This post was co-authored with Lauren Kane, COO of BioOne, who also contributed the analysis.

It may not come as a surprise that the working title of this post was “Mind the Gap: The Third Rail.” Over the past two years, we have been discussing issues of gender inequity in scholarly publishing — in person, at various conferences and events (for example at the 2016 Society for Scholarly Publishing [SSP] conference), and via the written word, in Learned Publishing, Digital Science’s popular blog, and here on The Scholarly Kitchen. Whether in comments sections or session Q&As, these discussions inevitably turned to the issue of compensation.

There is no shying away from the fact that this is a loaded topic. It is also part of a much broader debate that transcends both gender and our own industry. In the United States, women, on average and across all industries, earn about 80 cents on the dollar to their male counterparts (though the precise figure varies from study to study). But is this also the case in scholarly publishing? There are no substantial, current data on this question of which we are aware, so we decided to start small, analyzing publicly available data for nonprofit scholarly publishing organizations. We know that this does not represent the full picture; rather, our intent is to kick-off the conversation and spur on a broader analysis.

Methodology
As many of you know, US nonprofit organizations are required to publicly disclose their annual tax returns (form 990s*), essentially justifying their exemption from income tax. These documents are a treasure trove of information, including the compensation of what the Internal Revenue Service (IRS) calls “Key Employees and Highest Compensated Individuals” (found in Schedule J of each report) — individuals receiving compensation in excess of $115,000 and/or those in the top 20% of all organizations’ earners.

For the purposes of this post — which is intended to provide a snapshot of this issue, rather than a full representation — we analyzed all organizational members of two major industry bodies, the SSP and the Association of Learned and Professional Society Publishers (ALPSP), for which a publicly available form 990 was available. In most cases, the 990s were for fiscal year 2014, the most current posted return for the majority of organizations. This yielded a sample of 46 US-based nonprofit organizations, mostly scholarly societies, as well as a few not-for-profit presses, aggregators, and service organizations. By virtue of their SSP/ALPSP memberships, we can assume that publishing is an important activity for these organizations, though for some (e.g. the American Association of Cancer Research) it is only a small part of their mission.

Among the 46 organizations surveyed, the 990s collected yielded an aggregate 321 individuals listed as key/highly compensated employees. Part-time employees who were indicated as such, or anyone with a pro-rated salary (for instance, employees that started or finished mid-year) were omitted from the analysis. To ensure parity, total organizational compensation (Column E) was recorded, which includes base compensation, bonuses, retirement, and nontaxable benefits received from the organization.

The average total (gross) organizational revenue of this sample was $59,874,671; the median was $16,321,137. Publications revenue (where reported) represented, on average, 55% of this income — with an average of $34,265,466 and median of $11,209,355. The publications revenue range was very broad, from $200K (the Association of Pathology Chairs) to nearly $500M (the American Chemical Society).

Analysis
Of the 321 individuals in the sample, 40% were women (130) and 60% were men (191). The average compensation was $314,586. At $287,291, women earned 86% of their male counterparts’ $333,164 average — slightly better than the US average. Like total and publications revenue, the range of individual compensation was quite broad — from $115K (the reportable minimum) to over $2M. Given this, median compensation is probably a more accurate indicator; women fared better on that basis, earning 93% of their male counterparts’ compensation — a median of $244,613 versus $264,140 for men.

While these aggregate results are compelling on their own, there is more to learn from analyzing compensation by job title and function, both on average and in relation to organization size. This is particularly true when looking at the very top — among CEOs / executive directors.

Of the 46 organizations surveyed, 19 had a female chief executive (41%), most of whom led proportionately smaller organizations. Of the 16 organizations earning in excess of $50M in total revenue per year, women led only five, while at the other end of the spectrum, women led six of the 10 organizations with total revenue less than $5M. In our Learned Publishing study of scholarly publishing leadership demographics, we demonstrated that, despite their estimated three-fifths industry majority (58%), women occupied just one-third of CEO positions for the 186 organizations surveyed — including both nonprofit and for-profit, as well as companies outside the US. While keeping in mind the smaller sample size, it is encouraging that proportionately more of the organizations surveyed here had a female chief executive. Still, we have a way to go before our leadership reflects the overall gender make-up of our industry.

Turning to average compensation, there is a more significant gender gap among CEOs than in our larger sample. The average CEO earned $453,670. At $388,732, female CEOs earned just 78% of their male counterparts’ average compensation of $499,368. Unsurprisingly given the relatively small sample size, there are some interesting outliers, including a gender gap in female CEOs’ favor among the 13 organizations earning total revenue from $25M – $100M. Results by organizational revenue are shown in the chart below.

When looking beyond organizations’ chief executives, variances among titles and very small sample sizes made analysis difficult or inadvisable. Instead, we associated individuals based on their listed title with one of eight “departments.” The results are summarized in the table below. With the exception of those working in education/policy, where women’s average compensation was 4% higher than their male counterparts, average compensation for men exceeded that for women in every department. In marketing/communications, where our sample included 15 women and 4 men, women earned, on average, 22% less than their male counterparts.

In terms of the number of male versus female key employees in each department, it probably doesn’t come as a surprise that the widest disparities came in information technology, where men outnumbered women in the sample 29 to 6, and in finance, where men outnumbered women in the sample 21 to 9—or that the reverse was true in marketing/communications, where women outnumbered men 15 to 4.

Thoughts and Next Steps

It’s clear that a larger sample — one that includes for-profit entities and organizations based outside the US — would be more representative of our industry as a whole. And a more comprehensive survey (generated from active organizational responses, rather than a limited IRS form) could also provide valuable demographic information beyond gender, allowing for additional analysis based on race/ethnicity, age, education, location, and more. Such information would answer a number of questions that will undoubtedly arise in the comments (and that we share!), such as whether and to what extent differences in individuals’ level of education, years of experience, and location impact total compensation. An even larger study might also explore company policy in relation to flex-time, child care, and required travel — and how this impacts advancement for both women and men.

So what next? In our view, more data is required. While there are industry compensation surveys available, such as GuideStar’s nonprofit compensation report, they are costly, and often lump together disparate organizations (e.g., different kinds of publishing and media organizations). One solution would be for industry organizations such as SSP or ALPSP to individually or collectively organize and execute a demographic salary survey for the scholarly publishing industry. The (anonymized) results, openly and freely shared, would provide a tremendous professional development resource for the benefit of all.

One final point: industry and organizational benchmarking is a very useful and persuasive tool when applying for a new job or advocating for a salary increase at your current organization. We strongly encourage everyone (women and men) to remember that they should be paid based on their position, their performance, and their experience, and not on any other factor. This includes salary history, which it is now illegal for hiring employers in Massachusetts to require, with similar legislation being reviewed in other states.

*Form 990s, for US-based nonprofit organizations, can be accessed via a free personal account at www.guidestar.org.