Many communities aspire to create a more inclusive economy characterized by equitable economic growth for a diverse workforce. But practical plans for achieving this goal are harder to come by.
Despite the best of intentions, data limitations often make it challenging to accurately assess, let alone improve, the diversity and equity of a regional economy. Traditional government sources of labor market information (e.g. Bureau of Labor Statistics) provide comprehensive coverage backed by robust statistical methodologies — but typically lack the precision to accurately capture real-world job titles, let alone pinpoint the skills needed to transition from one job to another.
Fortunately, real-time data sources, like employer job postings and professionals’ online profiles, provide exactly the kind of precision those traditional sources lack — but can have comparatively uneven coverage across industries and job types.
The good news is, these complementary data sources can be integrated to amplify their strengths, overcome their limitations, and provide a best-of-both worlds data set that serves as a roadmap to greater equity and economic mobility.
While “integrating complementary data sources” may sound a bit theoretical, the results couldn’t be more practical. In fact, the economists, data scientists, and researchers at Lightcast have already used this approach to empower and guide strategies for education providers, non-profit organizations, policy makers, and other stakeholders in economic and workforce development. In this article, we’ll show how communities can use this data to unlock diversity, equity, and upward mobility in three basic steps:
Get a more precise picture of current workforce diversity
Identify promising career pathways based on skill similarity
Create a better equity benchmark and deploy targeted reskilling programs
1) Get a more precise picture of current workforce diversity
As usual, an effective strategy begins with an honest, accurate assessment of current conditions. In this case, you’ll want to establish a baseline for the diversity (or lack thereof) of your region’s workforce, and how it varies across different sectors and jobs.
At a basic level, this information can be obtained from a source like the Bureau of Labor Statistics (BLS). But there’s a catch: BLS data is classified by SOC, O*NET, and NAICS codes which provide a useful-but-simplified way of understanding the labor market. After all, in the real world, it’s rare that an individual’s job title is identical to the SOC code classification the BLS would assign to them.
Professional profile data, on the other hand, is incredibly precise and lets us analyze the labor market in terms of the actual job titles individuals and businesses use. This data does have its own quirks, like its tendency to offer stronger coverage of “white collar” jobs where individuals are more likely to share job changes and skills via their online profile. But these are not the same quirks as BLS data. Which is why integrating the two sources is so effective: the official government statistics keep the real-time data anchored in reality while the real-time data enhances the granularity of the government sources.
In practical terms, integrating these two data sources means that we can go from having ethnicity and gender estimates for 400 – 500 occupations in a given metropolitan area to producing diversity figures for over 1,300 detailed occupations. These detailed occupations better reflect the actual job titles used by employers and workers, and therefore allow us to analyze the nuances between different specialties within broad career areas.
For example, whereas BLS data can tell us what percentage of Medical Records Specialists are women of color, it can’t tell us how many of those women work as medical coders vs medical record technicians vs clinical systems analysts. Detailed occupation data, enhanced by online profiles, can.
This level of precision isn’t just “nice to have” or for merely informational purposes. These differences matter, both in terms of the skills required to do the job and the wages individuals can command in the economy. Consequently, this data is essential for spotting skill similarity and potential pathways between roles. And, as we’ll see in the next two sections, this skill-level insight is critical to developing actionable strategies that drive real change.
2) Identify promising career pathways based on skill similarity
In this step, we’ll pivot from diagnosing the challenge to exploring solutions. Specifically, we’ll seek to identify promising career pathways that are both feasible and desirable. For the moment, we’ll focus on the pathways themselves before re-applying the diversity, equity, and inclusion lens in the third step.
To assess feasibility, we can consider things like how much overlap there is in the required knowledge, skills, and abilities for a given “origin occupation” and a related “destination occupation.” We can also look for similar education or work experience requirements, along with whether or not there is sizable and growing demand for the destination occupation (after all, if there are no job openings for the destination occupation, then transitioning into it hardly represents a realistic career move, even if there is significant skill overlap).
We then introduce another set of metrics to determine the desirability of the job transition. Wage differential is a key ingredient here. Does the destination occupation pay more or offer better advancement opportunities than the origin occupation? If so, it’s probably a desirable transition. But we can also use indicators like automation risk to create an even more holistic, forward-looking assessment of a destination occupation’s desirability.
As before, we’ll need to integrate both traditional LMI and real-time data in order to accurately identify and map these promising pathways. In this case, that means starting with the Department of Labor’s O*NET database and the list of knowledge, skills, and abilities (KSA’s) that it provides for each occupation. Just like we saw with the BLS’s diversity data, this provides a helpful, well-structured starting place for our research.
But, it also faces similar limitations: a relatively-slow-to-update taxonomy that doesn’t capture the nuance and detail needed to create a true-to-life, high-fidelity picture of how job requirements connect in the modern labor market. For that, O*NET data must be augmented with real-time skill data captured from actual employer job postings. This insight provides the up-to-date detail needed to pinpoint the skills similarity between roles with greater confidence and clarity.
Just as important, it also enables us to identify the bridge skills and last-mile skills needed to successfully transition from one role to another. The Equation for Equality, a recent Npower report focused on women of color in tech, provides helpful definitions for these terms (and a powerful example of how this data can help to advance equity in a particular sector of the economy):
Bridge Skills are the skills that constitute the gap between the skills that a worker has now and the skills that they will need in their target job.
Last Mile Skills are skills that are specific to a particular employer, team, or role and can be developed through onboarding.
From “The Equation for Equality: Women of Color in Tech”
With this insight, community leaders can create, fund, and support targeted training programs laser-focused on the specific skill gaps that keep workers from making the jump to desirable destination jobs. And, combined with the detailed diversity analysis from step one, these programs can be targeted to reach underserved minority populations who are overrepresented in lower-wage jobs. Which brings us to step three.
3) Create a better equity benchmark and deploy targeted reskilling programs
Now comes the exciting part where we put it all together. From step one, we have detailed, job-level data outlining workforce diversity in our region. We know which occupations are disproportionately filled by minority or disadvantaged populations. And we know which high-growth, high-wage jobs are noticeably lacking in diverse representation.
From step two, we can identify realistic career pathways based on the similarity of skills, experience, and education required for various jobs. We’ve also evaluated the desirability of those transitions based on factors like whether or not the destination job offers a higher wage and has an acceptably low automation risk.
Putting these insights together, we can create a better benchmark for workforce diversity and craft an actionable plan for beating that benchmark in the future.
Establishing the benchmark
For our benchmark, we can now construct an “Equation for Equality Index,” like the one used in the aforementioned Npower report. This involves comparing the diversity of the current workforce in a given sector of the economy (say, IT or allied health or media production, etc.) with the diversity of the broader “skills-similar workforce,” comprised of individuals who work outside the target sector but who have enough skills overlap to be considered a viable talent pool.
From “The Equation for Equality: Women of Color in Tech”
This index provides a meaningful benchmark that is equal parts realistic and idealistic.
Realistic, because it is rooted in hard data about the number of underrepresented individuals working in skills-similar jobs
Idealistic, because it shows the potential for greater diversity if those individuals can be reskilled for higher paying, in-sector jobs
Beating the benchmark
To improve the equation for equality index number, community leaders can implement programs that teach the specific bridge skills and last-mile skills needed to move individuals from their current role into a higher-growth, higher-wage job. Furthermore, these programs can be targeted towards individuals working in career areas with the highest percentage of diverse representation. In this way, communities can design programs to naturally oversample the most diverse sectors of the economy, thereby boosting equity and economic mobility at the same time.
Conclusion
There’s nothing new about communities pursuing economic growth. But the recent focus on ensuring this growth is inclusive and equitable has highlighted the need for better data — data that not only captures how things are, but illuminates the path to how they could be.
The team at Lightcast is committed to delivering this data through innovative, customized approaches that combine the best of traditional and real-time sources. If you’d like to explore how this data can help unlock inclusive growth in your region, please reach out. We’d love to learn more about your community and partner with you to help it thrive.
Learn more about Lightcast’s consulting work by checking out our highlight reel of major projects from 2021.
