Recent History

In Part 28 of the Pension Series, I stated that age, tenure, and (in some cases) gender mattered more than any singular pension design element during a pensionable worker’s Golden Albatross pension decision. I based that statement on the evidence from my master’s thesis research. For my thesis, I ran a survey for pensioners and pensionable workers which I called the Golden Albatross survey. To complete the thesis I statistically analyzed the results of the survey, which I discussed in Pension Series Part 28. This post expands on the gender portion of those findings by focusing on the impact of defined benefit pensions for women in retirement.

Much like Pension Series Part 24 (Black Pensions), for this article I examine if and how the Golden Albatross decision-making framework should be modified for a specific sub-set of pensionable workers. In this case, that sub-set is women. In doing so, I specifically focus on the disparities in retirement savings between men and women and the impact of defined benefit (DB) pensions on women’s retirement outcomes. I conclude that the Golden Albatross calculation is different for female versus male pensionable workers.

A Friendly Reminder

As a reminder, the Golden Albatross (TGA) is a decision point many pensionable workers reach where they decide if staying for their pension is worth it. Pension design elements are the features or penalties built into defined-benefit pensions that entice or compel workers to stay for the pension (often called the generosity level). The features act as carrots, while the penalties serve as sticks.

Also, as a reminder, statistical analysis of my survey data showed that female participants seriously considered leaving their pensionable jobs significantly more than male participants. In a few cases, my analysis showed that female participants left, or remained undecided about leaving, at a substantially higher rate than their male counterparts, who elected to stay. However, while I opined in both my thesis and in Pension Series Part 28 about why that might be, I never discussed what it meant for women in retirement. Thus, this article plugs that gap in my thinking using an evidence-based approach.

The Disparity in Women’s Retirement Savings

In order to examine women’s pensions and their impact on retirement, it’s best to review the disparities between men’s and women’s retirement savings first. Fortunately, that’s not hard because there’s a lot of data on this issue, not only in the US but across the developed world. Unfortunately, that data shows distinct disparities between men’s and women’s retirement savings.

In 2017, the World Economic Forum noted that women’s global retirement savings balances were typically 30% to 40% smaller than men’s (WEF, 2017). Although a complex issue, the WEF labeled women’s “lower career salaries and longer periods out of the workforce” as causes for this disparity (WEF, 2017, p. 12). In other words, the gender pay gap, where women earn less than men for the same job, and disruptions in work history due to maternity and parenting absences, drive these differences.

The Organisation of Economic Cooperation and Development (OECD) listed additional causes of the male-female retirement savings disparity in 2018. Among the 34 OECD member-nations they studied, they noted lower work-based retirement plan coverage rates for women than men; lower participation rates among women in work-based retirement savings programs; poor design of work-based retirement savings plans which excluded many women; and a tendency for women to invest more conservatively than men also contribute to the problem (OECD, 2018). As the OECD explained, some of these reasons are structural, while others are cultural.

Gender Pay Gap Impact

What does that look like in everyday life? In the US, women earn approximately 81% of what men make, all other factors being equal (Bond et al., 2020; BLS, 2021). Thus, in 2020, US women earned a weekly median of $891 while men earned $1,082 (BLS, 2021). Furthermore, in the US, women’s working careers are shorter. US women work for 29 years on average, while men work 38 years (Garnick, 2016).

These disparities translates to fewer retirement savings for women versus men in the US. For instance, the 2021 T. Rowe Price Retirement Savings and Spending Study reported that women’s average 401k account balance was $50,700 versus $92,500 for men (Ward, 2022). Furthermore, in 2020, the Olson Group’s Benefits Team noted that (Scholl, 2020):

According to a 2019 report by the Employee Benefit Research Institute (EBRI), the average retirement savings shortfall for women is nearly twice that of men. Per the study, single women have an average retirement savings shortfall of $72,883 versus $37,690 for single men.

This means that both sexes save less than they should, but women are worse off of the two.

The Disparity in Women’s Retirement Income

Unsurprisingly, the women’s retirement savings gap translates into a retirement income gap, both worldwide and in the US. For example, the OECD found that (OECD, 2018):

On average … combining public and private sources, women aged 65 and older receive 26% less [retirement] income than men.

Of course, that’s an average of the 34 nations studied, meaning some countries fared better while others fared worse. Their data for the US showed an approximate 35% difference in retirement income between men and women (OECD, 2018).

More recently, the US’s National Institute on Retirement Security (NIRS), a think-tank, only reported a 17% median disparity in annual retirement income between male and female-led retirement age households (Bond et al., 2020). The difference in these rates may be due to different data sources or comparison methods. However, even though the NIRS report is half-as-bad as the OECD’s, the reported difference is still depressingly big. In dollar terms, it translated to an average of $57,144 for men and $47,244 for women (Bond et al., 2020).

In addition to all the above-cited reasons for the male-female income and retirement savings gap, the NIRS report added that in the US, late-career divorces, longer life-spans, and societal pressures to act as unpaid caregivers for family members also played significant roles in reducing women’s average retirement income (Bond et al., 2020). On the plus side, women who stayed married, divorced early, or were widowed, often had higher rates of retirement income at the household level than late-divorcees or separated women (Bond et al., 2020).

Caveats

If this all sounds rather complicated, it is. The NIRS report points out that the male-female retirement income gap isn’t even the most distinct disparity. Examining retirement income through the lens of race, age, or socio-economic status yields as great or greater divides than gender (Bond et al., 2020). So, I don’t want to oversell the point.

On the other hand, if you start combining factors like race and gender, you’ll see those disparities compound in many cases. For instance, retirement age black women-led households earned a median income of $37,548 in 2016 (Bond et al., 2020). That’s a 38% difference compared to the $60,540 for retirement age white male-led households (Bond et al., 2020). Those combinations don’t always yield worse results, though. Retirement age Asian women-led households earned $66,012 annually in 2016, more than any other combination of race and gender (Bond et al., 2020).

Women’s Retirement Outcomes

What does all this mean for women’s retirement outcomes? At the headline level, it leads to some rather grim statistics if you use a nation’s income poverty level for measurement. Admittedly, that’s a low bar, but it’s the same one most economists seem to use and the one I used for Pension Series Part 24 (Black Pensions). In any case, the US Department of Labor notes that in the US (Boiman & Khawar, 2021):

For all these reasons and more, women typically are retired for a longer period of time than men and with fewer assets. Unfortunately, women age 65 or older are about 43% more likely than men to live on an income below the poverty level. And about 65% of the elderly poor are women.

In the UK, where retiree poverty levels are on the rise overall, 2020 government figures show that 20% of women lived on or below the poverty line, while only 16% of men did (Independentage.org, 2021). As AgeUK.org noted, that marks a 6% rise in the women’s rate since 2012, which occurred (AgeUK.org, 2021):

Despite the rise in women’s State Pension age meaning the number of female pensioners in the UK has fallen by around 800,000 since 2012-13 (from 7.1m to 6.3m)[iii], the number living in poverty has actually increased by around 260,000 (from 990,000 to 1.25 million).

Rough Translation

As an aside, most nations refer to their retirees as “pensioners,” meaning they surpassed the official retirement age and receive income from government-sponsored retirement schemes. Unlike my use of the term throughout this website, “pensioner” is not a specific reference to a defined benefit (DB) pension, although many state-sponsored retirement schemes (like Social Security in the US) act like one. Thus, in New Zealand, where I currently live, 14% of female pensioners live at or below the poverty line. In comparison, only 6.6% of male pensioners do (Arnold, 2021).

More Caveats

Again, though, there’s some vital context to the women’s retirement poverty issue. As NIRS notes, in the US, from age 65 to 79, the male-female poverty rates in the US are relatively similar, with only a 0%-2% difference depending on the age group (Bond et al., 2020). It’s only from age 80 and up (when many more women become widowers) that poverty rates spike much higher (4%) for women over men (Bond et al., 2020).

Furthermore, other demographics like marital status and race impact male-female retirement outcomes differently. For instance, marriage mitigates the poverty differences completely. Only 3% of married men and women retirees in the US live below the poverty line (Bond et al., 2020). Others, like race and ethnicity, amplify the disparities. For example, female Latin (Latina) retirees are the worst off, with a 17% poverty level compared to 5% for white women (Bond et al., 2020). Flip to page 9 of the NIRS report (embedded below) to learn more about those nuances across age, race/ethnicity, and marital status.

Defined Benefit Pension Information

As one might expect from the above discussion about the effects of various demographics on retirement outcomes, comparing the poverty rates for women with and without access to defined benefit (DB) pensions makes for exciting reading! That assumes you can find reports that examine this particular issue. As with my Black Pensions article, I struggled to find academic or government-level research on women’s DB pensions.

The best document I found is a different NIRS report (from the one cited above) entitled The Pension Factor 2012 (TPF 2012), which I referenced heavily for the Black Pensions article. I do so again for the remainder of this article. It’s a US-centric document since NIRS is a US think tank, and I’ve not encountered an international equivalent.

In my opinion, if you’re a female and/or non-white pensionable worker, TPF 2012 is worth reading cover to cover. I linked to it above and embedded it below. If reading isn’t your thing, here’s a link to its YouTube video.

Unfortunately, the gender-separated categories for those with or without DB pension access include a spouse’s pension. Thus, it’s not a complete apples-to-apples comparison between women with or without a DB pension. However, it’s as close as I can find.

The Impact of Defined Benefit Pensions

In my opinion, TPF 2012 holds two headline comparisons applicable to the women’s pension issue. The first is the percentage of US women retirees without DB pension access who live in poverty vs. those with DB pension access. Whereas 18.4% of retirement-age women without DB pension income live in poverty, only 2% of those with DB pension income do (Porell & Oakley, 2012).

That’s a massive decrease! However, maybe even more important for the audience of this blog, pension income increases the percentage of retirement-age women with income at or above 200% of the poverty line from 46.7% to 78.3% (Porell & Oakley, 2012). In other words, DB pension income appears to move a large percentage of women from retirement poverty to retirement middle-class. This allows women the chance to enjoy retirement, not just survive it.

The second headline comparison in TPF 2012 catalogs how DB pension income significantly decreases food, shelter, and medical material hardships for retired-age women in the US. For example, the percentage of women experiencing at least one annual medical hardship in retirement (i.e., too poor to pay for it) shrinks from 10% to 4.6% for those with DB pension income (Porell & Oakley, 2012). Annual food and shelter hardships also drop precipitously for those with DB pension income. However, the medical issue is probably the most important for US women since their medical costs are 7% to 11% higher than men’s in retirement (Garnick, 2017).

Correlation vs. Causation

Far be it for me to rain on the women’s pension parade of good news, but while the above statistics imply causation, it could be a correlation. In other words, DB pension income may not be the driver for the difference. The category for women with access to pension income in TPF 2012 included a spouse’s pension. So, the steep drop in poverty percentages could be the marriage effect that I noted in the More Caveats section above.

In fact, elsewhere in TPF 2012, it’s stated that including spousal DB pension income effectively doubled DB pension coverage for retirement-age women from 20.6% to 41.8% (Porell & Oakley, 2012). So, it’d be hard to conclude that the positive effects of marriage weren’t playing a role. The question in my mind is, how significant of a role does marriage play compared to the positive impact of a DB pension’s steady monthly income?

Comparing poverty rates among a sub-group of women, like divorcees or widowers, both with and without DB pension income, could’ve provided better insight. However, the data that the authors of TPF 2012 analyzed, which came from the US Census Bureau’s Survey of Income and Program Participation (SIPP), may not have allowed it. That said, I don’t want to undersell the positive effects of DB pensions for women in retirement either. They definitely play a positive role in women’s retirement outcomes, regardless of the correlation or causation issue.

Advantages of DB Pension Income for Women

In light of the correlation vs. causation question for retirement outcomes, let’s discuss the advantages and disadvantages of DB pension income for women instead. We will do this while specifically keeping the root causes of the male-female retirement income disparity in mind.

The main advantage of DB pension income for women is that they cannot outlive it in retirement. This is true regardless of whether or not the retirement income stems from the woman’s pension or her spouse’s. That assumes survivorship was elected on the spouse’s pension. Considering that 2016 US Social Security data revealed that 2/3rds of married women who reach age 65 outlive their husbands by 11.5 years on average (Garnick, 2017), it’s no wonder why total survivorship is the default setting in the US for annuitized pensions (OECD, 2018).

If a DB pension has a cost of living adjustment (COLA), especially one indexed to inflation, then the pension is doubly advantageous for women. Not only will women avoid outliving their pension income, but they would maintain the same purchasing power throughout retirement — at least for that income stream. US women lucky enough to have access to a DB pension with an inflation indexed-COLA and Social Security (also indexed to inflation) stand to benefit even more. They can essentially write off more of their monthly retirement expenses against fixed income that maintains its purchasing power throughout retirement. With the US year-over-year core inflation rate at 40 year highs as I write this article (BLS, 2022), that too should not be undersold.

Battle of the Sexes, Pension Style

Furthermore, women benefit more than men from the efficiency of pension annuities over insurance annuities. As noted in Pension Series Part 25, US government studies show that annuities provided by DB pension plans are far more efficient than those offered by insurance companies. In other words, pension plans provide participants larger annuities than insurance companies can for equivalent invested amounts. Other studies have shown similar results (Rhee, 2014).

Part of that reason is that many pension funds have economies of scale that dwarf insurance markets (Rhee, 2014). However, this efficiency compounds in favor of US women in certain types of DB pension plans because the US Supreme Court ruled that unisex life expectancy tables must be used by DB pension plans for certain calculations (Garnick, 2017). Thus, this unisex requirement results in a complicated wealth re-distribution from men towards women in the same pension plan (Wilmore, 2001; Turner, 2010).

That’s great news for women! I say that because it is the exact opposite effect of what I discuss in my first pension lump-sum article, where women with their longer life spans end up subsidizing men when they opt for lump sums (Turner, 2010). Especially interesting, though, is that since women earn less than men for the same work (as discussed above), some economists believe that the wealth distribution from men to women for annuities within DB plans is fair (Wilmore, 2001). While I can’t speak to the fairness issue, what I can say is that under the right conditions, annuitizing a pension, as opposed to taking it as a lump sum, makes a lot more sense for women than men.

Disadvantages: The Pension Formula

The primary disadvantage of DB pensions for women is that they lock in women’s reduced earnings from their working years via mathematical formulas that use salary as an input. This results in smaller annuities than men in retirement. For example, in 2010, the average annual DB pension annuity value in the US was $22,238 for men but only $15,307 for women (Porell & Oakley, 2012).

Interestingly, I also noted this disadvantage in Pension Series Part 24 for black versus white pensionable workers and retirees. However, since most DB pension formulas also consider the number of years worked (tenure), women are doubly disadvantaged since they work for 9 fewer years than men on average (Garnick, 2017). The negative effect of these two disadvantages on women’s pension values is well-documented among economists and retirement scholars (Turner, 2010).

Disadvantages: Backloading

Pension formulas can also favor final salaries, such as an average of a worker’s last 3 years’ worth of salary. This creates a powerful incentive for pensionable workers to stay in their current job because their final working years are typically their highest-paid (Elliot & Moore, 2000). This is known as backloading, which I explained in Pension Series Part 25. However, I didn’t explain in Part 25 that backloading negatively impacts women more than men.

In truth, the requirement that pensionable workers stay at a single employer for decades to accumulate considerable pension benefits is ill-suited for many women due to their lack of “career orderliness” (Doerpinghaus & Feldman, 2001). Meaning, as previously discussed, women come and go from the labor force more than men. DB pension plans don’t accommodate these irregular employment patterns well. In fact, depending on a DB plan’s rules, backloading can impose severe negative consequences for leaving a pensionable job earlier than the normal retirement age (Taylor, 2000). The primary mechanism for these consequences is a frozen pension, which stops accumulating value once an employee leaves their current employer and cannot be transferred (Childs et al., 2002).

Thus, it’s not surprising that significantly more women than men considered backloading during their Golden Albatross decision, according to my survey. Given that women work shorter careers and leave the workforce more often than men, they run a greater risk of a frozen pension. Thus, it’s only natural that they’d consider this negative consequence more than male participants in the survey.

Advantages and Disadvantages: Public Jobs

In the US, many women and non-white workers combat wage disparity and discrimination by seeking out government (public) jobs at the federal, state, or local level. Government jobs attract non-white and women workers because they feature public pay scales that apply equally to all genders, races, and ethnicities. They are also governed by federal and state anti-discrimination and affirmative action measures (Cooper & Wolfe, 2020). As a result, in 2018, women made up 60% and Blacks made up 13.5% of all public workers at the state and local levels in the US (Cooper & Wolfe, 2020). Both figures were proportionally larger than the overall percentage of women and Blacks in the workforce (Cooper & Wolfe, 2020).

However, this over-representation by women and non-white workers in state and local government jobs leaves them disproportionally exposed to state and local budget constraints. As the Economic Policy Institute noted (Cooper et al., 2012):

Between 2007 (before the recession) and 2011, state and local governments shed about 765,000 jobs. Women and African Americans comprised about 70 percent and 20 percent, respectively, of those losses.

Furthermore, the US’s state and local government employment rates lagged behind the private sector’s recovery after the Great Recession. For example (Cooper et al., 2012):

From February 2010 (the month the labor market “bottomed out”) to January 2012, the United States experienced a net increase in total nonfarm employment of more than 3.2 million jobs, while state and local government employment fell by 438,000.

Disadvantage: Pension Safety

What does any of this have to do with women’s pensions? Well, state and local public jobs are overwhelmingly pensionable jobs. In March 2021, 86% of all state and local government workers in the US were covered by a DB retirement plan, of which 75% of eligible workers participated (Topoleski & Meyers, 2021). Unfortunately, though, the average funding shortfall for US local and state-run pensions systems was 29% in 2020 (Aubrey et al., 2020).

The budgetary constraints that cause layoffs at state and local levels are the same ones that cause state and local governments to underfund their DB pension liabilities (Cooper & Wolfe, 2020; Hiltonsmith, 2016). Thus, women in state and local government jobs in the US are overexposed to layoffs and the underfunding of their DB pension plans at the same time during economic downturns. Needless to say, if a DB pension can’t be relied on as promised in retirement, it won’t help deliver the positive retirement outcomes mentioned in previous sections.

Disadvantages: No Social Security

Approximately 25% of US state and local government employees do not participate in Social Security (Quinby et al., 2020). Instead, their erstwhile Social Security (SS) contributions are redirected into their pension plans along with their regular contributions. As a result, those workers won’t be eligible for SS payments in retirement without some other SS-eligible job history. According to federal regulations, however, these workers’ pension plans must provide equal or better compensation for that lack of SS, in addition to their regular pension payments. A 2020 Center for Retirement Research (CRR) study found that, on the whole, these state and local pension funds had lived up to that obligation … so far (Quinby et al., 2020).

Regardless, this redirecting of SS payments into pension plans represents a potential concentration of risk for those pensionable workers, especially if their pension plans are heavily underfunded. Not only that, but for women, who rely more heavily on SS than men in retirement (Bond et al., 2020; Hiltonsmith, 2016), it removes a valuable, inflation-protected, and diversified retirement income stream. Interestingly, that reliance by women on SS is greatest for black women-led households, where it makes up 57% of all retirement income (Bond et al., 2020). It’s the least great for Asian women-led families, where it only makes up 34% of all retirement income (Bond et al., 2020). White women (53%) and Latinas (46%) fall between. Notably, all women’s ethnic and race categories relied on SS more than their male counterparts (Bond et al., 2020).

Advantages: Black Women in Public Jobs

I found a few DB pension-related statistics that apply only to black women in pensionable public jobs in the US. They definitely fall on the advantages side of the argument. The first is that (Hiltonsmith, 2016):

…nearly a quarter (24 percent) of retired black women without public pensions are in poverty, while just 3 percent of black women with public pensions are. 

That seems significant, but I cannot verify if the author controlled for other potential causal factors, like marriage. That said, I also found this interesting statistic in the same document (Hiltonsmith, 2016):

Social Security and public pension income is particularly important for black women: together, the two income sources accounted for 63 percent of the total retirement income of black women, the highest share of any race and gender; black men, in comparison, received 51 percent of their income from the two sources combined.

What All of This Means for TGA

Compared to men, women have a dizzying number of additional considerations to include in their Golden Albatross (stay-or-go) pension decision-making process. While some of those considerations, like lesser pay over a career, are similar to those for black workers, others, like living longer in retirement, are not. Thus, unlike my blanket advice to black workers that they should exercise more caution than white workers when deciding to leave a pensionable job (due to the overwhelmingly positive impacts of pensions on their retirements), I have no such guidance for women.

However, women should modify the Golden Albatross decision-making framework to accommodate more factors than those discussed in my book … and elsewhere on this blog. Women shouldn’t simply line up the number of valuable pension design elements on one side; their personal needs and desires on the other; and then make a stay-or-go decision. I don’t believe that would lead to the most well-informed decision a woman could make, which is the whole point of the Golden Albatross framework.

Working at Cross-Purposes

Certainly, starting with one of the methods in my book won’t hurt because they provide a basis for understanding the intrinsic and objective worth (or lack thereof) of a DB pension. However, I realize that the Golden Albatross situation for pensionable women calls for more nuance. I say that because many of the additional factors women need to consider work against each other.

For instance, an annuitized pension lessens the likelihood of women outliving their retirement savings. On the other hand, since the average woman’s career is shorter than the average man’s and tends to feature more breaks in service, the backloading effect built into many pension formulas lessens an annuity’s value for the woman. Furthermore, if one or more of the traditionally valuable pension design elements that I examine in my book, like a COLA, is present, value calculations get complicated quickly.

The Bottom Line

Therefore, women must determine whether or not their DB pensions can deliver some of the positive retirement effects I discussed throughout this article on a case-by-case basis. Once done, they can  then determine if their pension is worth it. Naturally, determining whether or not staying for those outcomes is worth it will vary depending on the woman’s situation. A single, black, head of household, and pensionable woman is in a completely different set of circumstances than a married Asian woman with access to her husband’s pension. Thus, the relative impact of a pension on the retirement outcomes for those two different women will also vary. To expect anything else would just be unrealistic.