as gaps in entitlements do not emerge until women are in their 30s – the age when they are most likely to take a break from full-time employment to care for children (Chapter 3).

Lower incomes are another leading driver of the gender pension gap. The average gender pay gap for full-time employees in the OECD currently stands at around 13% (OECD, 2020[3]). Chapter 3 shows that income differences are a main driver of the gender pension gap in Finland and the United States.

The fact that women tend to earn less than men, increases their likelihood to leave full-time employment to be caretakers, and the link between pay and employment patterns becomes a circular problem. The lack of affordable childcare in many jurisdictions means that women may not see much financial benefit in returning to work full-time after maternity leave. An example calculated in the context of the Australian system shows that 90% of the additional income of working a fourth day a week on a AUD 60 000 annual salary (around 20% less than the average salary) would be lost to care costs, taxes, and lost welfare payments, and there would be zero financial gain from working a fifth day (Dale and St John, 2020[4]). Low salaries therefore lead women to take more part-time work or career breaks to be caretakers, which in turn leads them to earn lower salaries. Caretaking also increases a woman’s reliance on the salary of the spouse, ultimately leading to a lack of her own entitlements to retirement income.

The lack of their own retirement income entitlements puts women in a much more vulnerable position in the case of divorce, and the reduction in retirement income is larger for women than for men following divorce. The financial impact of divorce on retirement income in the United Kingdom is 50% for women, compared to 33% for men (Jethwa, 2019[2]).

While factors linked to participation in the labour market are the main drivers of the gender pension gap, other secondary factors come into play and are also worth recognising. Lower levels of financial education and literacy may lead women to engage less in retirement planning. Women demonstrate lower levels of financial knowledge in the majority of OECD countries (OECD/INFE, 2016[5]). Women are also underrepresented in the fields of science, technology, engineering, and mathematics (STEM). This trend seems to be driven by stereotyping and expected gender roles rather than actual interest and ability, as the educational paths of boys and girls only start to diverge around age 15 (OECD, 2017[6]). Nevertheless, this divergence ultimately leads females to industries where employers are less likely to offer an occupational retirement savings arrangement, with one exception being the public sector where women are highly represented and have access to an occupational plan. Chapter 3 shows that lack of access is a significant driver of the difference in coverage of these types of arrangement in the United States, and therefore also for the gender pension gap.

Women also frequently demonstrate higher levels of risk aversion than men, which can translate into a preference for lower-risk investments and therefore lower returns on their retirement savings. Chapter 2 presents evidence that this is linked to differences in attitudes towards risk and competition that are shaped by societal factors and expectations rather than an inherent difference in preferences (Croson and Gneezy, 2009[7]; Gneezy, Leonard and List, 2008[8]). This bias can be reinforced by financial advisors who may be influenced by gender stereotypes (Roszkowski and Grable, 2005[9]).

Nevertheless, some of these societal constructs do seem to be changing over time, leading to a gradual reduction in the gender pension gap in a number of countries. Women are participating in the labour market at higher rates. The proportion of working-age women having a job in the OECD has increased from 46% in the 1990s to 52% in 2017 (OECD, 2019[10]). Changing attitudes towards the role of women has aided this trend. The proportion of the UK population supporting the idea that women should stay home and care for children has decreased by 35 percentage points since the 1980s (Government Equalities Office, 2019[11]). Women’s education is also improving, with 57% of bachelor’s and master’s degrees across the OECD obtained by women in 2014 (OECD, 2017[6]). These trends have contributed to higher participation of women in occupational retirement savings arrangements. The gap in coverage for occupational arrangements is shrinking in Germany, and females aged 30-40 now have higher participation in occupational arrangements in the United Kingdom than men (Now Pensions, 2019[12]).

Despite these positive developments, the gender pension gap remains significant and needs to be reduced. The following section discusses some of the options available to address the drivers of this gap within the funded pension system.

There are numerous options for retirement savings arrangements to avoid exacerbating the negative impact on retirement income from the drivers of the gender pension gap summarised in the previous section. While they cannot address the drivers themselves in all cases, the design of the plans should at least account for and accommodate gender differences that can lead to lower eligibility, lower participation, lower and less frequent contributions, less regular career patterns, lower returns, lower individual rights, and lower retirement income. As such, they should aim to have a gender neutral design.

Women have less opportunity to access retirement savings arrangements. This is largely because women are more likely to work in industries that do not provide access to an arrangement and are more likely not to meet eligibility requirements defined in terms of minimum salary or hours worked.

The proportion of women covered by an occupational arrangement is lower than for men in many countries. One reason for this is that employers are less likely to offer such arrangements in the private sector industries where women tend to work.

Mandating employers to establish an occupational arrangement is one way to improve access for women, even if participation is not mandatory for the employee. Several countries demonstrate a negative relationship between the proportion of women in a sector and the availability of occupational plans in that sector (Chapter 1). Chapter 3 shows that one of the main drivers of the lack of participation in occupational arrangements by women in the United States – where employers are not required to set up a plan – is that the industries in which women are more likely to be employed are less likely to offer an occupational arrangement.

As an alternative to mandating the offer of an occupational plan, the government could provide incentives for employers to establish occupational arrangements for their employees. The SECURE Act in the United States, for example, offers a tax credit for small employers to help cover the costs of setting up a plan and educating their employees about it, and encourages them to establish Multiple Employer Plans with other small employers to mitigate the administrative expenses.

Increasing the availability of personal arrangements could also improve the participation of women. The coverage gap for personal plans is much lower than that for occupational arrangements, and participation by women in personal plans is equal to or higher than by men in several countries. This indicates that the availability of personal plans allows women who would like to save for retirement, but may not otherwise have access to an alternative arrangement, to do so.

Eligibility requirements to participate in occupational arrangements should be relaxed in order to be more inclusive of women. Many countries impose criteria for employees to be eligible to participate in the retirement savings arrangement offered by the employer that disproportionally impact women. Such requirements are usually expressed as a minimum income threshold or a minimum number of hours worked (OECD, 2019[1]). As such, they are more likely to exclude women from participating, as women tend to earn lower salaries and more often hold part-time jobs. The impact of income thresholds is evident in the United Kingdom, where automatic enrolment has succeeded in increasing the participation of eligible employees, but had little impact on the 10 percentage point gap in participation between men and women among all employees (Chapter 4).

Measures should also be in place to encourage women who have access to a retirement savings arrangement to participate in and contribute to it. This can be done by nudging and providing incentives to participate, as well as engaging women in retirement planning with targeted educational workshops and communication that convey the importance of having their own savings for retirement.

Mandatory participation for employees is effective at reducing the coverage gap and furthermore automates the link between women’s increased participation in the labour market and their participation in a retirement savings arrangement. For example, women are not underrepresented in occupational arrangements in Finland, where participation in an earnings-related arrangement is mandatory for most types of workers, even the self-employed. In the United States, female participation is higher in plans where contributions are mandatory (Chapter 3). Women’s participation in the mandatory superannuation in Australia has increased along with their participation in the labour market.

As an alternative to mandates, automatic enrolment into a retirement savings arrangement can also increase women’s participation. The requirement in the United Kingdom to automatically enrol eligible employees in a workplace pension scheme succeeded in eliminating the participation gap between eligible males and females, which stood at three percentage points when the policy was introduced. Where automatic enrolment is not required, governments can provide incentives to encourage employers to adopt it. In the United States, the SECURE Act provides small businesses with a tax credit for implementing automatic enrolment of employees into their occupational scheme, on top of the tax credit for establishing a plan.

Subsidies to new members of retirement savings schemes can be helpful to encourage low-income individuals – and therefore also women – to have an account and start saving for retirement. Several countries, in particular those having automatic enrolment (e.g. New Zealand, Poland, Turkey) have introduced an initial one-off subsidy that individuals receive if they do not opt out of the retirement savings scheme. While the initial “kick-start” subsidy for the KiwiSaver in New Zealand has been discontinued, it contributed to the initial take-up and popularity of the plan, even for those who were not automatically enrolled. The scheme has also been successful among women, with over 50% of participants being women (Inland Revenue, 2019[13]). Nevertheless, their participation cannot be attributed solely to the kick-start subsidy, as the proportion of women participants has not significantly dropped since it was eliminated, and matching contributions also likely contribute to the KiwiSaver’s popularity.

Chile also pays a subsidy to new low-income members of their defined contribution system, but rather than a one-off payment, the subsidy is for workers who join the system in their early working years. Individuals between the ages of 18 and 35 having a wage lower than 1.5 times the minimum monthly wage are eligible for the subsidy for each contribution that they make during the first 24 months. As such, it intends to encourage low-income individuals to start to contribute at a young age, thereby having a larger long-term impact on retirement savings. When the subsidy was introduced in 2011, more than half of the recipients were women (Hinz et al., 2013[14]).

Improvement in women’s financial literacy and knowledge of the retirement system is needed to overcome their reluctance to deal with financial matters and to close the gap in financial knowledge relative to men. In line with the OECD Recommendation on Financial Literacy, programmes to improve women’s financial literacy should be modified to be relevant for women. Chapter 2 shows that several jurisdictions have developed financial education programmes that specifically target women to help them to understand the importance of saving and preparing for retirement (OECD, 2013[15]). New Zealand has a Women in Super programme that organises meetings and events to educate women on the superannuation system and their specific retirement needs. Singapore’s programme Financial Education for Mature Women targets middle aged women to help them prepare to be financially independent in older age.

Educational efforts will also need to help women to overcome their lower levels of confidence with respect to financial matters. Education involving peer groups and providing financial advice can help women to overcome their lack of confidence (OECD/INFE, 2013[16]). A successful example of a financial education campaign is the EMPOWER (Embracing and Promoting Options for Women to Enhance Retirement) programme in Wisconsin (Chapter 2). The programme developed communication materials targeted specifically at women and organised educational sessions for women only. The programme succeeded at increasing women’s participation in the occupational plan by 2.6%, having a significant impact on the participation gap, and was particularly successful among younger women and those with lower earnings (Anderson and Collins, 2017[17]). Another example is Laborfonds, an occupational pension fund in Italy, which organises educational sessions to explain the benefits of the fund and how it works (Chapter 4). They have succeeded in increasing women’s participation by targeting industries with a large representation of women. Other organisations focus more on the provision of advice, such as the women’s group Frauenzentrale Zurich in Switzerland, which has set up a popular pension advice session for women (Leybold-Johnson, 2017[18]).

The way in which the importance of saving for retirement is communicated can also have a large influence on how successful the messages will be in encouraging women to save. One study showed that messages framed in a positive way to emphasise the benefits of saving for retirement were much better received by participants than negative messages emphasising the risks of not saving. For those not yet saving, negative messaging was particularly ineffective. Interestingly, this was the case for both women and men, even though women used more negative language when discussing saving for retirement (Behave London, 2019[19]).

Women contribute less to their retirement savings plans due to lower wages and a higher likelihood of being in part-time work. Career breaks linked to caretaking can also lead to contribution gaps and shorter contribution periods for women compared to men. This can be addressed through higher contributions from employers, employees, and/or spouses; financial incentives to contribute; subsidies for caretaking activities; allowance for catch-up contributions; tailored fee structures; and targeted communication efforts.

A lack of employer contributions to employee’s retirement savings schemes can contribute to lower retirement savings for women. For example, Chapter 3 shows that in the United States women are less likely to be members of plans into which employers contribute.

Eligibility requirements for employees to receive employer contributions to their retirement savings plans should not penalise low-income women. Australia, for example, has a minimum income threshold under which employers are not required to make the mandatory superannuation contributions for their employees on top of their paid wages. Total remuneration across multiple employers is not taken into account. Low-income women therefore miss out on this additional compensation.

A total remuneration approach to employee compensation could help to ensure that low-income women in particular are not completely excluded from employer-provided retirement income benefits and miss out on this compensation that they would otherwise receive. This could be a problem, for example, where the scheme has voluntary employer contributions and the employee is not eligible to participate (Dale and St John, 2020[4]). With a total remuneration approach, compensation takes into account the total monetary value of all benefits received. Any contributions to a retirement savings scheme would be deducted from this amount, so ineligible individuals would still be entitled to the compensation even if it is not paid to the retirement savings scheme.

Employer contributions can also be encouraged through financial incentives. In Germany, for example, employers making contributions on behalf of low-income employees receive a tax allowance.

An additional source to increase contributions levels for women is their spouse. Permitting spousal contributions would allow the spouse to make contributions to their partner’s retirement savings to compensate for lower salaries or for any time out of work to perform caretaking responsibilities. Many countries allow for such contributions, either to the partner’s account directly (e.g. Australia, Hungary, Lithuania, Spain, the United Kingdom) or to a separate account set up by the spouse (e.g. Canada, the United States). The same contribution limits typically apply, that is the spousal contributions count against the contribution limits for the recipient spouse, including any contributions they have made themselves. As such, the same financial incentives in place to contribute to one’s own plan also apply for the spouse’s plan (e.g. tax deductibility).

Where there are financial incentives to contribute to a plan, allowing for spousal contributions may lead to this option being used solely to optimise taxes at the household level. Indeed, the few individuals making use of this option in Australia tend to earn higher incomes, and therefore likely benefit more from the tax concessions provided. As such, the benefit for the majority of women is likely to be minimal.

Spousal contributions have the large benefit of allowing women to accrue their own individual rights rather than rely solely on the entitlements accrued by their spouse on behalf of the household. In this way the receiving spouse is less penalised by the service she provides for the household through unpaid activities such as caretaking.

As women are more likely to have low incomes, they are also more likely to be able to benefit from financial incentives targeted directly at those with low incomes. Such incentives can encourage them to contribute more and more regularly. They most often take the form of tax credits – either for the low-income earner or their spouse – or matching contributions and subsidies.

Tax credits targeting low-income earners allow women with lower incomes to contribute higher amounts, all else equal. Several countries (e.g. Australia, Korea, the United States) provide larger tax credits specifically for low-income contributors.

Financial incentives can also encourage spouses to make additional contributions towards the retirement savings of their low-income partner. For example, the Spouse Super Contribution Tax Offset in Australia provides a tax credit to the spouse contributing on behalf of their non-working or low-income partner.

Matching contributions or subsidies can also target low earners, providing an additional incentive for them to contribute regularly to their retirement savings plan and increasing the amount that they can accumulate. For example, in Australia the government matches voluntary contributions by low-income earners, which seems to be effective at increasing contributions for this group. A reduction in the match rate and maximum benefit corresponded with a reduction in contributions, and while low-income groups are less likely to make voluntary contributions, those that do have higher contribution rates than other income groups (OECD, 2018[20]). Another example is in Germany, where the government provides subsidies for savers in the Riester plans that provide higher relative benefits for those with low-income, and as such have been effective at attracting low-income earners (OECD, 2018[20]). Furthermore, over the period 2008 to 2013, women received the majority of these subsidies (Klammer, 2017[21]).1

Subsidies for having children and caretaking leave can help to counter the negative impact on retirement income for women who have children. These can take the form of contributions paid during maternity leave, potentially extended for a longer period of parental leave, to those who work part-time to care for children, or per-child subsidies.

Ideally, contributions for women on maternity leave would continue at the same rate as when they were working. In practice, the benefit to women varies depending on who pays the contribution and the earnings base on which contributions are made. In most countries, contributions to mandatory, quasi-mandatory, or occupational pension arrangements can continue during maternity and parental leave (Chapter 4). Often, contributions towards mandatory plans are taken over by the government. However, in some countries the contribution is voluntary, paid by the plan sponsor, or the employees themselves. Furthermore, contributions are not always based on the woman’s full salary, but rather a percentage of salary or a flat amount, though in some cases contributions can be even higher than when women were working. Some benefits are only for the period of maternity leave, while others continue during parental leave and/or up to a maximum duration.

Contributions may also continue during time off of work for caretaking that does not necessarily follow the birth of a child, though this is less common. In Chile, for example, parents are entitled to take medical leave with their full salary to care for a child with a serious illness. A study in the United Kingdom showed that a policy to top up pension contributions for carers – including both those working part-time and those caring full-time – had a significant impact on the gender pension gap, as women are typically the ones working less to care for children (Jethwa, 2019[2]).

Per-child subsidies paid by the government, usually over a fixed period of time, are another way of compensating for the child penalty on retirement savings regardless of the time taken off of work, and is an approach found in several countries. However, if offered unconditionally, this type of subsidy may be less effective in promoting retirement savings over the long term. While the child subsidy in Chile has been effective at increasing the number of women with a retirement savings account, it has not led to increased contributions (Chapter 4). Similarly, the balances in retirement savings accounts for women in Korea who received a child subsidy remained low because they did not make additional contributions afterward (Hinz et al., 2013[14]). In contrast, Germany requires a minimum level of contributions to the Riester plan in order to receive the child subsidy, which encourages regular savings behaviour.

Allowing any contribution limits to be carried forward to future years would allow women to be able to make up any lost contributions during time off work for maternity leave or caretaking. Several jurisdictions allow for this, with the period that can be carried forward ranging from three years (the United Kingdom) to indefinitely (Canadian registered retirement savings plans).

Communication and education for financial matters is more effective if it is personalised, and if made in a timely manner corresponding to specific life events and “teachable moments” (OECD, 2019[22]). Communication to encourage women to contribute to their retirement savings plan should therefore be tailored to their specific situation and period of life.

Personalisation of communication can go a long way to better capture the attention of the individual and help them to understand what action they could or should take given their situation. In one example carried out by Mercer, employees were sent short, personalised videos about the impact that additional contributions could have on their retirement savings and income in retirement (Chapter 4). The videos also linked the level of contribution and retirement income with the expected quality of life that they could provide in retirement. Women were particularly receptive to this type of communication, and were significantly more likely to watch the video to the end, as well as more likely to increase their pension contributions after they watched it (Read, 2017[23]).

The timing of the communication also matters, and can be more effective if provided at the relevant moment in a woman’s life, such as around the birth of a child. Nationwide Pension Fund in the United Kingdom provides a link on their website with specific information for new parents, explaining under what conditions pension contributions and insurance coverage will continue. It also takes advantage of this moment to encourage new parents to review their beneficiary information in case of death (Nationwide Pension Fund, 2020[24]). The Finnish Centre for Pensions provides a calculator for new parents to understand the financial impact of the length of the parental leave that they take, showing at which point their benefit accrual will start to suffer if they take a longer period of leave. Verve Super, a pension fund for women in Australia, offers free coaching and guidance, and helps women with decisions relating to specific questions, such as contributions during parental leave.

Women are more likely to have shorter careers and career breaks, as well as low and irregular contributions to their retirement savings plans. This is driven by lower incomes on average and taking time off from full-time work for caretaking. The design of retirement savings arrangements could better accommodate the career patterns of women by being more flexible with respect to contribution levels, improving the portability of occupational arrangements and adapting the fee structure to accommodate lower balances.

Given the career pattern of women is not always constant, contribution rates should be able to vary over time to allow women the flexibility to balance immediate needs with long-term savings goals, and to contribute more when they are better able to. This could be done by allowing women to choose from a range of contribution levels, as is the case for the KiwiSaver plan in New Zealand. Allowing for contribution holidays could also be helpful to allow a temporary break in contributions during periods of greater financial difficulty, particularly for low-income caretakers.

Women are more likely to move in and out of the labour market to accommodate the caretaking needs of their family. This means that they may not have access to a plan during these periods, and when going back to work they may have to contribute to a different scheme with the new employer.

Being able to contribute to the same plan regardless of whether they are employed or not would facilitate more continuity in retirement savings for women who take time off from working. Nevertheless, this would need to be accompanied with the allowance for flexible contributions in order to be most effective. Requirements around the frequency and level of contributions seem to be a barrier for individuals to make use of such arrangements in practice (OECD, 2020[25]).

For women changing employers, it is also important to avoid the accumulation of several accounts linked to each employer to help women build up more retirement savings in total. Having several small accounts could result in higher overall fees, unnecessarily deteriorating women’s retirement savings. Furthermore, having several accounts impedes active engagement with retirement planning as it reduces the visibility of the future retirement income potential that the retirement savings will be able to provide.

Ensuring that people are able to contribute to the same plan even when changing employers is one way to avoid several small accounts. This could be done by having a centralised institution managing the collection and payment of contributions, so that they can direct all contributions on behalf of a given member to the pension fund of her choice. This is the approach taken for the KiwiSaver in New Zealand and the Premium Pension in Sweden. Another approach is to de-link the choice of provider from employment, so the individual can choose to which account all of her contributions are paid, as is done in Mexico.

An alternative to having a single account would be to facilitate the transfer of the existing account when changing employment. This would be more effective if done automatically, as employees do not always take the necessary steps to make the transfer. In Australia, the tax office plays a central role in facilitating the consolidation of small accounts and preventing individuals from opening new accounts unnecessarily. It can track small inactive accounts to consolidate them into the current active account, and now automatically provides new employers with the employees’ active account details. Their website also includes a “consolidate my accounts” button which automates the consolidation for savers.

Smaller and less regular contributions due to lower incomes and periods away from work can mean that the balance in women’s retirement savings accounts starts very low and grows slowly. As such, fixed fees that the provider charges can have a relatively large impact on the balance of the account, and be a barrier for women to accumulate retirement savings.

Fee structures for retirement savings accounts could adapt to help small savings accounts to grow and be more forgiving to lower contributions during time off of work for caretaking. In particular, fee structures should avoid charging fixed fees, particularly for small balances.

Women tend to demonstrate more risk-aversion than men – a bias which is often reinforced by financial advisors – leading them to invest in more conservative strategies that offer potentially lower returns. Appropriate default investment strategies and objective assessments of individual risk tolerance can help women overcome their conservative bias.

Nudging women towards more balanced-risk investment options would help to overcome their tendency to select a more conservative strategy and take on a reasonable level of risk for the long-term horizon when saving for retirement. Numerous studies have shown that most retirement savers will remain in the default investment option that their savings plan offers, and offering an appropriate default investment option is a key recommendation of the OECD Roadmap for the Good Design of Defined Contribution Pension Plans. The default strategy should therefore be one that appropriately balances risk with the need to protect the retirement savings from severe market downturns. Indeed, there is some evidence that the introduction of qualified default investment alternatives in the United States, has reduced the difference in investment returns between men and women (Garnick, 2016[26]).

The appropriate default option should aim to be the optimal strategy for the typical career of women participating in the plan. Lifecycle default strategies are common in most OECD jurisdictions. Such strategies balance risk-taking with protection by gradually reducing the exposure to risky assets as the individual approaches retirement. However, other strategies could be considered that better account for women with low-income or those who take more career breaks. For example, alternative assets could potentially provide stable returns in the long term without the need to de-risk, offering a better solution for those having less regular contributions and an uncertain retirement date (Pensions Policy Institute, 2019[27]). Other options could aim to optimise the risk of the default strategy taking into account the expected importance of other sources of retirement income, namely the public system.

Objective assessments of risk tolerance could help women to avoid biased investment recommendations that are more conservative than appropriate for their actual risk appetite. According to the OECD Recommendation on Financial Literacy, women should have access to appropriate, independent and adequate financial advice. Financial advisors may be influenced by gender bias and reinforce women’s tendency towards risk aversion by advising them to invest more conservatively regardless of their demonstrated risk tolerance. To overcome this bias, assessments of risk tolerance could be automated to provide an objective view of risk appetite. Numerous robo-advisors already implement such assessments on their platforms to determine the range of investment options and vehicles that would be appropriate for the saver (e.g. Yomoni in France). As digital investment platforms become more prevalent, such assessments could be more easily implemented for retirement savings.

Not having sufficient benefit entitlements of their own puts women at increased risk of poverty in old age. While income resources are usually shared between spouses, which compensates for women having lower incomes and taking time off of work to care for children, this is not always the case for retirement benefits and savings. As a result, many women may lose access to or part of those retirement benefits if they divorce or upon the death of their spouse. To ensure that women share a part of their spouse’s retirement benefit accrual as well as their income, entitlements can be split either while the spouse accumulates them or upon the divorce of the couple. Communication around the options available should increase women’s awareness of the possibility and importance of splitting retirement assets upon divorce.

Individual benefit entitlements could be increased for women by allowing members to transfer part of their own retirement income benefit entitlements to their spouse. The split can be made at any point in time, from the initial contribution to the accrued benefits to the withdrawal of assets or payment of benefits at retirement.

Allowing for a split of entitlements as they are accrued ensures that spouses will have their own entitlements throughout the accumulation period. Australia allows individuals to divert a portion of their mandatory contribution to the superannuation system to a spouse’s account. Sweden allows a portion of entitlements earned over the last calendar year to be transferred to a spouse’s account once per year, though a charge is levied on the transfer to recognise that they most often are transfers from men to women, who can expect to receive retirement income benefits over a longer period of time. Iceland offers significant flexibility to share benefit entitlements with spouses. The split can be done during accumulation for past and/or future benefits in accrual, or alternatively the split can apply to retirement income based on current and/or future accruals. However, the take-up of this option remains very low, with less than 0.1% of accounts splitting their entitlements (Chapter 4).

Alternatively, the split can be executed upon termination of the accumulation of entitlements, whether this is due to leaving the plan or to retirement, though protections to the spouse in case of death may still be in place before that. The transfer of assets accumulated at retirement to a spouse’s account is one withdrawal option available in Portugal. The Netherlands allows accumulated entitlements to be split either when a member leaves the arrangement or up to a year before the arrangement pays a retirement income. If the member does not request a transfer within a certain time period, benefits will automatically be paid to a partner upon the death of the member when the partner reaches pensionable age.

Facilitating the split of benefit entitlements upon divorce will help to protect women who have relied on the income of their spouse and therefore accumulated fewer of their own entitlements, and allow them to have a higher retirement income than they could have on their own. Most countries allow for the split of entitlements upon divorce, and view retirement benefits as the joint property of the couple.

Explicitly requiring a split of retirement assets upon divorce, not only total assets, would better contribute to women’s financial security in retirement. Making the split automatic, as the Netherlands is planning to do, would further facilitate the transfer of entitlements to women and ensure that they receive the retirement benefits that they should be entitled to. Alternatively, certain criteria could determine whether the split is justified. In Denmark the split of benefits is considered when there is a large difference in retirement benefits accrued by the spouses or if one spouse has saved less for the sake of the family.

Relying solely on court orders to split retirement assets at divorce – the approach taken in a majority of countries – does not often result in a split. Oftentimes women are not aware that this is a possibility, and lawyers may be reluctant to promote this option due to the complexities involved in splitting some types of benefit entitlements. Less than 4% of divorce orders in 2018 in the United Kingdom included an attachment order to divide the pension entitlements accumulated during marriage once they are in payment (Now Pensions, 2019[12]). Nevertheless, this rate is higher when considering all types of pension split orders. In 2011-2012, a sample of cases indicated that 14% of cases accounted for a split of pension assets in some manner (Woodward and Sefton, 2014[28]).

Splitting accumulated entitlements at the time of divorce rather than splitting future benefits when they will be paid may better ensure that women receive the benefits they should be entitled to. This allows for a clearer split based on entitlements accumulated during marriage, and avoids complications relating to age differences and the timing of retirement income payments as well as the risk of losing benefits altogether in case of the death of the ex-spouse. Nevertheless, for entitlements that are more difficult to split during accumulation due to a lack of individual rights, the law should allow for adjustment of the retirement income to be paid to accommodate age differences.

Practical considerations as to how the split is made for non-defined contribution (DC)-type arrangements should also be taken into account. In theory the split can either be done by splitting entitlements within a plan, thereby adding the spouse as a member of the scheme, or by transferring entitlement to another plan. The former solution may present some logistical challenges that could increase the cost of the split, reducing the benefit for the spouse (Serenelli, 2020[29]).

Information on the possibility and the process of splitting retirement assets upon divorce should be available to women at the time of their divorce. Women need to be more aware of the possibility of splitting retirement assets, as well as the impact that not splitting these assets can have on their retirement security. As with communicating on the financial impact of taking time off to care for children, providing this information at a time that corresponds with relevant life events – both at marriage and at divorce – could improve the impact that this information has on the financial decisions of women. Nationwide Pension Fund provides a good example of communication with their Pensions and Divorce Leaflet, which describes the options available and the procedure to follow in simple terms (Nationwide Pension Fund, 2020[24]). It also includes a reminder to update beneficiary information, which could change following divorce.

Women can expect to spend a longer period of time in retirement, putting them at a disadvantage relative to men as to how long their retirement savings have to last. For defined contribution-type arrangements, this often translates into a lower monthly retirement income for the same account value, and an increased risk of a loss in purchasing power. Options for the design of the system that could help to address this inequality include equalising retirement ages, basing retirement income on unisex mortality rates, providing an explicit subsidy to women, promoting survivor income benefits, and encouraging the availability of pay-out options that increase retirement income over time.

Women should not systematically retire earlier than men given that they can already expect to spend a longer time in retirement if retiring at the same age. Retiring earlier than men can put women at a disadvantage by shortening the amount of time they have to accumulate benefit entitlements and increasing the length of time that benefits are paid. While most OECD countries have the same retirement age for both genders, several still have an earlier retirement age for women or reduce the retirement age for women who have children.

While the targeted retirement ages should not be different between men and women, there should still be some flexibility around the allowed retirement age to allow women to accommodate potential caring responsibilities. Women nearing retirement age can often have caretaking responsibilities for elderly parents, an older husband or grandchildren. Flexibility around the age at which they retire can therefore be valuable to them to be able to balance their financial and family needs.

Calculating retirement income based on unisex mortality rates would equalise the amount of income that women and men receive, all else equal. Nevertheless, this solution cannot be effectively applied in all retirement schemes. First, unisex mortality will only be effective in collective arrangements that can benefit from longevity risk pooling. Using unisex mortality rates for the purpose of calculating programmed withdrawals would simply increase the risk that women would run out of assets in retirement. Second, unisex mortality rates are not likely to be effective at increasing women’s retirement income where annuitisation is voluntary within the scheme. In the United States, annuitisation by men within occupational DC schemes – where unisex rates are obligatory – is rare because males can simply transfer their assets out of the scheme and purchase an annuity offering a higher income priced with gender-distinct rates. This effectively results in the calculation of retirement income using female mortality rates, with no additional benefit to women from the requirement for unisex rates (OECD, 2016[30]).

Where annuitisation is mandatory, an additional benefit of requiring a retirement income based on unisex mortality rates is that it may also encourage women to save more for retirement. Following the introduction of unisex pricing in Germany, women seem to save in Riester plans – which require the annuitisation of the accumulated balance at retirement – beyond their normal savings behaviour (Jusufovic, 2015[31]).

Direct subsidies, either during accumulation or at the point of retirement, could compensate women for having higher life expectancies and increase the level of retirement income that they receive. However, such subsidies could potentially have a negative impact on women’s wages and employment depending on how they are financed. Some employers in Australia voluntarily pay an additional contribution to the retirement savings plan of female employees as a way of promoting gender equality and positioning themselves as an attractive employer for women. As the additional contribution is voluntary, there is no negative impact on the employability of women. Nevertheless, gender discrimination laws that prohibit offering different compensation depending on gender could be a barrier to such subsidies in many countries. As an alternative, the government could provide a subsidy. A recent reform proposal in Chile included a subsidy to women at the point of retirement to increase the amount of retirement income that they would receive.

Having pay-out options that include the payment of an income to a surviving spouse can protect women having a partner from the risk of poverty in old age. A pay-out option offering survivor income may not always be available, particularly in less developed retirement markets that are not yet mature. In contrast with inheriting assets as a lump sum, survivor incomes will also protect the surviving women from the risk of outliving those assets. Nevertheless, survivor income is typically lower than the retirement income paid while the primary beneficiary is alive, so this solution is not a substitute for increasing women’s own benefit entitlements.

Where joint benefits are available, take-up can be encouraged through the use of hard or soft compulsion. Several OECD countries with DC arrangements require survivor income benefits for spouses. For example, married men choosing an annuity option in Chile are required to take a joint-and-survivor annuity rather than an individual annuity. Behavioural nudges can also be effective at increasing the take-up of joint and survivor annuities. The United States significantly increased the proportion of retirees taking survivor benefits through their occupational defined benefit plan by imposing this option as a default for married individuals and requiring the additional administrative step of the agreement of both spouses to opt out of this option (Chapter 4). Part of this increase, however, may also be attributed to an increase in the availability of this option.

Pay-out solutions that offer increasing payments over time can protect individuals from a loss of purchasing power over time. Because of their longer lives, women are more exposed to the risk of losing purchasing power due to the compounding effects of inflation. Options that provide indexation to an inflation measure are the most effective solution to protect women from inflation risk. Most jurisdictions allow for this option at least on a discretionary basis, if not a mandatory one. Nevertheless, indexation to inflation may not necessarily be the most efficient option for individuals to hedge the risk of a loss of purchasing power in retirement. Where inflation risk is not easily hedged, such options can entail significant risk to the providers and therefore involve substantial cost to the individual (OECD, 2016[30]).

As an alternative to full indexation to inflation, arrangements relying on risk sharing, where retirement income is not fully guaranteed, can aim to increase payments over time depending on investment and longevity experience. Such arrangements are likely to be able to pay higher retirement incomes as well due to the lower risk exposure for the provider (OECD, 2020[25]). Numerous countries have retirement income arrangements that aim to keep up with inflation. The bonus policy for the ATP in Denmark includes this objective. Another example are conditional indexation arrangements in Canada, which tie the indexation of retirement income payments to the financial strength of the scheme.

Much progress has been made in societies over the last decades in reducing the gender differences driving the gender pension gap. Societal attitudes about the role of women as homemakers and caretakers have been evolving. This has contributed to an increase in female participation in the workforce, leading to increased access to retirement savings schemes. This has also enabled women to accumulate retirement savings and entitlements of their own rather than relying solely on those of their spouse. Furthermore, there is a downward trend in the gender pay gap, providing women with more resources to put aside for their retirement. Women are also more often completing higher education, giving them access to higher paying jobs and leaving them in a better position to effectively plan and manage their financial future.

The ongoing COVID-19 crisis is threatening this progress, and is exacerbating many of the drivers of the gender pension gap. Women face an increased risk of unemployment, as they are more likely to work in sectors that have been heavily affected by restrictions, namely hospitality, tourism and retail sectors. Women also make up the majority of part-time workers, who are at higher risk of being laid off in times of crisis. When schools close, women are taking the burden of ensuring the education of their children. Physical distancing norms mean that women may also have to take over the caretaking responsibilities from grandparents who were contributing to caring for grandchildren in order to protect their elderly parents from the virus. These pressures may cause women to reduce their working hours to be able to fulfil the additional caretaking responsibilities, or prevent them from returning to the labour market. They may also lead women to reduce or suspend their contributions to their retirement savings plans.

Women are also less likely to be in a position to fully participate in the post-COVID recovery phase and to be able to overcome the detriment that the pandemic has had on their retirement preparedness. Their risk aversion in investment may make them less likely to invest in the stock market and benefit from the economic recovery. Women will also be more adversely affected by policies allowing the withdrawal of retirement savings during this period. Their lower balances on average mean that it will be even harder to make up these losses as they will miss out on the compound returns. The increase in divorce rates following lockdown measures will leave women in a more vulnerable position as they could lose the retirement benefits of their spouse.

In the wake of the pandemic, policy makers are therefore facing an even greater challenge to reduce the gender pension gap and improve gender equality. Fixing this problem will necessarily have to go beyond improving the design of retirement savings arrangements, and will need to target the major drivers behind the gender pension gap, particularly gaps in labour market participation and pay.

To target the drivers of the gender pension gap, pension funds themselves can make a difference with the way they choose to invest their members’ assets, provided their investment decisions remain in the best interest of members. They can invest in projects that promote social infrastructure that address the sources of the gender gap, such as childcare and financial education. They can also actively engage as shareholders to ensure that the companies they are investing in are also working towards gender equality in terms of pay and executive-level representation, and can sanction those who fall short with their voting decisions. Such investment behaviour will eventually have a real impact on how companies address gender gaps in their employment practices, and the benefits will translate into higher salaries and higher retirement savings for women.

The current situation calls for more drastic measures to reduce existing gaps linked to employment and to help women return to work in the recovery from this crisis so that they may continue to effectively plan for a secure financial future. The gender pay gap, disincentives linked to the cost of care, and low financial capabilities all need to be tackled – in addition to the design of retirement savings arrangements– if we are to continue to progress in closing the gender pension gap.


[17] Anderson, D. and J. Collins (2017), Can knowledge empower women to save more for retirement?, https://crr.bc.edu/wp-content/uploads/2017/09/wp_2017-12.pdf (accessed on 25 February 2019).

[19] Behave London (2019), Gender, Age & Pension Savings: New research for the pension industry into communication biases.

[7] Croson, R. and U. Gneezy (2009), “Gender differences in preferences”, Journal of Economic Literature, Vol. 47/2, pp. 448-474, https://doi.org/10.1257/jel.47.2.448.

[4] Dale, M. and S. St John (2020), Women and Retirement in a post COVID-19 world, https://cdn.auckland.ac.nz/assets/business/about/our-research/research-institutes-and-centres/RPRC/OtherPapers/Women%20in%20Super%20September%202020.pdf.

[26] Garnick, D. (2016), Income Insights: Gender Retirement Gap.

[8] Gneezy, U., K. Leonard and J. List (2008), Gender differences in competition: Evidence from a matrilineal and patriarchal society, NBER Working Paper Series, http://www.nber.org/papers/w13727.

[11] Government Equalities Office (2019), Gender Equality Monitor, HM Government.

[14] Hinz, R. et al. (eds.) (2013), Matching Contributions for Pensions: A review of international experience.

[13] Inland Revenue (2019), Annual Kiwisaver Statistics, https://www.ird.govt.nz/about-us/tax-statistics/kiwisaver/datasets.

[2] Jethwa, C. (2019), Understanding the Gender Pension Gap.

[31] Jusufovic, K. (2015), To Riester or not to Riester: A matter of pension regulations in Germany.

[21] Klammer, U. (2017), The family and gender dimension in German pension reforms.

[18] Leybold-Johnson, I. (2017), Pensions - what women need to know, https://www.swissinfo.ch/eng/gender-pension-gap_pensions-what-women-need-to-know/42970284.

[24] Nationwide Pension Fund (2020), Change in Circumstance, https://nationwidepensionfund.co.uk/active/benefits/change-in-circumstance (accessed on  September 2020).

[12] Now Pensions (2019), Facing an unequal future Closing the gender pensions gap, https://www.nowpensions.com/wp-content/uploads/2019/07/NP-gender-pensions-gap-report.pdf.

[3] OECD (2020), OECD Employment Outlook 2020: Worker Security and the COVID-19 Crisis, OECD Publishing, Paris, https://dx.doi.org/10.1787/1686c758-en.

[25] OECD (2020), OECD Pensions Outlook 2020, OECD Publishing, Paris, https://doi.org/10.1787/67ede41b-en.

[1] OECD (2019), “Are funded pensions well designed to adapt to non-standard forms of work?”, in OECD Pensions at a Glance 2019: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/1231a36d-en.

[10] OECD (2019), Pension Markets in Focus.

[22] OECD (2019), Smarter financial education: key lessons from behavioural insights for financial literacy initiatives.

[20] OECD (2018), Financial Incentives and Retirement Savings, OECD Publishing, Paris, https://doi.org/10.1787/9789264306929-en.

[6] OECD (2017), The Pursuit of Gender Equality: An Uphill Battle, OECD Publishing, Paris, https://doi.org/10.1787/9789264281318-en.

[30] OECD (2016), Life Annuity Products and their Guarantees, OECD Publishing, Paris, https://doi.org/10.1787/9789264265318-en.

[15] OECD (2013), Women and Financial Education: Evidence, Policy Responses and Guidance, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264202733-en.

[5] OECD/INFE (2016), International Survey of Adult Financial Literacy Competencies.

[16] OECD/INFE (2013), OECD/INFE Policy guidence on addressing women’s and girls’needs for financial awareness and education.

[27] Pensions Policy Institute (2019), “The gender pension gap: can it be closed?”, PPI Briefing Note 114, https://www.pensionspolicyinstitute.org.uk/media/3185/20190604-the-gender-pensions-gap-can-it-be-closedpub.pdf.

[23] Read, E. (2017), Pension gender gap, how can we respond?, https://www.slideshare.net/OECD-DAF/eve-read-pension-gender-gap-how-can-we-respond.

[9] Roszkowski, M. and J. Grable (2005), “Gender Stereotypes in Advisors’ Clinical Judgments of Financial Risk Tolerance: Objects in the Mirror Are Closer than They Appear”, Journal of Behavioral Finance, Vol. 6/4, pp. 181-191, https://doi.org/10.1207/s15427579jpfm0604_2.

[29] Serenelli, L. (2020), “German pension equalisation changes likely to hinder providers”, IPE, https://www.ipe.com/news/german-pension-equalisation-changes-likely-to-hinder-providers/10047767.article?utm_campaign=107549_11.9.20%20ipe%20daily%20news&utm_medium=email&utm_source=IPE&dm_i=5KVE,2AZH,C5K8L,8U0R,1.

[28] Woodward, H. and M. Sefton (2014), Pensions on divorce: an empirical study, Cardiff University.


← 1. However, this also includes women receiving the per-child subsidy, which go to the mother by default.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2021

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.