Written by

The Labor Market Is a Black Box

Data-Driven Intelligence, Programmatic| Views: 1255

Education is important. Particularly in the labor market.

But sometimes education gets confounded with certification.

Higher education: sends a very strong signal in the labor market (although not always rightly so) [source: Dilbert]
Higher education: sends a very strong signal in the labor market (although not always rightly so) // [source: Dilbert]

No Idea How to Evaluate Low-Skilled Workers

Last week a Forbes article by Adam Ozimek has caught quite some attention.


The piece in a nutshell:

(a) In the Labor Market College Degrees Are Used as Heuristics – Employers look at attended schools, majors, awards, and GPAs to infer certain skills in job applicants.

(b) The Same Is Not True With High School Degrees – Researchers found that employers do not use secondary education certifications/degrees for ability signaling.

(c) The Labor Market for Low-Skilled Workers Is Opaque – often the lack of information results in a less effective labor market.

Employees can’t be put at best use, generate less output, and earn less than they could.

‘Random’ Is The Recruitment Process

Unlike with white-collar employees, applications for low-skilled positions are less standardized (e.g. no LinkedIn to fall back), provide less comparable data points, and require repetitive manual input from the applicants (example form).

In addition lots of employers who hire for low-skilled positions also have smaller recruitment budgets (i.e. not feasible to conduct personalized tests; etc.).

Labor market for low-skilled workers is opaque - just like a blackbox [source: wikicommons]
Labor market for low-skilled workers is opaque – just like a blackbox [source: wikicommons]

The result:

Job applicants for low-skilled positions cannot properly signal their skills/education/knowledge.

The hiring process becomes a black box with a ‘random’ outcome.

Consequences on a Big Scale

To see the overall impact on the U.S. labor market we can look at 2 figures:

#1: Education Is a Not Spread Evenly by Race

On aggregate Hispanic or Latino populations have the lowest educational attainment of the U.S. labor force.

This is particularly alarming (c.p.) as Hispanics/Latinos are the youngest population group and will stay longest in the U.S. labor force.

Also Black/African American and White demographics have significant parts of their population without attainment of a tertiary education degree.

Access to education is spread evenly [source: bls.org]
Access to education is not evenly spread [source: bls.gov]

This puts certain demographics more at risk to be discriminated by hiring practices in the low-skilled labor market.

#2: Educational Attainment and Its Impact on Unemployment Rates

Statistics show a link between the level of education and ease of getting a job.

Unemployment by educational attainment [source: bls.org]
Unemployment rates by educational attainment [source: bls.gov]

The correlation between education and unemployment rates can have different causes.

Unemployment 1: Skill Mismatch

  • I don’t have a X degree, therefore I don’t have the skills.
  • That’s why I don’t have a X skill to match the X skill demand in the market.

Labor markets change constantly due to technological improvements, comparative advantages (to outsource labor to other regions), policy changes, migration, etc.

For this kind of unemployment it is important to anticipate supply and demand of labor and set incentives and methods to upskill the affected labor force.

Unemployment 2: Information Mismatch

  • I don’t have X degree, therefore I don’t have a certificate signaling my X skill.
  • That’s why I don’t get recognized in the market for having a X skill to match the X skill demand.

As the quoted article indicates, some unemployment originates from an opaque labor market. Employers can’t source and assess large parts of the labor force and hence cause higher unemployment, higher underemployment, and a low labor force participation.

To solve this problem it would help to better understand which skills are critical for low-skilled positions and develop methods to infer these skills from alternative data points.


An alternative information layer for the low-skilled workforce has the potential to change the following:

  • Increase transparency in an opaque job market
  • Provide a tool for employers that significantly improves their recruitment
  • Reduce discrimination towards job applicants that might have the skills but lack the credentials

About Spencer Parra

Spencer Parra is the VP of Product Management for advertising and data products at Radancy. In that capacity, Spencer and his team of product managers, program managers, data scientists, and data analysts work to develop products in a data driven mindset. As the leader of Advertising products, he works to bring a holistic full funnel approach to Radancy’s advertising technology stack with Programmatic Jobs at its foundation. Through data products, he tells the story of media performance via Radancy’s Metrics Gateway and helps ensure data is democratized through Radancy’s unified platform. Spencer came to Radancy from the Perengo acquisition in mid 2019 where he served as Lead Product Manager and a member of the founding team. With Perengo, he worked towards the vision of leveraging the same rigor and concepts from ecommerce advertising technology to the recruitment advertising space. Prior to Perengo, Spencer launched and supported in-app advertising products at Criteo as a solutions engineer. Spencer holds a B.S. in Aerospace Engineering from the Massachusetts Institute of Technology.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to top