It has been an interesting summer. Over and over, I kept hearing that things were changing. Not only were new postdoc offices popping up all over the place-- the University of Medicine and Dentistry of New Jersey , New York University's Sackler Institute , the University of California, Irvine , the Cleveland Clinic Foundation?s Lerner Institute , to name but a few--but policies governing the postdoc experience were moving beyond the talking stage and into practice. The University of Alabama at Birmingham and Stanford, for example, recently instituted grievance policies for postdocs . Several institutions are establishing new benefits policies . And minimum salary policies are becoming more common.
While it is important to rejoice at these developments, there is still much work to be done. Earlier this summer the Postdoc Network asked postdoc offices to share with us their postdoc compensation and benefits policies. At the same time, both the Committee on Science, Engineering, and Public Policy (COSEPUP) and the Federation of American Societies for Experimental Biology (FASEB) were engaged in similar activities. The Postdoc Network made arrangements to pool responses from all three surveys and secured the assistance of Renuka Chander, the member research manager at the American Association for the Advancement of Science (publisher of Next Wave), and ReData of Bethesda, Maryland, to help us tabulate and interpret the data sets. These data analyses--and this summary presentation--are fully the responsibility of the Postdoc Network.
A close look at the results of these surveys illustrates several areas of progress--including those mentioned above--but it also highlights a number of areas in which further effort is still required.
The PDN survey was distributed by e-mail in May 2002 using the PDN listserv  and the Association of American Medical College's GREAT group  listserv. The PDN requested the following information:
(1) Postdoc compensation.
What is the high/median/low salary/stipend for first-year biomedical postdocs at your institution? Does this increase with years of experience? Are biomedical postdocs compensated at a different rate than other postdocs, e.g., chemistry or physics?
What benefits are available to all postdocs, regardless of funding source? What percentage is covered by the institution?
(2) Institutional support for postdoc associations and offices.
Does your institution support a postdoc office or association?
What is the annual operating budget? From where is it derived?
We received 34 responses. The respondents in all but two cases were administrators or faculty from the respective institutions. The two postdoc respondents were chairs of well-respected postdoc associations.
Thirty-four institutions responded to the follow-up survey. According to Deborah Stine, associate director of COSEPUP, their goal was to compare results with the earlier survey and measure whether any change had occurred. However, due to inherent limitations of sample size and second-round nonresponders, this goal was not achievable. As a result, COSEPUP did not pursue publication of their findings.
The FASEB survey. FASEB is concerned about inequities in postdoc benefits within and among institutions. They are assessing the administrative, financial, and philosophical issues, on the part of both training institutions and funding agencies, which underlie the inequities of benefits access. ?The FASEB Science Policy Committee carried out an informal survey of  institutions, including a few government institutions, to determine provision of health benefits to postdocs, as well as access to retirement programs,? explained Heather Rieff, a senior policy analyst at FASEB. Like COSEPUP, FASEB decided not to publish their results. The FASEB Science Committee has also been working to address the need for mentoring and career planning advice during the postdoctoral period. In early September they will be releasing an "Individual Development Plan," a tool to be used by postdocs and mentors.
In all, 63 institutions participated in these three surveys (see sidebar). Some participated in two or all three, but multiple responses were merged in considering their responses. Twenty-six of the 63 institutions were public academic institutions, 25 were private academic, 10 were government or nonprofit, and two were industry-based organizations. These proportions roughly match the distribution of postdocs as given in the 1997 NSF Survey of Earned Doctorates. This survey reported that 80% of postdocs were employed in academia, 12% in government, and 8% in industry.
The following institutions participated in at least one of the surveys ...
Arizona State University
Of these institutions, 56 provided health benefits to all postdocs, 23 had postdoc offices, and 25 had an active postdoc association. Other findings included:
(1) The median salary for first-year biomedical postdocs is $30,000 to $39,000 at both public and private institutions.
(2) Health insurance is widely available to postdocs at public and private institutions, but private institutions are more likely to offer dental and life insurance.
(3) At private institutions, those that have a postdoc office or postdoc association (PDA) are more likely to offer a benefits package (Fig. 1).
(4) Institutions that have a postdoc office or association are more likely to have formal annual raise policies (Fig. 2).
(5) Health benefits and access to institutional facilities and child care, tuition, and housing subsidies are more likely at institutions with postdoc associations.
(6) The presence of a postdoc office is positively correlated with higher median salaries for postdocs.
(7) More private than public academic institutions have postdoc offices (Fig. 3).
Some trends are quite clear in these data. For example, it appears that if you want to pursue a postdoc, your best bet--from the compensation angle, anyway--is to look for a private institution with a postdoc office or association. The data do not allow us to determine whether or not this effect is causative, but the presence of an association or office dedicated to postdoc affairs appears to be a marker for good management in this area.
Other questions cannot be dealt with so easily. The combined surveys generated only limited data on compensation for nonbiomedical postdocs, although those that did come in indicate that biomedical postdocs are compensated at a lower rate than chemistry and physics postdocs. With only two industrial labs participating, the sample size was too small to make any general conclusions regarding postdocs in industry.
We had hoped to compare compensation and benefits between postdoc associates, fellows, and National Research Service Award (NRSA) recipients. However, not only were these data difficult to obtain, but that which was provided showed substantial variation between institutions. In general, NRSA recipients tend to receive fewer benefits than fellows, and fellows fewer than associates. The data were too complicated to offer a specific breakdown of benefits, even for a single class of postdoc.
Finally, most respondents did not provide data on budget and personnel for postdoc offices and associations. That which we did obtain suggests a pastiche of funding schemes, from a fee per postdoc (assigned to PIs) to funds derived from overhead costs. Some institutions did not provide the equivalent of one full-time employee devoted to postdoc affairs, while others had four staffers plus a dean.
Implications and Directions
Since the initial publication of the PDN database of postdoc organizations  in 2000, many offices and associations have been added--but several have also been removed. Postdocs report that administrative support is critical. Since the founding of the postdoc office at her institute, a postdoc reports, "It's not even the same institution as far as postdocs are concerned. The administration has already made several key changes that were considered impossible a year ago."
Survey data may provide the necessary push to implement change, especially when coupled with strong leadership and a sense of urgency. This has been the case for some postdoc associations, who have found surveys to be a useful advocacy tool . As we found in the PDN meta-analysis of postdoc association surveys , a coordinated effort, with a set of predetermined and tested questions, is likely to provide greater clarity. Actors in the postdoc policy arena need to establish partnerships before conducting surveys and policy reviews. Nevertheless, the data presented here are worth paying attention to. They illustrate that there has been progress on postdoc issues over the last couple of years. And that a postdoc office or association appears to be an important component in that progress.
It has been an interesting summer.