Caregiver Credits in France, Germany, and Sweden: Lessons for the United States

Introduction

In recent years, analysts and policymakers have directed increasing attention to securing the future of Social Security's old-age social insurance program. To date, the majority of proposed reforms have focused almost exclusively on the long-term fiscal sustainability of the program, with measures that would reduce benefits (for example, through an increase in the full retirement age (FRA)), increase revenues (for example, through an increase in the payroll-tax ceiling), or a combination of both. At the same time, many scholars have argued for the need to complement any such reforms with measures that would improve the adequacy of benefits for certain groups, especially those that would be particularly vulnerable to benefit cuts (Reno and Lavery 2009; US Senate 2010). Among the most vulnerable of those groups is that of women aged 65 or older, a group with poverty rates almost twice that of men in the same age group. According to Social Security Administration (SSA) figures, 11.9 percent of women aged 65 or older fell below the poverty line in 2008 compared with 6.7 percent of men (SSA 2008). Poverty rates were even higher when looking at certain subgroups of women, especially among the nonmarried (16.9 percent), widowed (15.4 percent), and divorced (19.5 percent) categories. Although many factors have led to those high poverty rates, one significant factor is the substantial gap in the labor force participation of many women because of providing unpaid care to children and sick or elderly relatives. Those gaps often result in shorter work histories, lower average lifetime earnings, and consequently lower benefits at retirement than men. This is especially a concern for the significant number of women who are not eligible for current-law spouse or survivor benefits, often called auxiliary benefits, based on the contribution record of a spouse.

This article focuses on an often-considered measure to improve the adequacy of Social Security's retirement benefits for women: the introduction of caregiver credits provided for time spent out of the workforce while caring for dependent children and sick or elderly relatives. Throughout the majority of European Union (EU) member states, caregiver credits have become an established component of public pension programs used by countries to pursue a number of objectives, including improving benefit adequacy for caregivers (primarily women but also men), promoting higher fertility rates, facilitating the return to the labor force following childbirth, and simply rewarding the act of providing unpaid care. Through an analysis of three of the longest running caregiver credit programs—in France, Germany, and Sweden—this article shows how these programs have been developed abroad. The article especially focuses on how these countries have addressed three key challenges: (1) the policy challenge of designing a program that targets a specific segment of the population, (2) the administrative challenge of determining an individual's eligibility for the credits, and (3) the financial challenge of funding the additional benefit.¹

The article first provides background on auxiliary benefits under the current-law Social Security program and examines the adequacy of those benefits, given recent changes in marital patterns and family structures in the United States (US). It then profiles caregiver credit programs in France, Germany, and Sweden. Finally, it concludes by discussing whether the experience of these countries provides any insight into the potential adoption of caregiver credits into the US Social Security program.

Protection for Caregivers under the Current US Social Security System

Under current program rules, caregivers are covered only indirectly by Social Security through spouse and survivor benefits. This section describes the current program and analyzes its effectiveness in reducing the risk of poverty among older women.

Program Rules

Social Security provides auxiliary benefits to the spouses and survivors of retired, disabled, and deceased workers. Table 1 summarizes the eligibility requirements and amounts of those benefits. Under current law, the spouse of a retired worker is eligible to receive 50 percent of the retired worker's primary insurance amount (PIA) if claimed at the FRA. Individuals who are eligible for benefits based on their own work history and their spouse's work history—referred to as dually entitled individuals—receive the worker's benefit plus the difference between that amount and the benefit they would receive as a spouse. For example, if a woman was dually entitled, she would receive the full benefit based on her own earnings record plus the difference between that amount and the benefit she would receive as a spouse (that is, 50 percent of her husband's benefit). Spouse benefits are also paid to individuals who are divorced, provided they were married to an insured worker for at least 10 years and are currently unmarried.

Social Security also provides benefits to surviving spouses of deceased workers. A widow(er) retiring at the FRA is eligible to receive 100 percent of the deceased worker's PIA plus any additional amount the deceased worker was entitled to receive based on delayed retirement credits. A reduced benefit, ranging from 71.5 percent to 99.9 percent of the deceased worker's PIA, is paid as early as age 60 (age 50 if disabled). As with the spouse's benefit, a divorced widow(er) is eligible for this benefit if he or she was married to the deceased worker for at least 10 years. With the exception of a few specific circumstances, a widow(er) must have been married to the deceased worker for at least the 9 months immediately prior to the worker's death.

Social Security's auxiliary benefits have become more generous since first introduced in 1939. Under the 1939 Amendments to the Social Security Act, the wife of a retired worker was eligible to receive a benefit equal to 50 percent of her husband's PIA provided she was aged 65 or older, and a widow in the same age group was eligible to receive 75 percent of the deceased husband's PIA if benefits were claimed at the FRA by the widow or the worker. Eligibility for spouse and survivor benefits was subsequently extended to men in 1950; in 1965, eligibility was also extended to divorced spouses and divorced survivors with a 20-year marriage history. (The duration-of-marriage requirement was reduced to 10 years in 1977.) In addition, benefit levels for survivor benefits have increased significantly over time, from 75 percent to 82.5 percent of the deceased spouse's PIA in 1961, and to 100 percent in 1972, where it remains today. (By comparison, within the Organisation for Economic Co-operation and Development (OECD), the United States is the only country with survivor benefits equal to 100 percent of the deceased spouse's PIA; most countries offer benefits ranging from 50 percent to 80 percent of the spouse's benefit).

In short, under current program rules, auxiliary benefits are based entirely on marriage, with no direct compensation for unpaid caregiving. Unpaid caregivers are only compensated based on their marital status, which can lead to some caregivers (for example, those in traditional breadwinner families with a single high-earning spouse) receiving generous Social Security benefits and others (for example, those who have never been married or who divorced their spouse prior to meeting the 10-year length-of-marriage requirement) receiving no compensation for their caregiving activities.

The Adequacy of Current Law Auxiliary Benefits

Taking the population as a whole, Social Security has been extremely successful in reducing poverty rates in the United States. As Table 2 shows, the percentage of persons aged 65 or older falling below the poverty line decreased from 28.5 percent in 1966 to 9.7 percent in 2008.² However, those broad trends obscure the significant variation that is evident when looking at the population by sex and marital status, as displayed in Table 3. In 2008, 11.9 percent of women aged 65 or older fell below the poverty line compared with 6.7 percent of men. In addition, there were significant differences across marital categories, with relatively low rates for all married persons aged 65 or older (4.9 percent) compared with the nonmarried (15.5 percent), widowed (14.4 percent), divorced (16.4 percent), and never-married (17.6 percent) categories. Furthermore, a higher percentage of women fell below the poverty line than men across all marital categories, especially in the nonmarried (16.9 percent of women compared with 11.6 percent of men), widowed (15.4 percent of women compared with 10.2 percent of men), and divorced (19.5 percent of women compared with 11 percent of men) categories.

This discrepancy in poverty rates between men and women, and the fact that the gap is wider for certain marital categories, has led many analysts to question the adequacy of auxiliary benefits in the current-law Social Security program. As those analysts note, marital patterns and the structure of the modern family are far different today than they were in the 1930s when Social Security was first enacted. For example, the traditional breadwinner model—consisting of a working husband and a stay-at-home wife—is no longer the norm in American society. This is manifest in the US divorce rate, which has risen dramatically since the 1970s; approximately a third of all marriages currently end within 10 years of marriage, and thus before the 10-year length-of-marriage requirement for auxiliary benefits has been met (Bramlett and Mosher 2001; Tamborini and Whitman 2007). The change in marital patterns is also observed in the percentage of individuals aged 65 or older who have never married (and are therefore by definition ineligible for spouse or survivor benefits), which is projected to increase from 2 percent (men) and 4 percent (women) in 1992 to around 6 percent (men and women) in 2040 (Favreault and Smith 2004).³ Finally, the structure of the family has also changed in recent decades, with dramatic increases in childbearing among unmarried women—from 18 percent in 1980 to 40 percent in 2007—and in the number of single-parent households—from 5 percent in 1970 to 9 percent in 2006 (Ventura 2009; Census Bureau 2007).

While these changes have undoubtedly had the effect of reducing the number of women eligible for auxiliary benefits, they are offset by developments in the labor force that have led to more women being eligible for benefits based on their own work history. According to data from the Current Population Survey, the labor force participation rate of women has increased dramatically in recent decades, from approximately 34 percent in 1950 to 60 percent in 2008. During this same period, women's share of the labor force grew from only 30 percent to 46.5 percent. In addition, women's earnings have increased relative to those of men, from around 62 percent in 1979 (the first year for which comparable earnings data are available) to around 80 percent in 2009 (BLS 2010).

That said, women continue to be at a disadvantage in many respects despite these recent gains. For example, they continue to hold more part-time jobs than men, with approximately 25 percent of women usually working part time in 2008 compared with only 11 percent of men. (This figure has remained relatively constant since 1984, when 27 percent of women worked part time.) Furthermore, the enduring gap in women's earnings relative to those of men remains a concern, especially given that women's earnings have seemingly peaked at around 80 percent of men's earnings, with little movement in the past decade (BLS 2010). Because of these trends, women will continue to earn less throughout their lifetimes than men, leading to lower Social Security benefits (and a higher risk of poverty) at retirement.

Because of the higher risk of poverty among women (the data show that this risk is unlikely to diminish in the foreseeable future), significant attention has been directed by analysts and advocates at measures that would improve the adequacy of benefits for women. The rest of this article analyzes one such measure—the possible introduction of caregiver credits to compensate for unpaid caregiving to dependent children and sick or elderly relatives. Caregiver credits have long been an important part of European pension systems, representing one of a number of strategies used to combat poverty among certain vulnerable populations.⁴

Improving Benefit Adequacy through Caregiver Credits

In recent decades, caregiver credits have become a near-universal component of public pension systems in higher-income OECD countries.⁵ As mentioned earlier, the primary objective of those systems is to improve benefit adequacy for women, whose separations from the labor force to provide care for dependent children and sick or elderly relatives often lead to lower average earnings and lower benefits at retirement. At the same time, caregiver credits are used for a number of secondary objectives, including promoting higher fertility rates, creating an incentive to return to work following childbirth, and simply rewarding the act of providing unpaid care. This mix of objectives has led to significant variation in the design of caregiver credit programs across Europe, as discussed later.

Although caregiver credit programs have been around as early as 1945, with the adoption of a childcare pension bonus in France, widespread adoption only started to gain momentum around the 1980s. During that time, caregiver credit programs went into effect in Sweden (1970s), Germany (1986), Norway (1992), and Switzerland (1998), among others. A second wave of pension reforms in the early 2000s brought the introduction of caregiver credits in Luxembourg (2002), Austria (2005), and Finland (2005). Subsequent reforms in many of those countries have had the result of expanding and strengthening these benefits, for example by extending eligibility to men and by expanding the programs to include not only periods caring for dependent children, but also periods caring for sick or elderly relatives.

This widespread adoption of caregiver credits in almost all member states of the EU is the result of a number of factors, with a few of the most important being (1) the gradual reduction in recent decades in the generosity of survivor benefits and the resulting need for new ways to protect women; (2) the reduction in the generosity of public defined benefit (DB) programs and the adoption in some countries of mandatory defined contribution programs, which more closely link old-age benefits to a worker's lifetime earnings; (3) the increased emphasis in many countries on family friendly policies that aim at reconciling childcare with employment; and (4) the increased dialogue (and ultimately cross-national policy coordination) that takes place as a result of being a member of the EU.

First, largely as a cost-cutting measure, countries have gradually reduced the generosity of survivor benefits in recent years, opting in many cases to replace those benefits with other measures, such as caregiver credits and earnings sharing.⁶ In Germany, for example, the benefit rate for widows was decreased in 2002 from 60 percent to 55 percent of the spouse's pension, and the indexation of survivor benefits was adjusted downward. In Sweden, the length of time survivors below the FRA could receive benefits was reduced. Reforms in other countries have replaced the payment method from a pension to a less generous lump-sum payment (for example, Denmark and the United Kingdom); made the benefit means tested (for example, Greece); or phased out the benefit for widows altogether (for example, Australia and Latvia). (For further discussion, see James (2009) and Monticone, Ruzik, and Skiba (2008).) In an attempt to lessen the impact of those reductions, countries have increasingly been turning to measures such as caregiver credits.

Second, to lessen the burden placed on public pension systems by demographic aging, many countries have implemented reforms to their pension systems in recent years, including reductions in the generosity of public DB pension benefits and the introduction of defined contribution (DC) pension programs to supplement, and in some cases replace, existing public DB programs.⁷ France, for example, increased its legal retirement age from 60 to 62 in 2010 and is steadily increasing the number of contribution years required for a full pension from 40 years in 2003 to 41.5 years by 2020. To mitigate the potential impact those changes could have on women, the French government has implemented (or further expanded) a number of policies to increase women's eventual pension benefits.⁸ In Sweden, a 1999 pension reform law replaced the former pay-as-you-go DB system with a pay-as-you-go notional defined contribution (NDC) system,⁹ supplemented by privately managed individual accounts. Under that new system, caregiver credits—along with other measures, including unisex lifetables and a minimum pension guarantee—are utilized to ensure that women are not disproportionately hurt by the system's closer linkage of benefits to contributions.

Third, countries across the EU have been actively pursuing a number of family friendly policies that improve flexibility for working parents, remove barriers to employment (particularly for women), and reverse the decline in fertility and birth rates (currently below replacement levels in every EU member state except Ireland¹⁰). That process has led to the adoption of a number of programs in recent decades, including not only caregiver credits, but also parental leave, family allowances, and childcare services. For example, in 1992, an EU directive (92/85/EEC) was passed that obligates states to provide a minimum of 14 weeks of paid maternity leave for childbirth.¹¹ In practice, almost all EU member states have passed more generous national leave policies, including Italy (20 weeks of maternity leave), Ireland (26 weeks), and France (16 weeks for the first two children and 26 weeks for the third child and subsequent children); see SSA (2010b). In addition, countries have further attempted to remove disincentives to female labor force participation by improving childcare services. In 2002, the European Commission set the target of providing childcare to at least 90 percent of children from age 3 to the mandatory school age and to at least 33 percent of children younger than age 3, by 2010 (EC 2002). Similarly, the EU has also directly promoted the adoption of caregiver credits, with recent studies emphasizing the importance of those credits in improving benefit adequacy for unpaid caregivers (EC 2006a and 2006b). Viewed from this larger context of EU family policy, pension credits are one of a number of tools used to reward the social contribution of women and to address the potential negative effects that childbirth can have on female labor force participation.

Finally, the EU has also arguably facilitated the expansion of caregiver credits simply by providing a forum for dialogue among the member states, many of which have long advocated the value of those programs. In the area of social protection, the formal manifestation of this repeated cross-national dialogue is the development of a common agenda—seen most recently in the so-called "Social Policy Agenda 2006–2010"—urging national governments to improve labor market conditions for women. At a more general level, the EU encourages an ongoing dialogue among policymakers in its member states through the "open method of coordination" (OMC), a framework for political cooperation that involves agreement on common objectives and common indicators to track member states' progress. The most recent objectives adopted in March 2006 include combating poverty and social exclusion, especially among the most vulnerable groups (including women). By creating a set of common goals and subsequently evaluating progress toward achieving them, the OMC allows EU member states to learn from each other, which can lead in many areas to a degree of policy convergence, with caregiver credits being just one example.

While caregiver credits have a long history in the EU, proposals to introduce credits in the United States have not met with success. To a large extent, this lack of US enthusiasm stems from three major concerns: (1) how to design a program that targets the credits to the correct population, (2) how to address the administrative challenges of such a program, and (3) how to pay for this new benefit. The rest of this section profiles France, Germany, and Sweden to examine how these countries have confronted those challenges.

Designing Caregiver Credit Programs

One of the main concerns voiced by many analysts looking at the possible introduction of caregiver credits in the United States is how to design those credits so that they reach the targeted population. Under certain designs, there is a danger that the credits would go to higher-income parents—those who can afford to take significant periods of time off from work—rather than to lower-income parents—who must combine caregiving with at least part-time work. That would be the case for one of the narrowest forms of caregiver credits, whereby credits are awarded only to parents who leave the workforce entirely. As this section shows, however, the three countries examined have all more precisely targeted benefits to certain types of caregivers.

When designing a program of caregiver credits, a number of issues must be addressed, including the following: (1) the number of years an individual will be eligible to receive the credits; (2) how the credits will be calculated; (3) who is eligible to receive the credits (that is, a mother, a father, or both); and (4) whether an individual has to leave the labor force completely to receive the credit. Table 4 provides a brief description of caregiver credit programs throughout Western Europe, showing a wide variation in the ways in which countries therein have confronted those issues. This variation is largely the result of the diverse goals these programs were designed to achieve, as is clear when looking more closely at some of the specific programs.

In France's general pension scheme—le régime general d'assurance vieillesse des travailleurs salaries¹²—there are three different forms of caregiver credits, each meant to achieve various objectives.¹³ The first form of credit (l'assurance vieillesse des parents au foyer, or AVPF) was introduced in 1972 largely to improve benefit adequacy among lower-income parents who exit the labor force or significantly reduce working hours in order to provide childcare. Under this program, parents who receive certain family benefits (such as the family income supplement, young child allowance, or parental leave allowance) and whose individual earnings are below certain thresholds (17,600 euros or US$23,028 for the first child and 30 percent more for subsequent children) are credited as if they were earning the minimum wage for up to the first 3 years of childcare. The credits are subsequently used upon retirement when calculating the insured person's old-age benefit. Each parent receives the benefit, provided they both meet the eligibility conditions; in practice, however, the benefit is paid predominantly to women, with approximately 33 percent of all women who retired in 2004 receiving the credit compared with only 1.5 percent of men. (In 2004, nearly 2 million people received AVPF, of which 92 percent were women.) On average, women born after 1945 receive 3 years of these credits (Bonnet, Buffeteau, and Godefroy 2006).

The second form of credit (les majorations de durée d'assurance, or MDA) targets women who do not benefit from the means-tested AVPF, especially those who continue to work after childbirth. Under the MDA, which was also introduced in 1972, a parent caring for a child under age 16 for at least 9 years receives up to 8 quarters (2 years) of coverage per child.¹⁴ (Under the French system, 40.5 years of coverage—increasing to 41 years by 2012 and to 41.5 years by 2020—are necessary to receive a full pension.) In other words, unlike the AVPF, the MDA does not credit a caregiver with a particular earnings level; rather, it simply reduces the number of total quarters MDA beneficiaries (including those with higher incomes) must work to receive a full benefit, allowing them to retire earlier than women without children. All quarters of coverage are generally granted to the mother; however, as of April 1, 2010, the second year of coverage can also be granted to the father or split between both parents upon request.¹⁵ The credit is granted irrespective of the caregiver's work activity, which can lead to the caregiver receiving coverage through MDA and through their own work activity simultaneously.¹⁶ According to the Pensions Advisory Council (Conseil d'Orientation des Retraites), more than 70 percent of women in the general scheme—and approximately 90 percent across all schemes—receive pension credits from the MDA upon retirement.

Finally, the third form of caregiver credits in France (majoration de pension pour trios enfants et plus) has the primary aim of increasing fertility rates by providing a pension bonus for parents with multiple children.¹⁷ To be eligible, parents must have raised three or more children for at least 9 years prior to the children's attainment of age 16. The bonus, which has been in effect since 1945, is equal to 10 percent of an individual's old-age pension and is awarded to both parents. According to the Pensions Advisory Council, approximately 5.4 million people (or 42 percent of all pensioners) received this bonus in 2004, with the average bonus amounting to 89 euros (US$117) a month, or 1,068 euros (US$1,405) a year.¹⁸

In Germany and Sweden, caregiver credit programs have also been designed to pursue a number of objectives, including improving the adequacy of benefits for caregivers and increasing the incentives for them to continue working (or to return to work after a brief absence) following childbirth. Under Germany's caregiver credit program (Kindererziehungszeiten), the parent who is "mostly" responsible for childcare receives the equivalent of 1 pension point per year for the first 3 years of his or her child's life.¹⁹ The credit is provided for children born after December 31, 1991, and is typically awarded to the mother; however, a father may also receive the credit upon written request to the German statutory pension insurance scheme (Deutsche Rentenversicherung Bund, or DRV-Bund). (Parents of children born from 1986—when the caregiver credit program was created—to December 31, 1991, receive credits for only the first year of their child's life.)

In addition, a 2001 pension reform law established additional credits—equal to one-half of the pension points a parent receives as a result of his or her work contributions, up to a maximum (own pension points plus bonus credit) of 1 pension point per year—for parents who continue to work while raising a child aged 3–10. In such a way, the credits provide an incentive for parents to return to work while providing childcare. (The law also provides parents who leave the labor force entirely to provide care for two or more children—one of which is younger than age 10—with an additional one-third of a pension point.)

In Sweden, caregivers are credited in the public pension system for any period of care for children aged 4 or younger. (Adoptive parents receive credits for the first 4 years the child is under their care, until the child reaches age 10.) The parent (mother or father) with the lowest earnings in the year prior to childbirth automatically receives the credit (called Barnårsrätt), unless they notify the Swedish Social Insurance Agency that the credits should be granted to the higher-earning spouse. There is no limit to the total number of years in which a parent can receive the credits throughout their lifetime; however, at retirement, benefits cannot be based only on those credits. (A parent must have at least 5 years of covered employment with an average income of at least two income base amounts²⁰ (102,200 kronor, or US$15,750 in 2010).) Additional requirements stipulate that the child must have been in the parent's custody and physically living with the parent for at least half the year, and both the child and parent must have been living in Sweden the entire year.

Swedish caregiver credits are calculated in three different ways, with each method targeted to specific types of caregivers. (The calculation method used is the one that is most beneficial to each individual caregiver.) Under the first calculation, the credit is equal to the caregiver's earnings in the year before the birth or adoption of the child. This calculation is beneficial for workers who had relatively high earnings in the year prior to childbirth or adoption and who significantly reduce work hours afterwards. Under the second calculation, the credit is equal to 75 percent of the average earnings in Sweden the year before childbirth or adoption. This calculation benefits workers who had relatively low earnings prior to childbirth or adoption and who reduce work hours significantly. Finally, under the third calculation, the credit is equal to one income base amount (52,100 kronor, or US$8,028 in 2011). This calculation provides a bonus to workers who continue to work approximately the same number of hours as before childbirth or adoption. In addition, it ensures that caregivers who return to work relatively soon after childbirth or adoption are not penalized by that decision.

While caregiver credit programs for childcare have become almost universal across the EU, programs that award credits for time spent providing care to elderly or sick relatives are much less common. Of the three countries analyzed here, only Germany has a program of credits for this type of care.²¹ In 1995, Germany introduced a new long-term care insurance program that specifically addressed the burdens placed on caregivers (particularly women) who often are forced to reduce the number of hours they work in order to provide unpaid home nursing care (Federal Ministry of Labour and Foreign Affairs 2009). Under that program, German caregivers receive pension credits for time spent providing unpaid care of at least 14 hours a week. To be eligible, a caregiver must work less than 30 hours a week and the person under care must receive benefits through the long-term care insurance program. The calculation of the credit is a factor of both the number of hours per week the caregiver spends providing unpaid care and the level of nursing care dependency.²² The pension credits are paid by long-term care insurance and range from about 0.25 pension points to 0.8 pension points per year, with no lifetime limit.

In sum, countries in the EU have taken a variety of approaches in designing their caregiver credit programs. The different approaches have allowed these countries to target specific populations, depending on the objectives that the credits are meant to achieve. In the case of France, the multiple forms of caregiver credits for childcare are meant to achieve a number of objectives, including improving benefit adequacy among lower-income parents, allowing caregivers (including those with high incomes) to retire earlier, and increasing fertility rates. Similarly, that mix of objectives is also apparent in Germany and Sweden, where caregiver credit programs attempt to improve benefit adequacy while at the same time providing a bonus to those persons who combine work with childcare responsibilities. Through these more complex designs, all three countries aim at ensuring that caregiver credits target the intended populations.

Administering Caregiver Credits

In addition to the challenge of designing a caregiver credit program that successfully reaches its targeted population, a second challenge is the administration of the program. Determining an individual's eligibility can be extremely complex to administer, especially in the case of caregiver credits for care provided to sick or elderly relatives (where determining eligibility requires verification that care has been provided). As this section shows, however, the three countries surveyed have all developed the infrastructure necessary to make the administration of their caregiver credit programs virtually automatic.

In France, caregiver credits under the general scheme are administered by the National Old-Age Pension Insurance Fund (la Caisse Nationale d'Assurance Vieillesse, or CNAV). To determine an individual's eligibility, CNAV simply requires that parents produce the birth certificate or, in the case of adoption, the adoption certificate. Caregiver credits are automatically awarded to the mother, but a father may also receive the credits by notifying CNAV in writing as to how the credits will be split (up to 4 quarters, or 1 year, of credits can be given to the father). If there is disagreement as to who should receive the credits, CNAV makes a decision based on information provided by the parents. For example, in the case of divorce, CNAV uses the divorce judgment to determine which parent has guardianship of the child(ren) and bases their decision on how to divide the credits on this information.

Under the German system, caregiver credits are administered by the DRV-Bund. When a child is born in Germany, he or she is registered with the registry office, which immediately passes the information on to the DRV-Bund. The credits are then entered into the mother's social insurance record, unless there is a written request by the parents to give the credits to the father. Parents who share childcare responsibilities can also state in advance which months should be credited to each parent, as long as the maximum of 3 pension points (1 pension point per year), shared between both parents, is not exceeded. The process of administering caregiver credits that are provided to persons caring for sick or elderly relatives is equally straightforward. Under the long-term care insurance program,²³ a medical determination is made, on both the level of care dependency and the number of hours of care an individual requires, by the Medical Review Board of the Statutory Health Insurance Funds (Medizinische Dienste der Krankenkassen, or MDK), under the supervision of the Ministry of Social Affairs in each province. Pension contributions are then transferred from long-term care insurance on behalf of the caregiver to the DRV-Bund. As such, the DRV-Bund has very little administrative role in the provision of those caregiver credits, and instead relies on the determination made under the long-term care insurance program by the MDK.

In Sweden, the administration of caregiver credits is also essentially automatic. All of the information needed to process the credits, including information on births, parental relations, custody, and place of residence, is contained in a civil registry maintained by the Swedish Tax Agency (Skatteverket). That data is transferred to the Swedish Pensions Agency (Pensionsmyndigheten) on a daily basis, along with earnings records, to determine which parent should receive the credits (absent a written request from the parents). As a result of this data exchange, the Swedish Pensions Agency has all the information needed to automatically award the credits, with very little additional effort needed by agency staff. The only administrative difficulties faced were at the program's initial implementation, when much of the information needed to award the credits was not yet computerized.²⁴

Funding Caregiver Credits

The final challenge involved in the introduction of a caregiver credit program is how to fund the program. Unlike the other two challenges, which have been addressed in a wide variety of ways, almost all countries are similar in their response to the funding issue, choosing to pay for caregiver credits out of general revenues or other taxes. In France, for example, both the MDA and the pension bonus are paid for by a public fund financed through a variety of earmarked taxes, including taxes on alcohol (D'Addio and Whitehouse 2009). Contributions for the AVPF are paid by the Family Allowance Agency (Caisse d'allocations familiales), which transfers the funds directly to CNAV.

Similarly, in both Germany and Sweden, the federal government finances caregiver credits for unpaid childcare through transfers from the state budget to the social security system.²⁵ In Germany, however, the government has gradually moved from transferring the entire value of the credits to transferring a fixed amount annually, an amount that falls well below the actual cost of the credits (Fultz 2011). Finally, and like the AVPF in France, credits for unpaid care of sick or elderly relatives are paid for by Germany's long-term care insurance program, which transfers the full value of the credits directly to the DRV-Bund.

Naturally, the cost of caregiver credits varies significantly across the countries surveyed depending on the generosity of the individual program. In 2006, France's expenditures were relatively high, with the MDA, the AVPF, and the pension bonus for multiple children representing approximately 13 percent of total pension expenditures in the public pension system. (Of this, 7.4 percent was for the MDA, 1.7 percent was for the AVPF, and 3.9 percent was for the pension bonus.) To put this in perspective, total pension expenditures amounted to approximately 183 billion euros (US$237 billion) in 2006, while the cost for the three credits was approximately 24 billion euros (US$31 billion). Furthermore, the cost of those programs is projected to increase dramatically in coming years, especially that for the AVPF, as more eligible workers reach retirement and claim benefits. (To date, only a small fraction of AVPF-credit beneficiaries has reached retirement age.) According to CNAV, the cost of the AVPF is projected to double between 2004 and 2015.

In Germany, the cost of the childcare credit is also significant, albeit much less than that of France. In 2010, caregiver credits cost the government approximately 12 billion euros (US$16 billion), compared with total pension expenditures of 238 billion euros (US$313 billion). In other words, caregiver credits represented approximately 5 percent of total pension expenditures. That said, the actual cost of the program is projected be much higher in coming decades, as many of the beneficiaries of the credits have not yet reached retirement age.

Finally, the cost of Sweden's caregiver credit program in 2008 was relatively modest compared with the other countries surveyed, making up only 2 percent of total pension expenditures. In 2008, caregiver credits cost the government approximately 5 billion kronor (US$709 million), compared with total pension expenditures at around 234 billion kronor (US$33 billion). That relatively low cost was likely the result of a number of possible factors, including the design of the program; characteristics of the beneficiaries (for example, how they were distributed among the three calculation methods and how quickly they returned to work); and the program's interaction with Sweden's parental leave policy, under which beneficiaries continue to contribute 7 percent of wages for old-age insurance.

Discussion and Conclusion

The extensive experience with caregiver credits in the three countries profiled provides a number of insights with respect to the potential use of credits in the United States. First, all three countries have caregiver credit programs, which are designed to reach specific target populations. In France, caregiver credits pursue a number of different objectives, which has led to the development of three forms of credits aimed primarily at caregivers with low income, caregivers who continue to work after childbirth, and parents of three or more children. A similar mix of objectives applies to Germany and Sweden, where credits are designed to benefit not only women (and increasingly men) who leave the workforce entirely, but also those who combine unpaid caregiving with at least part-time work. It is precisely this mix of objectives that has led to the existence of such varied forms of caregiver credits in these countries, and throughout the EU more broadly.

For the United States, the mix of objectives suggests that any consideration of caregiver credits should begin by focusing on the policy objectives the credits are meant to achieve—whether it is improving benefit adequacy for caregivers (especially those who lack coverage under Social Security's current auxiliary benefits), increasing equity among men and women, or to pursue some other objective or mix of objectives—and then designing a program that pursues those objectives by targeting the intended vulnerable populations. As the European cases show, this targeting can be achieved fairly simply through the qualifying conditions and the calculation method chosen for the credits.

Second, the administration of caregiver credits has not presented any major difficulties in any of the countries profiled. For childcare credits, administration typically involves the parents supplying a birth or adoption certificate to prove guardianship. In Sweden, the process is fully automated, with the government being alerted automatically of births on a daily basis. In the case of the AVPF in France, the administrative process is also simplified by linking eligibility to the receipt of other family benefits.²⁶ The challenge is certainly greater for caregiver credits provided for unpaid care of sick or elderly relatives, given the need for more detailed information, including proof that a certain number of hours of care has been provided and that an individual is in need of such care. In Germany, however, this process can also be simplified by basing eligibility on the receipt of benefits by the core recipient, benefits that are determined on the basis of the degree of dependency for daily living.

While France, Germany, and Sweden have had remarkably little difficulty administering caregiver credits, the challenges would potentially be greater in the United States, at least initially. For example, much of the data that SSA would require to automate the administration of caregiver credits is either not immediately available or not complete. At present, SSA lacks birth records for all but the youngest generation, and the cost and difficulty of acquiring older records (held in most cases by individual cities or counties) would likely be prohibitive. In addition, the birth records of children are not always linked in SSA files to the records of their parents, a situation that would inevitably lead to many women not automatically getting the credits that they are eligible to receive. That said, the administrative burden could be reduced by simplifying the requirements at the beginning of the program's implementation; for example, by requiring parents, in order to receive the credits, to present a birth certificate at the time that they claim Social Security retirement benefits. (As noted earlier, a similar situation faced Sweden at the implementation of their caregiver credit program, leading the country to simplify the award process.)

Third, the cost of caregiver credits varies considerably across the countries surveyed depending on the design of the specific program, showing that there is nothing inherent in caregiver credit programs that makes them cost prohibitive. In a country such as France, the cost is expectedly high (and is projected to grow rapidly in coming decades) because of having three very generous programs that cover significant proportions of the population. As in Sweden, however, the cost of caregiver credits can be kept relatively low. These examples suggest that the cost of a US caregiver credit program would depend entirely on the specific design that was chosen and the generosity of the new benefit. However, an additional policy issue to keep in mind for the United States is that any new costs associated with caregiver credits would have to be funded in a program that has historically been reluctant to rely on general revenue financing.

While the near-universal adoption of caregiver credit programs across the EU suggests that the challenges of design, administration, and cost are not insurmountable, what is surprising is how little attention has been given to the issue of the effectiveness of those programs. To date, there have been very few studies analyzing the effectiveness of caregiver credit programs, and the studies that do evaluate those programs show mixed results. In their study on Sweden, for example, Stahlberg and others (2006) found that women have a higher return on their lifetime contributions than men because of caregiver credits, unisex lifetables, and the guaranteed minimum pension. Of those measures, caregiver credits were found to have the least impact on replacement rates, primarily because the other tools have been so effective. (According to Zaidi (2007), the average Swedish woman can expect to have her benefits increased by about 10 percent from childcare credits.) In France, the effect of caregiver credit programs on benefit adequacy is unknown, although the Pensions Advisory Council has noted that the absence of those programs would undoubtedly lead to women having lower retirement benefits and would probably force many of them to work more years to receive a full benefit, than at present.

In the only cross-national study to date that analyzes the effectiveness of childcare credits, D'Addio and Whitehouse (2009) found that credits do unquestionably improve pension entitlements for women. According to their simulations, the absence of caregiver credits in the countries where they now exist would lead to a 3 to 7 percentage point drop in gross replacement rates (for career interruptions between 3 and 15 years). However, there is significant cross-country variation, with credits having relatively little effect on replacement rates in some countries (for example, Denmark, Hungary, and Sweden), and large effects in others (for example, France, Germany, and Luxembourg). This variation is largely the result of the design of pension systems and the presence of other tools (for example, residency-based benefits, minimum income guarantees, and so forth) that mitigate the relative importance of the credits. It should be noted, however, that the study examined replacement rates of women as a whole; further dividing the populations into subgroups of women (particularly by income) would very likely show the impact of credits to be stronger on some subgroups rather than others.

Despite the relative lack of studies on the effectiveness of caregiver credits in the EU, credits appear to have at least a modest impact on improving benefit adequacy for many women across the region. The question to be explored and answered, however, is whether a similar effect would occur if caregiver credits were adopted in the United States. In the international literature on this defining issue, studies considering the impact of caregiver credits in the United States have also produced mixed results, largely as a result of the assumptions used and the specific form of credits used. Favreault and Sammartino (2002), for example, considered the potential impact of adding caregiver credits (specifically, crediting parents with half of the average wage for up to 5 total years in which they provide care to a child younger than age 6) as a supplement to current auxiliary benefits. The authors found that while the credits would have modest effects overall, they would be particularly well-targeted to women at the bottom of the lifetime earnings distribution (see also Favreault and Steuerle (2007)). Similar results are shown in studies that consider caregiver credits in a situation where current spousal benefits are eliminated; it was found that those credits would be more effective than current spousal benefits at reducing poverty for low-income groups and minorities (Herd 2006).

In contrast, other studies have been more pessimistic on the potential impact of credit proposals. Iams and Sandell (1994), for example, considered the poverty-reducing effect of dropout years—where up to 5 years of unpaid caregiving are excluded from the benefit calculation—and found that credits had very little impact and were targeted primarily at higher-income women. According to the authors' results, most potential beneficiaries will receive auxiliary benefits at retirement, thus mitigating the need for caregiver credits. What these studies suggest is that the effectiveness of caregiver credits on old-age poverty rates will vary significantly depending on the design of the credits (including whether a parent is awarded a fictional salary for years of caregiving, or if those years are simply excluded from the benefit calculation) and on how the credits would fit into the current Social Security system (that is, as a replacement or as a supplement to current-law auxiliary benefits).

In short, the EU experience shows that caregiver credits can be designed in multiple ways and often for multiple objectives, including improving benefit adequacy for caregivers (primarily women), promoting higher fertility rates, facilitating the return to the labor force following childbirth, and rewarding the act of providing unpaid care. The specific caregiver credit design chosen has a direct impact on how many individuals are eligible for the credits and, consequently, how much the credits cost. In addition, while this article has focused specifically on caregiver credits in the EU, it bears repeating that these credits are only a small part of a larger set of measures undertaken in these countries to protect women. Other measures, such as minimum pension guarantees and a broad range of family benefits, are used alongside caregiver credits to protect the most vulnerable populations.