Flexible spending accounts: by participating, are employees borrowing from their future?

Flexibel spending accounts (FSAs) have become popular fringe benefit options, used for medical and dependent/child care expenses. They reduce tax liability, thereby increasing taxpayers' disposable income. Pretax dollars are used to pay qualified health care and/or dependent care expenses. Thus tax is not paid on that portion of income that is contributed to an FSA. The amount an employee wants to contribute for any particular year must be decided before the beginning of the applicable year (normally by November 30).

While the benefits received under an FSA may be tax free, because of the potential reduction of eventual social security benefits, employees may be borrowing from their future. This article will discuss the relevant factors and reveal who should use FSAs.

Limitations of FSAs

After an employee decides to set up an FSA, he must choose a specific FSA contribution amount to be withheld from his paycheck on a pretax basis (before Federal income tax, FICA and FUTA amounts are calculated) and deposited in his FSA. Whether state and local income taxes are required to be paid depends on the appropriate statutes; however, the authors are not aware of any state in which these FSA contribution amounts would be taxable. The result is that the employee does not pay taxes on the amount contributed to the FSA.

Contributions to an FSA that are not used in the same period (year) are forfeited; there is no refund of any unspent balances. Once the employee has specified an amount to be contributed to the FSA, he will either use it or lose it. The amount contributed to the FSA is used to reimburse the employee for certain medical and/or dependent/child care expenses incurred during this period(1) that have not been reimbursed by any other plan.

Social Security Benefits Calculation

FSAs may reduce a taxpayer's social security benefits when social security income is less than the maximum social security tax base for that year.

When contributions to an FSA reduce social security income, future social security benefits are also reduced. According to the Social Security Administration rules, the benefits are based first on the average indexed monthly earnings (AIME). To compute AIME, each year's earnings (up to the maximum social security income taxable for that year) are indexed by the cost-of-living factor for that year to inflate them to current dollars. Next, the average of the 35 highest indexed amounts is computed. (For those born before 1929, fewer than 35 years are used.) Last, that quotient, which represents an annual average, is divided by 12.

A young employee, early in his career, may likely work for more than 35 years, and expect raises greater than the cost of living. This employee would benefit by taking fringe benefits on a nontaxable basis during these early years. Even though this strategy does not maximize yearly social security income, if the employee works more than 35 years, the earlier years' earnings will not be among the 35 highest, and therefore will not be used in computing AIME.

After computing AIME, the primary insurance amount (PIA) must be calculated. The PIA is the monthly amount the social security recipients will receive in today's dollars, adjusted for future inflation. PIA is computed by taking 90% of the first $387 of AIME, plus 32% of the next $1,946 of AIME, plus 15% of the AIME in excess of $2,333. While the percentages remain the same, the upper end of each dollar bracket will rise with inflation.

As indicated in the PIA calculation, the monthly social security benefit increases at a decreasing rate as income subject to social security increases. Therefore, in order to make the most appropriate decision regarding FSAs, a taxpayer must know his effective marginal PIA bracket (90%, 32% or 15%). (Note: All computations in this article assume that the taxpayer's current PIA bracket does not change. In other words, any change in a taxpayer's current income will be due to inflation. This assumption will be relaxed later.)

Another assumption, which is not always the case, is that social security benefits will not be subject to income tax in the year receive. The taxation of social security benefits, a concern for higher income individuals, increases the attractiveness of FSAs.

Decision Variables and Assumptions

A major variable to be considered in making the FSA decision is the employee's current salary level. The lower the salary level, the less beneficial an FSA will be because the current tax saving will be less (15% tax bracket) and the loss of future social security benefits will be greater (90% PIA bracket). On the other hand, the higher the salary level, the more likely an employee will want to take advantage of an FSA because the current tax saving will be more (28% or 31% tax bracket), while the loss of future social security benefits (15% PIA bracket) will be less significant.

The effect of state and local income taxes must also be considered, although an in-depth analysis of this factor is beyond the scope of this article. Depending on the employee's state and city of residence, the avoidance of state and city income taxes through the use of an FSA could provide additional savings. In borderline cases, state or city income taxes would make the nontaxable treatment afforded by FSAs attractive, even if social security benefits would be reduced by the FSA election.

To adjust for inflation, a number of variables used in the social security system and the Federal tax system (tax brackets, AIME factors, PIA brackets and social security benefits) are subject to indexing. The initial analyses conducted in this article, then, are based on the uninflated and undiscounted approach, assuming that the future inflation rate and the taxpayer's expected after-tax rate of return on investment will be identical. This would occur if the taxpayer could invest funds at an after-tax rate of return on investment equal to the inflation rate. This simplifying assumption is used in the initial analysis in order to evaluate a conservatively extreme condition. The effects of relaxing this conservative assumption to evaluate another extreme condition and a more reasonable condition will be discussed later. Generally, it can be said that relaxing the conservative assumption renders the FSA choice more attractive.

Using this simplifying assumption, social security benefits can thus be computed based on today's social security rules since they will be indexed with future inflation. With the social security system and the Federal tax system variables adjusted yearly for inflation, the current tax savings can be compared with social security benefits lost because the authors assume that the after-tax rate of return on investing the current tax savings equals the inflation rate used to increase all of the variables. In other words, it is assumed that the after-tax real rate of interest (interest net of inflation) is zero.

Based on this simplifying assumption, the FSA decision is affected by the number of postretirement years in which the taxpayer will receive benefits and whether the spouse is collecting social security based on the employee's earnings. If the spouse's PIA is less than 50% of the employee's, the spouse's benefits will be based on the employee's benefits. If this is the case the spouse will receive 50% of the employee's amount.

Results

* Lower tax brackets

Exhibit I on page 319 compares the current effect on taxes and earned income credits to the future loss of social security benefits when $100 is sheltered in a medical care FSA. Column 1 provides computations for an employee who is in the 90% PIA bracket and 15% Federal income tax (FIT) bracket. The 90% PIA bracket applies to those taxpayers whose average annual earnings over the past 35 years, indexed to current dollars, are less than $4,656. Such an employee is assumed to be in the 15% Federal income tax bracket , but is more likely to be in the 0% tax bracket.

TABULAR DATA OMITTED

In the 15% bracket, an FSA of $100 would save $15 in income tax for one year. There would also be a savings of $7.65 in FICA tax, for a total savings of $23 (rounded up). However, $17 in earned income credit (EIC) would be lost. In addition, $51 in social security benefits would be sacrificed [$100 [devided by] (35 years x 12 months) x 90% x 12 months x 20 years], assuming the employee is collecting for 20 years. Thus, the net loss from the decision to have an FSA is $45. The loss would be even greater, $71 [($51 x 150%) + $17 - $23], if the employee's spouse is expected to collect social security benefits based on the employee's earnings. Thus, taxpayers in this income bracket should not use a medical FSA if they plan on collecting social security benefits for more than two years.

Columns 2 applies to taxpayers in the low end of the 32% PIA bracket (average annual earnings, indexed to current dollars, between $4,656 and $7,100) who, for the most part, will be in the 15% tax bracket and the 7.65% FICA tax bracket. With such low annual earnings they are still among those who would lose $17 of EIC for each $100 reduction in reported income.

Thus, the FSA will save $23 in current Federal income and FICA taxes for each $100 contributed, while costing $17 in EIC and $18 in social security benefits over 20 years [$100 [devided by] (35 years x 12 months) x the 32% PIA rate x 12 months x 20 years].

If the spouse also collects on the employee's benefits, this family would lose $27 in social security benefits over 20 years. As shown in the exhibit, employees in this income range should avoid FSAs unless they expect to live a rather short time after retirement. The results are even more pronounced if the spouse were collecting on the employee's PIA. The couple would not want to use an FSA if they collected benefits for more than approximately four years, depending on the potential avoidance of state and local taxes. Further analysis reveals a break-even income tax bracket for the 32% PIA bracket of 27% for a single social security collector. In other words, at a 27% marginal Federal plus state plus local income tax bracket, a taxpayer would save as much as he would lose in an FSA. When a taxpayer's marginal tax bracket is in excess of 27%, he would be wise to use an FSA.

Column 3 applies to taxpayers in the higher end of the 32% PIA bracket. Here, they do not lose EIC as they reduce reported income. From $7,100 to $11,250, there is no effect on EIC; from $11,250 to $21,250, they gain $12 in EIC for each $100 reduction in reported income. (Above $21,250, no EIC is available.) Thus, as the note in the exhibit explains, taxpayers in this special range ($11,250 to $21,250) should be quite tempted to use the FSA. The decision is somewhat more difficult for others: All taxpayers in this column would lose $18 in social security benefits over 20 years of collecting individually for every $100 in reported income in a year, or $27 if their spouses were also collecting.

In column 4, as average reportable income over the past 35 years exceeds $28,000, the PIA rate drops to 15%. In this PIA bracket, a $100 reduction in reported income in a year costs a social security recipient only $9 in reduced benefits over 20 years of collecting, or $14 if the spouse is also collecting. As the exhibit shows, one would have to expect to collect for many more years before deciding to reject the FSA option.

* Higher tax brackets

Columns 5 applies to someone in the 15% PIA bracket and the 28% Federal income tax bracket. Thus, an FSA contribution of $100 will save $36 in Federal income tax and FICA tax. Assuming the taxpayer has been earning the equivalent of $42,000 in today's dollars for the past 35 years, the AIME would be approximately $3,500, with or without an FSA contribution. The PIA would be approximately $1,146 with or without an FSA (the difference would be only four cents).

In this situation the AIME decreased by the same amounts as for individuals in lower tax brackets. However, the PIA decreased by a much lesser amount due to the lower marginal PIA bracket assigned to the higher income taxpayer. Thus the social security benefits over 20 years would be $275,056 without the FSA, and $275,047 with the FSA, representing a loss of $9 in social security benefits and a gain of $36 in current tax savings. The taxpayer would have to collect benefits for 80 years before an FSA would be disadvantageous.

If the taxpayer's spouse also collects benefits based on the taxpayer's PIA, their total collections over 20 years would be $412,585 without the FSA and $412,571 with it. Thus, $14 worth of benefits would be given up in order to save $36 in current taxes. Therefore, an FSA would be advantageous unless they both lived to collect for 51 years.

Column 6 relates to taxpayers who would be above the social security tax minimum, currently $55,500. Thus, the savings related to an FSA contribution would not reduce the social security tax, only the Federal income tax. On the other hand, an FSA contribution would still be to their advantage, because their social security benefits would not be reduced. It should be noted that for earnings of less than $130,200 but over $55,500 taxpayers will save the 1.45% Medicare tax. Taxpayers in the 31% Federal income tax bracket are essentially in the same position, except that they get an additional $3 worth of tax savings for every $100 in an FSA, because of their 3% higher tax bracket.

As mentioned earlier, one of the assumptions regarding earnings stability will now be relaxed, and the impact on a young taxpayer expecting his salary to increase faster than inflation will be examined.

One possibility would be that future salary would increase so much that the current year's earnings would not be included among the 35 highest indexed amounts in computing AIME for social security benefits. In this case, there are no losses of future benefits, and the FSA should always be taken.

The other possibility is that the current year's salary would be included among the 35 highest, but the higher future earnings will have the effect of pushing the AIME into a lower marginal PIA bracket. For example, a taxpayer currently earning $25,000 would be in the 32% PIA bracket, but if future earnings increase faster than inflation, he could ultimately collect benefits in the 15% PIA bracket. In this case, Exhibit I could still be used, but the advice should be read from under the future PIA bracket, not the present one. That is, current taxes saved would still be $23, but benefits would be computed such that the 51 years of collecting individually, or 33 years of collecting as a couple, would be the applicable cutoffs (per column 4).

A converse possibility - which could strain the stable earnings assumption - would be that of an older taxpayer expecting a drop in income, due to partial retirement. Most likely, the taxpayer would have already accumulated the 35 highest indexed years before partial retirement, so that any future reduced earnings will not affect social security benefits; therefore, the exhibit is still valid. In any case, it would be quite unlikely that subsequent reduced income would be low enough to cause the 35-year average to be pushed to a higher PIA bracket, while high enough to be included in the 35-year average in the first place.

* Dependent care FSAs

Exhibit II on page 322 relates to dependent care FSAs, which follow the same general pattern as the health care option. However, all wage earners have another desirable alternative if they do not take the FSA: the dependent care credit. Given this, the FSA benefits must be compared not only to all of the factors as the health care FSA, but also to the lost dependent care credit that would otherwise be available. Exhibit II reveals that the FSA is a bit more difficult to justify for incomes below $55,500. The taxpayer needs to be collecting social security benefits for a shorter period than with a health care FSA to decide whether to reject the dependent care FSA.

TABULAR DATA OMITTED

For example, taxpayers with incomes below $24,000 would rarely want to choose the FSA. As column 1 indicates, the lost dependent care credit would exceed any taxes saved, even before social security benefits are considered. There is an exception, discussed in the note to Exhibit II, for taxpayers in the $11,250 to $21,250 income range. Taxpayers in this range can get increased EICs as they reduce their reported income. Thus, especially toward the high end of the range - where the dependent care credit is lowest - they might collect social security benefits for a short enough period of time (thereby suffering less from the reduced benefit level) to justify the FSA option.

The taxpayers in columns 2 and 3, in the income ranges of $24,000 to $35,800, need to have very low estimates of their life expectancies before they would choose the FSA. In the PIA range below $28,000 they would be losing too much in potential social security benefits. Above $28,000, they would lose less ($9 in benefits over 20 years if they collect on their own, $14 if the spouse collects also). Still, this would rarely offset the minimal current savings in taxes when reduced by the current $20 lost dependent care credit.

In column 4, the marginal income tax rate increases above $35,800. Now, the current tax savings would justify the FSA, unless the taxpayer lived a very long time with the distress of reduced benefits. Finally, in column 5, the FSA is always justified for income levels above $55,500, because there is no reduction in benefits.

Again, the assumption regarding earnings stability may be relaxed. For an upwardly mobile young professional, the advice in Exhibit II generally is still valid. This is because below $28,000 adjusted gross income (AGI) dependent care FSAs can rarely be justified, based strictly on the current year's tax effects. Even before possible social security benefit losses are considered, the loss of the dependent care credit offsets the gain from any tax saved. For individuals with salaries above $28,000, the minimum PIA bracket has already been reached, and consequently the exhibit is applicable because the amounts used to compute the benefits remain valid even as future earnings increase.

For an older employee contemplating future partial retirement, the situation is the same as under the medical care FSA. Future reduced earnings will not alter the advice in the exhibit because of the negligible probability that the future earnings will cause a change in the PIA bracket.

* Relaxing the return on investment assumption

The analyses presented thus far have been based on the assumption that the future annual inflation rate and the expected after-tax rate of return on investment are identical. As stated earlier, this is a conservatively extreme assumption.

In relaxing that assumption, two other situations can be analyzed. One involves the extreme case of a taxpayer faced with an after-tax rate of return on investment that is 12 percentage points greater than the annual inflation rate. This might apply to a taxpayer whose best investment alternative is paying down a credit card balance. The other situation is a more moderate assumption that the after-tax rate of return on investment will be three percentage points greater than the annual inflation rate.

The results reveal that an after-tax rate of return that is greater than the inflation rate causes the FSA to be a more attractive choice. The higher the interest rate, the less significant are the future social security benefits lost by using an FSA. The analysis is complicated, however, by the fact that the introduction of an interest rate necessitates the consideration of the number of years the taxpayer has remaining before retiring.

When interest rates are considered, a medical care FSA is always justified, except for extremely low income levels. At a 3% interest rate, taxpayers with incomes below $7,100 should not use the FSA. At a 12% rate, only those with incomes below $4,656 should avoid the FSA.

The dependent care FSA is more complicated because of the lost tax credit. As shown in Exhibit III, on page 323, taxpayers with incomes below $24,000 should never use the FSA, regardless of interest rates, except when the EIC is involved. The EIC causes taxpayers in the $11,250 to $21,250 income range to get increased EIC as they reduce their reported incomes, usually justifying the FSA. At the other end, taxpayers with incomes above $28,000 should, as a rule, always take the FSA. For incomes between $24,000 and $28,000, the exhibit provides the appropriate advice.

TABULAR DATA OMITTED

Additional Factors

An FSA may result in savings to the employer because an employer must pay the same FICA amount as the employee. Thus, because an FSA may reduce an employee's income subject to social security tax, it may reduce the employer's social security tax expense as well.

All taxpayers should obtain their Personal Earnings and Benefit Estimate Statement from the Social Security Administration to help determine their marginal PIA rate. This is important because the PIA rate is influenced by the employee's social security income for the current and previous years.

There is a greater reduction of social security benefits from the FSA decision if the spouse will be receiving social security benefits based on the earnings of the employee rather than the spouse's own earnings. On the other hand, if the spouse's benefits are based on the spouse's own social security income, the employee might be persuaded to treat the fringe benefits as nontaxable income. However, the spouse's social security benefits are not a major consideration for those who have extremely low or extremely high income.

Conclusion

An individual's income history can be use to determine the impact of a flexible spending account on future social security benefits by taking into consideration the employee's marginal PIA rate; the Federal income tax rate; the potential effect on the earned income credit; the spouse's retirement situation; as well as state and local taxation. This study found that the greater one's income, the less effect an FSA has on social security benefits, which enhances the value of having an FSA. Lower and middle income taxpayers, however, need to consider carefully whether the benefits of having an FSA exceed the costs. And finally, the number of years remaining until retirement is relevant as one considers varying the rates of return on investment.