Postdoc job titles and classifications are commonly assigned by institutions based on the source of funding for the postdoc. This decision determines whether or not the postdoc qualifies for health benefits, retirement fund options, and even access to the institutional library. Sherry Queener, associate dean of the Indiana University Graduate School, explains the situation at her university as follows: "Postdocs typically are divided into at least two categories: employees and trainees. These divisions are often driven by [the] funding mechanism, rather than real differences in the purpose of the training/work. Thus, both categories of postdocs might work side by side on similar problems. Because these two categories of postdocs require different fringe benefits and because they often receive salaries determined by the funding mechanism, the universities are placed in the position of creating inequities, i.e., different pay for the same work."
Indiana University is not unusual in this regard. Institutions are guided--or hampered, as the case may be--by local and federal policies regulating the way they treat employees and trainees. What follows is the second in a series of articles examining current law regulating the taxability of fellowships , employee-employer relationships, and special rules for trainees. Because I am not a lawyer, accountant, or tax specialist, I strongly urge you to seek the advice of a professional to determine your specific rights and responsibilities.
Investigator-initiated (institutional) grants
Unlike postdocs funded by an individual fellowship, postdocs supported by NIH R01s or similar investigator-initiated grants are hired to work on a specific project. Because they are being compensated for services, i.e., there is significant quid pro quo, postdocs working under these circumstances are considered to be employees ( Loo v. Commissioner, 1954; 22 T.C. 220; see Part I). One result: Employee postdocs fill out W-2 forms, and the institution manages the withholding of income tax and FICA tax, as well as paying FUTA tax (see sidebar).
Some funding agencies specify that postdoctoral fellows supported with their funds be considered employees (see for example Postdoc News Bytes, 11-1-2002 ). Others specifically forbid that fellows be employees. National Institutes of Health (NIH) fellows under the National Research Service Act (NRSA) fall into the latter category. Both types of fellowship are subject to income tax.
The Congress of the United States enacted the NRSA Program in 1974 (42 USCS 288) to help "ensure that highly trained scientists will be available in adequate numbers and in appropriate research areas to carry out the Nation's biomedical and behavioral research agenda." Under this congressional authority, NIH awards individual NRSA postdoctoral fellowships (F32's in NIH talk). In 2002, NRSAs were renamed the Ruth L. Kirschstein National Research Service Awards.
Until 1984, the NRSA program was specifically mentioned in Section 117 of the IRS code: "Any amount paid to, or on behalf of, an individual from appropriated funds as a national research service award under section 487 of the Public Health Service Act shall be treated as a scholarship or fellowship grant." NIH refers to IRS Section 117 in its June 1, 2000 F32 policy document, PA-00-014,  which says: "The taxability of stipends, however, in no way alters the relationship between NRSA fellows and institutions. NRSA stipends are not considered salaries. NRSA fellows are not considered to be in an employee-employer relationship with NIH or with the institution in which they are pursuing research training. It is therefore inappropriate and unallowable to charge costs normally associated with employment (such as FICA, workman's compensation, unemployment insurance, etc) to the fellowship."
Private Letter Rulings (see PLR 9851002, 199933021, 200042027, and 200226005, and Part III of this series) on the subject of postdoctoral fellowships all use the NRSA as a model, noting that the IRS does not consider the research and training activities sponsored by NRSAs as constituting the performance of services. The PLRs cite Revenue Ruling 83-93, based on Treasury Ruling 1.117-4 [26 CFR 1.117-4] as the basis for establishing the precedent that, like fellowships, income derived from an NRSA or similar training grant does not constitute "wages," as specified under FICA.
University administrators cite these rulings as the reason NRSA fellows cannot participate in institutional pension plans. There are many different kinds of retirement accounts; some include contributions from the individual, and some include contributions from the institution. It is not clear whether FICA or FUTA payments are inextricably linked to retirement accounts of any kind. According to Wally Schaffer, Research Training Officer at NIH, "The University of California has considered contributions to retirement accounts to NRSA recipients. So my assumption is that these concepts can be teased apart. We've even considered the idea of allowing some of the funds provided under NRSA to be used as a contribution to a retirement account. None of the lawyers we've talked to have indicated that this would be prohibited."
So if it isn't wages, why pay taxes?
If the NRSA stipend isn't considered compensation, why must postdoc fellows file quarterly tax statements? Why don't some postdocs get any record (besides the check itself) of receiving a stipend? The Treasury ruling cited above states that for training grants and fellowships:
Such amounts are not subject to income tax withholding by the administering institution, making the recipient responsible for filing of quarterly taxes. As with fellowship awards, such amounts are not subject to FICA or FUTA. The administering agency is not required to file Forms W-2 or any returns of information under Section 6041 with respect to these grants. Because such amounts do not represent payment for services, they are not subject to Self-Employment Contribution Act (SECA) taxes imposed by Section 1401.
Such amounts are not subject to income tax withholding by the administering institution, making the recipient responsible for filing of quarterly taxes.
As with fellowship awards, such amounts are not subject to FICA or FUTA.
The administering agency is not required to file Forms W-2 or any returns of information under Section 6041 with respect to these grants.
Because such amounts do not represent payment for services, they are not subject to Self-Employment Contribution Act (SECA) taxes imposed by Section 1401.
In the past, NIH has encouraged institutions to report NRSA income on 1099 forms. Schaffer explains that, "we stopped doing this when it was pointed out that the instructions for 1099 forms specify that they are not to be used to report fellowship or scholarship income." Since the W-2 is not an appropriate reporting vehicle, either, the NIH has concluded that there is no way for institutions to report out NRSA income. Some institutions still use the 1099, but others issue no forms to fellows. However, if IRS instructions are followed, trainees and fellows are required to report this income. "My guess is that many do not," says Schaffer.
The benefits of overhead
Another way that individual grants differ from institutional grants is in the overhead allowed. Institutional grants usually charge at least 25% overhead; these funds are used to pay for and administer benefits and professional development services, along with the cost of electricity used in the lab and library, water and sewer for the lab and bathrooms, security patrols on campus, research administrators to review contracts, and much more.
The Office of Management Budget (OMB) Circular A-21 spells out the allowable overhead costs for researchers employed on federal grants. This means postdoc employees on federal grants are included when calculating grant overhead. This translates into access to institutional benefits plans and full access to services and facilities.
This is the same OMB circular that specifies the following: "compensation paid as, or in lieu of, wages to students performing necessary work are [sic] allowable provided that there is a bona fide employer-employee relationship between the student and the institution for the work performed." That is, graduate students doing work paid for with federal grant funds must be employees. There is no wiggle room on this--unless, that is, a proposed OMB revision to Circular A-21  is accepted.
According to Robert Hardy at the Council for Government Relations, an association of research universities, "This change is based on recommendations made by the National Science and Technology Council  a couple of years ago to improve the government-university research partnership and reflected in OMB Memorandum M 01-06 . Consistent with the OMB memo, the proposed revision would condition the allowability of tuition remission on federal awards on the basis of the individual's participation in the project, not on the existence of an employer-employee relationship for tax purposes." This means graduate students wouldn't need to be considered employees to benefit from federal overhead rates and, ultimately, this may have implications for nonemployee postdocs as well. You can find the text of the proposed revision at here  (see right-hand column of chart near the bottom).
In contrast to institutional grants, many private funding agencies provide little ($1000/yr) or no institutional allowance, and some restrict the use of these funds so that they cannot cover costs of benefits. NRSAs are generally accompanied by one of the most generous ($5500/grant) institutional allowances of any funding agency. The NRSA allowance can be utilized to cover the costs of health insurance coverage, explained Roslyn Orkin, Assistant Dean of Faculty Affairs at Harvard Medical School. "However, institutional allowances were also designed to cover costs of supplies, small equipment, and travel to scientific meetings. If this funding is instead used for health benefits, it is no longer available for professional usage." she adds In addition, because federal laws prohibit NRSA recipients from receiving additional compensation from NIH funding sources, most principal investigators do not have the resources to provide benefits to the postdocs who write and receive these prestigious awards.
A grand unified theory?
Many administrators and postdocs feel that fellows are being punished for writing and receiving individual grants (see, e.g., the Postdoc Network AAAS Meeting report ). Some universities have gone so far as to actively discourage postdocs from applying for NRSAs and other individual grants that prohibit employee status. A postdoc association representative at one southwestern institution explained that her college decided the equitable thing to do was "to eliminate retirement, vacation and sick-leave benefits to all postdoctoral fellows in the institute."
Emory University has crafted a policy that treats non-NRSA postdoctoral fellows as full employees. According to the Emory Controller's office, "The determining factor is the substance of the work and the relationship with the University. The service component is an important part of this determination." In this case, the institution is exercising its prerogative to establish an employer-employee relationship. Still, NRSA recipients are classified separately and have access to fewer institutional benefits and services.
Can a single employee classification be adopted for all postdocs, regardless of funding source? While conducting the Postdoc Network survey on postdoc compensation and benefits , we found that some universities had been audited by the IRS after trying to establish a uniform benefits policy for postdocs, and some administrators refused to comment on the record about their policies, for fear of an audit.
In Part III of this series, we will discuss applicable case law and federal policy.