Evaluation of Flexible Spending Accounts for Active-Duty Service Members

Unlike many large employers, including the federal government, the military does not offer flexible spending account (FSA) options to members of the armed services and their families. The House Armed Services Committee Report accompanying the National Defense Authorization Act for fiscal year 2021 directed the Secretary of Defense to report on the feasibility of implementing flexible account options that allow pre-tax payment of dependent care expenses, health and dental insurance premiums, and out-of-pocket health care expenses for service members and their families (U.S. House of Representatives Report 116-617, 2020). As detailed by the directive, the report should include an assessment of the tax incentives when using FSA options and the financial advantages or disadvantages to service members and their families and should include identification of any legislative or administrative barriers to implementing these options.

Contributions to either a health care FSA (HCFSA) and/or dependent care FSA (DCFSA) reduces the amount of income subject to income and payroll taxes, thereby reducing the individual's tax liability. FSAs interact with other tax incentives in the U.S. tax code, potentially reducing or even eliminating the potential tax savings to individuals with an FSA. From the standpoint of the employer, such as the U.S. Department of Defense (DoD), FSAs reduce Federal Income Contribution Act (FICA) taxes that consist of both Social Security and Medicare taxes but impart a cost in the form of administration costs and implementation or start-up costs. Thus, a priori, it is not possible to predict whether FSAs would represent a net benefit to service members and their families in terms of reduced tax liabilities or a net reduction in costs to DoD.

The Office of the Secretary of Defense (OSD) asked the RAND National Defense Research Institute to provide analytic support for its report to Congress, including an implementation plan should FSA options be implemented by DoD. This study documents that support. Our analysis focused on active-duty personnel and sought to evaluate the benefits and costs to service members of FSAs for dependent care and health care and the costs and savings to DoD. For contextual background, we drew information from policy documents and other sources on existing health care and dependent care benefits available to military members and civilians. To help understand the elements of cost and savings to DoD, we held structured discussions with subject-matter experts (SMEs) on administrative costs and we also gathered available inputs on administrative costs. We used publicly available tax simulation programs to evaluate the tax benefits and costs to service members of FSAs for dependent care and health care and compute the aggregate net change in tax liabilities to service members and the net costs and savings to DoD under a variety of assumptions regarding the share of members who might participate in the FSA options. Finally, we drew together the results of the analysis to assess the feasibility of providing FSA options to service members and discuss potential legislative changes that could improve the net benefit of FSAs to members and their families. We note that our tax-accounting approach does not consider how FSAs might change the members’ incentives to consume health care or dependent care or issues related to the efficiency or social welfare implications of providing a health and dependent care benefit through the U.S. tax code.

Key Findings

Many Members Would Have Few or Even No Eligible FSA Expenses Under Current Law

For members to take advantage of an FSA, they must have eligible expenses, usually in the form of dependent care expenses and out-of-pocket health care expenses for themselves or their family members. In the case of health care, most members would have few or no eligible out-of-pocket medical care costs associated with TRICARE because 82 percent of family members were enrolled in TRICARE Prime and incurred an average of $100 annually in out-of-pocket health care costs between 2018 and 2020. Meanwhile, 17 percent of family members enrolled in TRICARE Select and incurred higher costs, an average of about $500. HCFSAs also cover health-related expenses outside TRICARE, including over-the-counter medication, eyeglasses and contact lenses, and orthodontia expenses. Estimates for these expenses for the active-duty population are unavailable, but national estimates suggest that out-of-pocket expenses outside health insurance are about $400, although the U.S. population differs demographically from the military population and out-of-pocket expenses could also differ. We note that orthodontia expenses specifically have an estimated average cost of $3,100¹ in 2021 dollars among those incurring these expenses (Hung et al., 2021).

Dependent Care Expenses

In the case of dependent care, because of the way in which DoD provides child care benefits, an unknown share—and, potentially, a large share—of military families may not have eligible dependent care expenses. Using data from the 2019 Survey of Active Duty Spouses conducted by DoD (Office of People Analytics, Department of Defense, 2020), we estimate that 65 percent of military spouses are working and have sufficient earnings to qualify for a DCFSA and have sufficiently young children to potentially benefit from a DCFSA.² However, expenses for off-base child care subsidized by DoD are unlikely to be FSA eligible. The reason is that the subsidy offsets the eligible expenses for a DCFSA dollar for dollar; any combination of child care subsidy and DCFSA contributions cannot exceed the maximum DCFSA contribution limit of $5,000. That is, expenses that would be eligible for DCFSA contributions and expenses that are subsidized are not stackable.³ According to the 2019 Survey of Active Duty Spouses, 61 percent of spouses who use child care report using child care that is provided off base by a civilian provider. Many military families seek off-base care because care at on-base child development center (CDC) facilities is supply constrained, meaning that there are more members seeking care for their children at these centers than there are spaces for children, and CDCs have wait lists (Kamarck, 2020). We have no information on what share of these 61 percent of spouses accesses fee-assisted or subsidized care and therefore would be unlikely to have eligible DCFSA expenses. It is likely that some of this off-base care is unsubsidized (implying that expenses would be FSA eligible), as we discuss in the next subsection, but more information is needed on the extent.

FSAs Would Confer a Tax Savings for Many But Not All Members with Eligible Expenses

For those military families who do have eligible expenses, we estimate that the HCFSA option would confer a savings but that the DCFSA would not do so in all cases. Contributions to an FSA reduce the amount of income subject to taxation, thereby reducing the amount of income tax and payroll taxes (the employee's contribution to Social Security and Medicare taxes) that a member would pay, suggesting that FSAs would produce a tax savings for service members. However, FSA contributions interact with other components of the U.S. tax code, specifically the Child Tax Credit, the Child and Dependent Care Tax Credit (CDCTC), and the Earned Income Tax Credit (EITC), sometimes in ways that are not necessarily favorable on net to the taxpayer. These credits have phase-in, plateau, and phase-out points as income changes—and, by reducing taxable income, FSA contributions affect how much of these credits can be claimed and the tax liability of the taxpayer. Consequently, on net, the tax liability for service members does not necessarily decrease in all cases.

We estimate the total change in tax liability for married and unmarried service members under different assumptions about the number of children, whether spouses are working, and FSA contribution amounts, and we consider results by paygrade and by total family income (TFI) (before FSA contributions). Considering results by TFI allows us to consider cases where members might earn special and incentive pay (raising family income) or spouses have higher than average earnings. (We estimate that average spouse earnings were $25,000 in 2020).

In the case of dependent care, we estimate that the total change in tax liability from contributing $5,000 to a DCFSA would not always be advantageous to the service member. That is, we find cases where the tax liability remains unchanged or increases slightly.⁴ Figure 1 illustrates the change in total tax liability of contributing $5,000 to a DCFSA for a married service member in 2020. The figure shows the complex, nonlinear nature of the U.S. tax code with the tax change rising and falling at different earnings levels as certain credits phase in and out and as tax brackets change. In general, for lower levels of earnings, the tax benefit from contributing to a DCFSA is driven by increases in the amount of EITC that can be claimed, and to a smaller extent, increases in the amount of the refundable Child Tax Credit. As the EITC phases out at higher earnings levels, the tax benefit from contributing $5,000 to a DCFSA decreases. Once the EITC phases out (i.e., earnings exceed the means tested threshold to claim the EITC), the tax benefit from contributing to a DCFSA plateaus until earnings reach a level where the service member moves up to the next tax bracket and the reduction in income tax (before credits) increases. Additional nonlinearities in the tax code cause the tax benefit from contributing to a DCFSA to increase and decrease at higher earnings levels.

Married DCFSA-contributing service members who earn $30,000 to $50,000 annually and have one child will get less tax benefit than service members earning the same amount and who have two children (about $1,200 for one child and $1,400 for two children). However, once the salary goes above $60,000, having one child provides a better tax benefit than two children (generally about a $400-dollar difference, ranging from $0 to $1,000 for two children and from $400 to $1,400 for one child as the salary increases).

NOTES: Gross earnings are equal to family earnings without the FSA contribution deducted. The x–axis in the figure shows gross family earnings, meaning these are earnings without the $5,000 reduction from the DCFSA contribution. For married service members, we assume that the military spouse earns at least $5,000, so the couple qualifies for the maximum DCFSA contribution. We used publicly available tax simulation programs to estimate the change in federal income tax from contributing $5,000 to a DCFSA and add in the estimated payroll tax reduction of $382.50 to calculate the total tax change.

The results in Figure 1 show that over the range of $59,000 to $105,000 of TFI, a married service member with two children and $5,000 of eligible DCFSA expenses would not benefit from having the FSA; the change in total tax liability is virtually zero (or, more precisely equal to +$17.50). This occurs because of a loss in the CDCTC, which exceeds the reduction in tax before credits and payroll taxes.

We estimate that virtually all members with eligible health expenses would experience a tax savings from contributing to an HCFSA regardless of marital status and number of children with the tax savings increasing when HCFSA contributions are larger. Figure 2 shows the change in total taxes from contributing $500 to an HCFSA for a married member with a working spouse. Although all members are estimated to experience a savings in Figure 2, the amount varies with income because of the phase-in and phase-out of different credits at different income levels and because the tax brackets vary with income.

Service members who contribute $500 to HCFSA, have a working spouse, and earn $50,000 to $170,000 see the same tax benefits (which range from $100 to $200) whether they have no children, one child, or two children. Above and below that salary range, benefits increase as service members have more children (though none go above above $200).

NOTE: Gross earnings are equal to family earnings without the FSA contribution deducted.

Across Service Members, FSAs Would Confer a Net Tax Benefit

Ideally, to estimate the total net change in tax savings or costs across service members, we would compute the share of members who would likely participate in the HCFSA and DCFSA options and then compute the cost or savings for those members and for DoD. Because we lack sufficient data to estimate the HCFSA and DCFSA participate rates, we compute savings and costs under alternative participation rate assumptions and then use available, albeit incomplete, information on which participation rates seem the most likely.

For the DCFSA option, we use (1) Defense Manpower Data Center data for the number of service members on active duty in 2020 and the number with children under age 13 and (2) 2019 Survey of Active Duty Spouses on the share of service members with a working spouse to estimate the total tax savings to service members contributing $5,000 to a DCFSA under different DCFSA program participation rates. We estimate an annual aggregate tax savings of $30.7 million if 15 percent participate, $51.2 million if 25 percent participate, $102.4 million if 50 percent participate, and $204.9 million if 100 percent participate. A participation rate of between 25 percent to 50 percent may be more realistic than a 15-percent or 100-percent participation rate. As discussed earlier, 64 percent of employed spouses report using child care, an upper estimate of the rate of participation in a DCFSA. However, the 64-percent figure is too high as an estimate of a participation rate because not all of these military families with an employed spouse would have eligible DCFSA expenses given that some families receive subsidized off-base child care that offsets the tax benefit of a DCFSA dollar for dollar. The more realistic estimate is somewhere above 24 percent but below 64 percent. The 24 percent corresponds to our estimate of working spouses using child care who would use on-base child care.⁵ But some families with working spouses use off-base unsubsidized care that would be eligible for a DCFSA, so the participation rate would be higher than 24 percent, but we do not have an estimate of how much higher.⁶ Given this context, a participation rate between 25 percent to 50 percent may be more realistic leading to an estimated aggregate benefit of $51.2 million to $102.4 million.

In the case of an HCFSA, we assume that all active-duty service members are eligible to participate in an HCFSA. We estimate an annual aggregate tax savings of $19.8 million assuming 15 percent of active members participate in the HCFSA and contribute $500 annually. That figure increases to $33.1 million assuming a 25-percent participation rate to $66.1 million assuming 50 percent participate to $132.3 million assuming a 100-percent participation rate. Given that 17 percent of active-duty families were covered by TRICARE Select and averaged $500 annual out-of-pocket expenses, the actual participation rate may be closer to the 25-percent figure than the 50-percent figure. On the other hand, all members could have other eligible out-of-pocket expenses outside TRICARE, equal to about $400 for the civilian population, so participation could be higher.

FSAs Would Confer a Cost Savings to DoD but Could Result in Large, Albeit Unknown Implementation Costs

We also estimated the aggregate cost to DoD's accounting of administering FSA options and the savings to payroll taxes from reduced Social Security and Medicare taxes. We used information from the U.S. Office of Personnel Management (OPM) (U.S. Office of Personnel Management, 2018; U.S. Office of Personnel Management, 2020; and U.S. Office of Personnel Management, 2021) on the cost of administering the FSA options for federal employees, noting that administration costs are lower after the first year when more reserve funds are built up to handle overpayments and other situations. Table 1 shows the estimated cost implications under different assumptions about participation rates assuming members contribute $5,000 to a DCFSA (bottom panel) or contribute $500 annually to an HCFSA (top panel). Assuming a 25-percent participation rate for each program, we estimate that DoD would save $6.0 million annually for the HCFSA program and $31.6 million for the DCFSA program, after the first year. These figures exclude implementation costs and ongoing overhead costs, such as the cost of training personnel about the FSA options.

We held discussions with SMEs to better understand these implementation and ongoing overhead costs. The experts identified costs associated with developing the processes and contractual arrangements required to administer the FSA plans and the costs of providing financial literacy training to members. In addition, the pay systems would need to be modified to accommodate the FSA options. They noted that using off-the-shelf training and OPM to assist in plan administration would help reduce the first two costs, but the costs of adapting the pay systems could potentially be high if implementation involved adapting both the legacy pay systems and the forthcoming service-specific systems, ranging up to $28 million or more. The services are each undertaking the development of an integrated pay and personnel system, and each is expected to roll out these systems in the coming years, although on different schedules. Because of this change in systems and the different service schedules, the legacy pay systems would need to be modified, not just the systems under development. Implementation costs would be lower if DoD only implemented the FSA options in the new systems, thereby avoiding the necessity of changing the legacy pay system. But, even focusing on only the new systems, several of the experts observed that adapting or creating the pay systems needed to support FSAs could prove to be a substantial challenge, both in terms of cost and time, and could result in delays in the roll out of the new systems.

Legislative Changes Could Expand the Benefit of an FSA to Service Members

Our analysis estimates the costs and benefits to service members and their families of FSA options under current law. However, changes in current law could expand the value of an FSA to members. First, current regulations do not allow for health FSA funds to be applied to health and dental insurance premiums. Thus, enabling legislation would need to be passed to allow for payment of insurance premiums from FSAs for military personnel. Active members and their families have no premium or enrollment fee for TRICARE Prime or Select, but families pay a premium for dental care equal to $11.65 and $30.28 per month in 2022, respectively, for single members and members with families. Second, members who receive subsidized off-base child care would be unlikely to benefit from a DCFSA, because contributions to an FSA are offset by the subsidy. This issue could be addressed by eliminating the offset for the employer subsidy of child care. However, eliminating the offset would require legislation to make an employer child care subsidy and eligible DCFSA expenses stackable. Allowing them to be stackable would likely increase participation in the DCFSA. Finally, FSAs for service members could be implemented faster—perhaps up a year faster—if enabling legislation to support special features, such as payment of insurance premiums, was not required.

Wrap-Up

Introducing an FSA option could impart a tax benefit for some members and their families and save personnel costs for DoD, although the start-up costs of updating the legacy and new pay systems could be substantial. However, many military families would not have eligible DCFSA expenses, and families could have few eligible HCFSA expenses under current law, unless they use the HCFSA to cover expenses outside TRICARE, such as for over-the-counter medication and supplies. Changes in law could expand the value of FSA options to service members.