Receive daily articles & headlines each day in your inbox with your free ERE Daily Subscription.

Not logged in. [log in or register]

Metrics for STEM Women – a Critical Examination of the High-tech Approach

by and Aug 4, 2014, 5:21 am ET

STEMThere has been a great deal of publicity lately surrounding the lack of STEM women at high-tech firms. Unfortunately, we have to give two thumbs down to the diversity data from each of the top high-tech firms that have publicly released their numbers. Although the firms’ intentions were good, the limited scope of the metrics that they revealed do not provide the necessary information that STEM women need in order select which firm to join or the right information needed in order to encourage them to actually apply for a different tech job.

High-tech firms have two basic reasons for attempting to hire and retain more STEM women into key roles.

The first and most obvious is to meet EEOC legal requirements, but the second is related to improving business results. This business reason is critical because if you don’t have a sufficient number of women designing a product and providing customer service, it’s unlikely that the products and services that the firm offers will fit the needs of current and potential women users. This business reason means that simply hiring women is not enough, because it is also equally important to make sure that STEM women are placed in influential positions that directly impact a firm’s products, services, and the way that it does business. The current metrics that firms use are not clearly aligned with their business goals, and because they are only “quota metrics,” we find that they do not provide the in-depth information that is needed in order to identify the barriers that prevent STEM women from applying, accepting an offer, and succeeding in a new job.

It’s time for firms to move away from metrics that cover only standard quotas (where you only measure the percentage of women). Now is the time to introduce metrics that go further and measure STEM-womens’ influence at the firm and whether they are treated well enough that they stay at the firm for long enough to have a business impact. These broader goals of influence, inclusion, and retention goals obviously require not just metrics covering the percentage of women at the firm but also a complete STEM women scorecard, which helps to reveal the depth, impact, and movement of STEM women in a company. Our recommended STEM women metrics are broken up into three parts: metrics for current employees, metrics for recruiting, and metrics for a business case.


Part I — STEM Women Metrics Covering the Current Employee Population

Before you can increase the representation of women at a firm, you must improve the recruiting and retention processes. Metrics can only help you improve your recruiting and retention if the firm tracks and reports on the key issues that STEM women worry about before they consider applying for a new job or when they decide to quit a job at your firm. As a result, the metrics that we recommend focus on key job issues that STEM women often have, including: whether women frequently become managers and if they are represented in key jobs, get promoted rapidly, get paid fairly, and if they are represented in executive positions. We recommend corporations select from among the eight STEM women metrics listed here.

  1. The number and percentage of STEM women employees — this metric provides only a limited value because it does not cover important impact factors like the level of the positions that women fill, their pay, their promotion rates etc.
  2. The percentage of women among C- level executives and on the firm’s board of directors – one way to ensure that women have a high strategic business impact and visibility to STEM women on the outside is to have STEM women adequately represented on the board and among “C-level” executives.
  3. The average percentage of STEM women employees in management and leadership positions – having STEM women in management and non-executive leadership jobs will increase their business and product impacts, but it will also serve as a signal to other women that they can be hired into and promoted into management. As a result, firms should track and report the percentage of women and men who hold management and team leadership roles, particularly in technical business functions.
  4. The average percentage of STEM women employees in each key job family, including product design, sales, and customer service  if one of the goals of your firm is to have women employees in roles where they can influence the products and services of your firm, you must measure their percentage in key design and product development jobs. If another goal is to have women interacting with current or potential customers that include a high percentage of women, then you must also measure their percentage of representation in customer contact jobs including customer service and sales job families.
  5. The average promotion rate of STEM women employees compared to men – women have traditionally been promoted at a lower rate than men, so measure their promotion rate compared to men in key job families and career paths.
  6. The average retention rate of STEM women employees compared to men — spending large amounts in order to recruit STEM women only has a high return if they stay on the job long enough to have an impact. Measure the voluntary turnover rate of STEM women in key jobs and compare it to the rate for men.
  7. The average compensation of STEM women employees compared to men in the same job – women have frequently been underpaid compared to men; measure the average difference in pay between men and women in the same job. This compensation metric can be “leveled” based on experience, tenure, and education level.
  8. The number and percentage of STEM women employees over the age of 40 – because women frequently are concerned about discrimination as they age, measure the percentage of women in key jobs who are over 40.


Part II — A Recruiting Metric Scorecard for STEM Women

The recommended metrics above cover current employees, but firms should have a separate scorecard which covers the recruiting of STEM women each year. This separate scorecard focusing on recruiting is important because other than through mergers and purchasing firms with a high percentage of women employees, the only way to increase the percentage of women at a firm is to recruit them. Four powerful and revealing recruiting metrics that we recommend that firms consider include:

  1. The total number and percentage of STEM women new hires during the last year — although it’s a broad indicator, each year a firm should measure the number of women hired and their percentage among all new hires (excluding interns).
  2. The number and the percentage of STEM women new hires in management and team leadership positions — women considering a new job expect women to hold a representative number of management jobs. Firms should measure the number and percentage of STEM women who are externally hired each year into management and team leader positions.
  3. The number and the percentage of STEM women new hires in design, sales, and customer service jobs — if you expect STEM women to impact your product and customer service, you must measure the number and the percentage of STEM women who are hired each year into product and customer impact jobs (i.e. design, sales, and customer service).
  4. A metric to assess “where” in the recruiting process a firm is losing STEM women applicants – even if the firm succeeds in getting enough qualified STEM women to apply for its jobs, our research indicates that a significant percentage will drop out or be rejected long before a final hiring decision is made. If a firm expects to successfully hire more STEM women applicants, it must develop a system to identify precisely at what stage or stages of the recruiting process (i.e. resume screening, phone screening, interviews, or the offer process) a high percentage of STEM women are prematurely rejected or lost. The firm should also survey STEM women applicants and new hires to determine what worked and what didn’t work for them during the recruiting and hiring process. Obviously, this data would be used for continuous process improvement and it would not be revealed to the public. 


Part III — Business-case-related Metrics for STEM Women

Our research indicates that the best way to convince hiring managers and coworkers to identify and select more women is to demonstrate the business impact of hiring and retaining more STEM women. When both hiring managers and team members realize the size of STEM womens’ economic contribution, they are more likely to eliminate barriers that restrict their hiring. And finally the CFO’s office should work with HR in order to quantify the total revenue impact in dollars of hiring and retaining more STEM women. Obviously we don’t recommend that you externally report these metrics. Two business case metrics that we recommend are:

  1. The average on-the-job performance of STEM women new hires compared to men (i.e. quality of hire) — simply counting the number of women hired does not reveal the increased performance that newly hired STEM women can provide. Measure the on-the-job performance of newly hired STEM women (i.e. quality of hire) and then compare it to the average on-the-job performance rating (i.e. performance appraisal rating or forced ranking score) of newly hired men in the same job family. Firms should also strive to show whether STEM women produce innovations and innovative ideas at a rate higher than that of men.
  2. The correlation between STEM women and business results – demonstrate the economic value of having a representative workforce. One of the ways to do this is to show a direct connection or correlation between increasing the percentage of STEM women in a team and a direct improvement in business results for that team. By showing a high correlation, you can increase everyone’s support for hiring more STEM women.


Final Thoughts

Although many have applauded the recent release of diversity metrics by several high-tech firms, we find that the metrics that they have released can only be classified as shallow. They are shallow because they don’t cover important areas like inclusion, promotion rates, and the level of the positions held by STEM women. In most cases, the released metrics are merely statistics that are designed to meet the limited EEOC reporting requirements.

If firms want to be transparent and reveal where specifically in their organization that they have succeeded in making sure that STEM women are fairly represented, it’s time for them to develop a more comprehensive metric scorecard. In this article we have provided a variety of metric options that more effectively reveal to potential candidates and the current women employees whether a firm has succeeded in each of the areas that reveal a comprehensive inclusion of STEM women. Finally, we find that the time to move beyond simple quotas for STEM women and other diverse groups has long ago passed. The time to act is now.

This article is provided for informational purposes only and is not intended to offer specific legal advice. You should consult your legal counsel regarding any threatened or pending litigation.

  • Jim D’Amico

    Great article. As a metrics geek, I thoroughly enjoyed it! Really well thought out metrics, and detailed explanations.

    The one point that even tech companies seem to overlook is that there is strong evidence that organizations whose own diversity mirrors the diversity of their customers, out perform those who do not mirror by 3%-9% annually (Dr.Cedric Herring UIC, American Socialogical Association 2009).

  • http://www.imprintpdx.com Aimee Levens

    Seriously great article – there is so much oversimplification in STEM and this calls out what needs to be called out. Too many tech companies, especially startups, show off that they have women in leadership yet more often than not they are in sales, marketing, and admin roles – not technical leadership roles like engineering.

    And then of course there’s the rampant issue of how women are treated in the tech workplace (great article here: http://www.bizjournals.com/portland/print-edition/2014/07/18/opinion-selena-deckelmann-on-portland-techs-gender.html?ana=sm_ptl_upc37&b=1405698248^15005561). Lots claim they want more women then do nothing to change their behaviors &/or culture to attract them.

  • Keith D. Halperin

    Here are some facts and figures:

    FOR IMMEDIATE RELEASE: MONDAY, SEPT. 9, 2013

    http://www.census.gov/newsroom/releases/archives/employment_occupations/cb13-162.html

    Census Bureau Reports Women’s Employment in Science, Tech, Engineering and Math Jobs Slowing as Their Share of Computer Employment Falls

    Growth in women’s share of science, technology, engineering and mathematics occupations — commonly referred to as STEM jobs — has slowed since the 1990s, according to a U.S. Census Bureau report released today. Women’s employment in STEM has slowed because their share in computer occupations declined to 27 percent in 2011 after reaching a high of 34 percent in 1990.Blacks and Hispanics also remain underrepresented in STEM jobs.

    These statistics come from two reports released today: Disparities in STEM Employment by Sex, Race, and Hispanic Origin and The Relationship Between Science and Engineering Education and Employment in STEM Occupations. STEM workers include those who work in computer and mathematical occupations, engineers, engineering technicians, life scientists, physical scientists, social scientists and science technicians. It also includes managers, teachers, practitioners, researchers and technicians. The reports are an example of the important education and occupation statistics that the American Community Survey produces annually, allowing businesses, communities and civic leaders to make informed decisions on workforce development.

    In 2011, there were 7.2 million STEM workers accounting for 6 percent of the U.S. workforce compared with 4 percent in 1970. Half of STEM workers were employed in computer occupations, followed by engineers (32 percent), life and physical scientists (12 percent), social scientists (4 percent), and mathematicians and statisticians (3 percent).

    While women make up nearly half of the workforce, they were 26 percent of the STEM workforce in 2011.

    “We have seen an increase in women employed in STEM occupations, but they are still underrepresented in engineering and computer occupations that make up more than 80 percent of STEM employment,” said Liana Christin Landivar, a sociologist in the Census Bureau’s Industry and Occupation Statistics Branch and the reports’ author.

    Since the 1970s, women’s representation in the STEM workforce has grown in all occupation groups with the largest increase in social sciences from 17 percent to 61 percent. In 2011, women made up almost half of mathematical workers at 47 percent, an increase from 15 percent in 1970, and 41 percent of life and physical scientists, up from 14 percent in 1970. Women’s share of computer employment reached a high in 1990 of 34 percent, up from 15 percent in 1970, but it declined to 27 percent in 2011. Among all STEM occupations, women were most underrepresented in engineering in 1970 and in 2011. About 3 percent of engineers were women in 1970, compared with 13 percent in 2011. The 1970 estimates for social scientists, life and physical scientists and computer workers are not statistically different from one another.

    Earnings by Sex

    Men with a bachelor’s degree in science or engineering and employed full-time, year-round in STEM occupations earned $91,000, compared with women who earned $75,100 on average. Women with a science or engineering bachelor’s degree who were employed full-time, year-round in STEM occupations earned $16,300 more per year than women who had a bachelor’s degree in science or engineering but were not employed in a STEM occupation.

    Blacks and Hispanics make strides but still underrepresented in STEM

    Blacks and Hispanics have been consistently underrepresented in STEM employment. In 2011, 6 percent of STEM workers were black, increasing from 2 percent in 1970. Although the Hispanic population’s share of the overall workforce has increased significantly, its share of STEM employment has not shown a similar rate of increase. Hispanics made up 3 percent of the overall workforce in 1970; by 2011, their share had climbed to 15 percent. The Hispanic share of the STEM workforce in 1970 was 2 percent and increased to 7 percent in 2011. The estimates for the share of STEM workers who are black and Hispanic are not statistically different.

    Asian and non-Hispanic white workers were employed in STEM occupations at higher rates. Non-Hispanic whites held 71 percent of STEM jobs, but made up 67 percent of the total workforce, whereas Asians held 15 percent of STEM jobs compared with 6 percent of total jobs.

    The unemployment rates among science and engineering graduates varied by race and Hispanic origin. The unemployment rate among black and American Indian and Alaska Native science and engineering graduates was 6.6 percent. Of all science and engineering graduates, 83.3 percent were employed, 3.9 percent were unemployed, and 12.8 percent were not in the labor force.

    Asians and non-Hispanic whites who were science and engineering graduates and were employed full-time, year-round in STEM occupations earned more than any other demographic group ($89,500 and $88,400 respectively). The earnings estimates for Asians and non-Hispanic whites are not statistically different.

    Science and engineering graduates employed in STEM earned more than science and engineering graduates who were not employed in STEM. For example, the median earnings for employed blacks with a science or engineering degree was $58,000 but increased by $17,000 to $75,000 when employed in STEM. Similarly, Hispanics with a science and engineering degree earned $18,300 more when employed in STEM, increasing from $59,000 to $77,300. These three estimates for blacks and Hispanics are not statistically different from one another.

    Three in Four Science and Engineering Graduates not Employed in STEM Occupations

    Three in four science and engineering graduates were not working in STEM occupations in 2011. Instead, they were working in fields such as non-STEM management, law, education and accounting, and STEM-related occupations, such as health care.

    Science graduates were less likely to be employed in STEM because science employment typically requires graduate training, and many graduates may be employed in STEM-related fields, such as health care. On the other hand, engineering, computer, math and statistics majors were most likely to be employed in STEM because graduate training is not required for many engineering and computer jobs.

    “The statistics show that women are less likely to major in engineering and computer sciences, which may reduce their STEM employment options unless they go on to graduate school,” Landivar said.

    Of all STEM workers, 42 percent had a bachelor’s degree, 21 percent had a master’s degree, 6 percent had a doctorate degree and 1 percent had a professional degree. Women in 2011 were more likely than men to have a bachelor’s degree, but less likely to graduate with a science or engineering major. Women were 53 percent of college graduates and 41 percent of science and engineering graduates. Of female science and engineering graduates, only 15 percent were employed in STEM. Male science and engineering graduates were employed in STEM occupations at about twice the rate of women at 31 percent.

    Note: The Standard Occupational Classification Policy Committee issued recommendations in April 2012 on how to define STEM occupations. These two Census Bureau reports reflect the new recommendations. Other prior analyses, including a report released by the U.S. Department of Commerce Economics and Statistics Administration in July 2011, have defined STEM workers in slightly different ways.

  • bboyairwreck

    Loved the article Trena! It’s clear that if companies want to help their gender diversity, then they need to thoroughly understand the problem. The better one understands a problem, the better one can come up with more effective solution. To get a better grasp of the issue, it would require taking more metrics, as you are stating. Your additional metrics proposal is definitely a great suggestion if companies want to tackle diversity in the right way. I hope recruiters and even CEO’s in the STEM industry take note of this.