2027 Data Science Degree Underemployment Report: Which Graduates Are Most Likely to Work Below Their Education Level
Co-Founder and Chief Data Scientist
Many data science graduates are entering a market where demand is strong but entry-level screening is unforgiving. The 2024 Strada Institute and Burning Glass Institute report found that 52% of bachelor's graduates are underemployed one year after graduation, meaning their jobs do not typically require a degree.
For data science students, the risk is not the major alone; it is the gap between coursework, proof of applied skill, and employer expectations. This guide explains who is most vulnerable, which roles preserve career momentum, and how to avoid getting stuck in low-credential work.
Key Things to Know About Underemployment in Data Science Industry
- Data science graduates face lower risk than many nontechnical majors when they can show job-ready skills, but there is no official national underemployment rate for data science majors specifically; the best benchmark is the broader 52% one-year underemployment rate for bachelor's graduates reported by Strada Institute and Burning Glass Institute in 2024.
- The labor market still rewards degree-level data skills: the U.S. Bureau of Labor Statistics reported a May 2024 median annual wage of $112,590 for data scientists, far above common low-credential fallback roles such as customer service, administrative support, or retail sales.
- The highest-risk graduates are usually those with weak portfolios, no internship or co-op, limited SQL or cloud experience, shallow machine learning projects, and resumes that describe coursework instead of measurable business or technical outcomes.
- How likely is it for Data Science graduates to become underemployed?
- Is the college curricula for Data Science keeping up with employer expectations?
How likely is it for Data Science graduates to become underemployed?
Data science graduates are less exposed to underemployment than graduates from many broad, nontechnical fields, but they are not insulated from it. Underemployment occurs when a graduate works in a job that typically does not require a bachelor's degree, such as general clerical work, retail sales, call center support, basic data entry, or nontechnical operations roles.
The clearest national benchmark is the 2024 Strada Institute and Burning Glass Institute finding that 52% of bachelor's graduates were underemployed one year after finishing college. Data science graduates with strong applied skills should not assume that figure applies directly to them, but they should treat it as a warning: a degree alone is no longer enough to prove readiness for a technical role.
Risk varies sharply by preparation. A graduate who has completed internships, written production-style SQL, built deployable models, and explained business impact is competing for analyst, data scientist, machine learning, and business intelligence roles. A graduate with only classroom notebooks and no evidence of applied work may be screened into general office, reporting, or support positions.
The table below shows how underemployment risk usually differs across common data science graduate profiles. These are practical risk categories, not guaranteed outcomes, because employer demand varies by region, industry, and hiring cycle:
| Graduate profile | Typical underemployment risk | Why the risk changes | Better target roles |
| Internship plus portfolio plus SQL and Python | Lower | Employers can verify applied technical ability before the interview | Data analyst, junior data scientist, BI analyst, analytics engineer associate |
| Strong coursework but no internship | Moderate | The degree signals potential, but the resume may lack workplace evidence | Data analyst, reporting analyst, research analyst, operations analyst |
| General data science degree with weak projects | Higher | Course titles do not prove the ability to solve messy business problems | Entry-level analyst roles with structured training, technical support analyst |
| Late job search after graduation | Higher | Graduates may accept low-credential work to create immediate income | Contract analyst work, apprenticeship-style analytics roles, internships for recent graduates |
For students deciding whether the degree is worth it, the main takeaway is that data science still has strong labor-market value, but the payoff depends heavily on early career execution. Students who wait until graduation to build proof of skill are more likely to compete for jobs that use only a small part of their education.
Is the college curricula for Data Science keeping up with employer expectations?
Many data science programs have improved by adding machine learning, statistics, programming, ethics, and visualization, but employer expectations often move faster than academic catalogs. The most common mismatch is not that colleges teach the wrong theory; it is that graduates may not get enough practice with messy data, stakeholder questions, cloud tools, version control, model deployment, and measurable business outcomes.
The BLS projects data scientist employment to grow 36% from 2023 to 2033, which confirms strong long-term demand in the U.S. market. However, fast growth does not mean every graduate is ready for degree-level work on day one. Employers increasingly expect junior candidates to be useful with data pipelines, dashboards, experimentation, and communication, not only algorithms.
The table below compares common curriculum strengths with employer expectations. Use it as an audit tool when reviewing a program, choosing electives, or deciding what to add outside the classroom:
| Area | Often covered in programs | What employers commonly expect | Underemployment risk if missing |
| Statistics and probability | Hypothesis testing, regression, distributions | Ability to choose methods, explain uncertainty, and avoid misleading conclusions | Moderate |
| Programming | Python or R assignments | Readable code, Git, testing habits, reusable scripts, APIs, and collaboration | High |
| Databases | Introductory SQL | Joins, window functions, data modeling, query optimization, and warehouse concepts | High |
| Machine learning | Model training and evaluation | Feature engineering, bias checks, deployment awareness, monitoring, and business trade-offs | Moderate to high |
| Visualization | Charts and dashboards | Executive-ready storytelling, KPI design, and decision support | Moderate |
| Cloud and MLOps | Sometimes optional | Familiarity with AWS, Azure, Google Cloud, Docker, pipelines, or orchestration tools | High for technical data science roles |
A good program does not need to teach every tool, but it should create enough applied work for students to demonstrate transferable ability. If a curriculum is mostly theory, students should compensate with internships, capstones, open-source contributions, Kaggle-style work that goes beyond leaderboard chasing, or employer-sponsored projects.
Students considering graduate business or analytics leadership paths should also compare technical depth with management training. For example, an executive MBA may support experienced professionals moving into analytics strategy, but it is usually not the fastest fix for a new graduate who lacks SQL, Python, cloud, or portfolio evidence.
Data science sits in a stronger position than many majors because it connects directly to high-demand technical, analytical, and business functions. Still, national underemployment datasets often group data science with computer science, statistics, mathematics, analytics, or business fields, so readers should be cautious about any source claiming a precise standalone underemployment rate for data science bachelor's graduates.
One useful comparison is major-to-occupation alignment. Data science has clearer degree-level pathways than broad humanities or general studies majors, but it can be more competitive at the entry level than nursing, accounting, or engineering fields that have highly standardized hiring pipelines, licensure, or accreditation expectations.
The table below summarizes how data science compares with other broad major categories from an underemployment-risk perspective. It is designed to help students understand relative risk rather than predict an individual outcome:
| Major category | Typical degree-level pathway clarity | Common underemployment pressure | How data science compares |
| Data science, statistics, analytics | Moderate to high | Portfolio and experience expectations can be high for entry-level roles | Strong upside, but skills proof matters heavily |
| Computer science | High | Competition for junior software and AI-adjacent roles | Similar technical screening pressure; stronger if the graduate can code well |
| Engineering | High | Location, internship, and industry-cycle effects | Often more standardized than data science hiring |
| Business administration | Moderate | Broad degree may lead to sales, support, or operations fallback roles | Data science may offer stronger technical differentiation |
| Humanities and general social sciences | Lower to moderate | Degree-level role fit may require additional specialization | Data science generally has clearer technical job alignment |
The practical conclusion is that data science is not a weak major, but it is a proof-driven major. Graduates who can demonstrate applied work are better positioned than many peers. Graduates who rely on the credential alone may face the same underemployment pressures affecting the broader bachelor's labor market.
Yes, it can. Low-credential work may slow career growth when it prevents graduates from building the technical record employers expect for data roles. The risk is not simply a lower starting wage; it is the loss of time spent outside analytical workflows, technical teams, code review, data infrastructure, stakeholder presentations, and model evaluation.
AI has made this issue more important. Many entry-level tasks that once helped juniors learn, such as simple data cleaning, basic chart creation, and first-pass reporting, are increasingly supported by automation. That does not eliminate junior data jobs, but it raises the bar: employers may prefer candidates who can validate outputs, ask better questions, interpret results, and connect analysis to decisions.
Low-credential roles are not equally harmful. The following comparison can help graduates decide whether a fallback job is a bridge or a trap:
| Fallback role type | Career-growth effect | Signals to watch | Best use of the role |
| Data-adjacent operations role | Can be a bridge | Access to spreadsheets, databases, dashboards, KPIs, or process data | Convert routine reporting into measurable analytics achievements |
| Technical support for software or analytics tools | Can be a bridge | Exposure to product data, SQL logs, customer behavior, or implementation teams | Move toward product analytics, solutions engineering, or data analyst roles |
| General administrative role | Mixed | Limited analytical ownership or no technical stack | Create internal reporting improvements and apply externally quickly |
| Retail or customer service with no analytics path | High risk | No data tools, no internal mobility, no relevant metrics ownership | Use only as short-term income while upskilling and applying |
Graduates who plan to pursue advanced credentials should also be realistic about timing. A specialized master's degree may help if the barrier is advanced quantitative depth, but a fast doctoral option such as 1 year PhD programs online no dissertation is generally not a substitute for the hands-on technical experience employers expect in entry-level data science hiring.
The salary gap can be large because degree-level data roles sit in a much higher wage band than common fallback jobs. According to BLS May 2024 wage data, the median annual wage for data scientists was $112,590. That figure should not be read as a guaranteed starting salary for new graduates, but it shows the economic value of roles that fully use data science training.
By contrast, many low-credential fallback roles pay closer to general support or service wage levels. The table below compares degree-level data science pathways with fallback roles using BLS-style occupation categories and typical role alignment. Salaries vary by state, employer, industry, and experience, so use these figures as labor-market context rather than a promise:
| Role category | Degree utilization | Typical work | Salary context |
| Data scientist | High | Modeling, experimentation, predictive analytics, data interpretation | BLS May 2024 median annual wage: $112,590 |
| Operations research analyst | High | Optimization, forecasting, decision modeling, resource planning | Often degree-level and quantitatively aligned |
| Business intelligence analyst | Moderate to high | Dashboards, KPIs, SQL reporting, stakeholder analysis | Often a strong entry point into analytics careers |
| Customer service representative | Low | Customer issue handling, scripts, account support | Usually does not require a data science degree |
| Retail sales worker | Low | Sales floor support, transactions, customer assistance | Usually weak degree alignment unless it leads to corporate analytics |
| General office clerk | Low | Administrative support, filing, routine office tasks | Usually weak degree alignment unless it includes reporting ownership |
The financial risk becomes sharper for graduates with student loans. A lower starting wage can make repayment harder, delay savings, and reduce flexibility to relocate, complete certifications, or accept a short-term internship that would improve long-term prospects. The goal is not always to reject every lower-paying job; it is to avoid roles that provide neither adequate income nor a credible path to degree-level work.
Underemployment usually comes from a combination of market timing, weak experience signals, and unclear specialization. Data science is broad, so employers may struggle to interpret a resume unless the graduate clearly shows whether they are suited for analytics, machine learning, data engineering, business intelligence, research, or product work.
The most common barriers are practical and fixable, but they require action before graduation. These issues often decide whether a student gets interviews for degree-level roles or gets pushed toward general support work:
- No internship or co-op: Employers often use experience as a shortcut for readiness, especially when entry-level applicant pools are crowded.
- Portfolio projects that look like class assignments: Projects based on clean datasets, generic notebooks, and no business question rarely stand out.
- Weak SQL: Many entry-level analytics jobs require strong SQL before advanced machine learning matters.
- No cloud, Git, or deployment exposure: Employers may doubt whether the graduate can work in a modern technical environment.
- Unclear career target: A resume trying to be a data scientist, software engineer, analyst, and researcher at once can look unfocused.
- Local market mismatch: Some regions have fewer degree-level analytics openings, making remote search, relocation, or industry targeting more important.
- Late applications: Waiting until after graduation can mean missing internship conversions, campus recruiting timelines, and alumni referral windows.
Students should also avoid choosing programs only by name recognition. A strong program should provide employer-connected projects, career services familiar with analytics recruiting, access to internships, and enough electives to match the student's target role.
A data science resume should prove fit for a specific role, not simply list tools. Hiring managers and applicant tracking systems look for evidence that the candidate can solve business or research problems with data. The strongest resumes connect technical methods to measurable outcomes.
Before rewriting the resume, graduates should choose one primary target: data analyst, junior data scientist, BI analyst, analytics engineer, machine learning engineer associate, product analyst, or research analyst. Each target requires a slightly different mix of keywords, projects, and accomplishments:
- Open with a role-specific summary that names the target function, such as analytics, machine learning, BI, or experimentation.
- Move technical skills near the top and group them by category, such as programming, databases, visualization, cloud, statistics, and machine learning.
- Rewrite projects as outcomes: include the question answered, data used, method applied, tool stack, and decision supported.
- Replace vague phrases such as "worked with data" with specific actions such as "built SQL queries," "designed dashboard KPIs," or "validated model performance."
- Include GitHub, portfolio, or dashboard links only if the work is clean, documented, and relevant to the target job.
- Tailor each application to the job description, especially SQL, Python, Tableau, Power BI, Excel, cloud tools, experimentation, and stakeholder communication terms.
- Show domain context when possible, such as healthcare, finance, retail, logistics, education, cybersecurity, or marketing analytics.
A common mistake is using one resume for every posting. Data scientist, data analyst, and analytics engineer roles overlap, but they are not identical. A resume that looks too broad may be passed over in favor of one that matches the employer's immediate need.
Graduates who want to combine analytics with management may eventually compare MBA options, including the best AACSB online MBA programs. For immediate underemployment prevention, however, the first priority is usually a sharper technical resume and stronger evidence of applied data work.
Certifications can help, especially when they validate tools that employers actually use. They are most useful when paired with projects, internships, or job experience. A certificate alone rarely overcomes a weak portfolio, but it can reduce doubt about practical skills in SQL, cloud, BI, or machine learning workflows.
The table below summarizes certifications and credentials that can support degree-level data roles. Costs, exam formats, and vendor requirements can change, so students should verify details before enrolling:
| Credential area | Examples | Best-fit roles | When it helps most |
| Cloud data and analytics | AWS, Microsoft Azure, Google Cloud data credentials | Data analyst, analytics engineer, junior data engineer, ML associate | When job postings repeatedly mention cloud platforms |
| Business intelligence | Microsoft Power BI, Tableau credentials | BI analyst, reporting analyst, business analyst | When the graduate has strong analysis skills but needs dashboard proof |
| SQL and database credentials | Vendor or platform-based database certifications | Data analyst, analytics engineer, database-focused analyst | When SQL is the main interview barrier |
| Machine learning certificates | Applied ML or cloud ML credentials | Junior data scientist, ML analyst, applied research assistant | When supported by original projects and strong statistics |
| Project or agile credentials | Entry-level project management or scrum credentials | Analytics coordinator, product analyst, operations analyst | When targeting cross-functional analytics teams |
Students should choose certifications based on job postings, not social media trends. If 25 target postings mention SQL, Power BI, and stakeholder reporting, a dashboard credential may produce a faster return than an advanced machine learning certificate.
Use this sequence to decide whether a certification is worth the time and cost:
- Collect current U.S. job postings for the exact target role and region.
- Count repeated tools, platforms, and credentials across those postings.
- Choose one certification that matches a frequent requirement and fills a real gap.
- Build a project using the certified tool before applying.
- Add the certification to the resume only with evidence of applied use.
The best strategy is to treat employability as a four-year project, not a senior-year emergency. Data science students should build a record that answers the employer's core question: can this candidate use data to make a decision, improve a process, or build a reliable technical solution?
The following steps are most effective when started early, but recent graduates can also use them to move out of low-credential roles:
- Pick a target pathway by sophomore or junior year: Choose analytics, BI, machine learning, data engineering, product analytics, or research so coursework and projects reinforce each other.
- Master SQL before chasing advanced tools: SQL appears across many entry-level analytics roles and is often more immediately employable than niche modeling techniques.
- Build three portfolio projects with different purposes: Include one SQL analysis, one dashboard or business case, and one statistical or machine learning project with clear limitations.
- Secure internship experience before senior year: Internships, co-ops, research labs, and employer-sponsored capstones carry more hiring weight than isolated class projects.
- Network through alumni and local employers: Referrals matter because junior data postings can attract large applicant pools.
- Apply before graduation: Track applications, interviews, referrals, and skill gaps instead of waiting for the diploma to start the search.
- Evaluate fallback jobs carefully: Prefer roles with data systems, metrics, reporting, or internal transfer potential over jobs with no analytical exposure.
- Keep salary expectations realistic but strategic: A lower first salary may be acceptable if the role builds technical evidence; a low wage with no data work is a higher-risk trade-off.
Program selection also matters for nontraditional learners. Students returning to school later in life may need flexible scheduling, transfer-credit review, and career support that understands analytics hiring; resources on online degree programs for seniors can be useful when comparing flexible education formats.
The most employable data science graduates are not always the ones with the most advanced math or the longest tool list. They are the ones who can show credible evidence of applied problem-solving, communicate trade-offs, and enter the market before desperation pushes them into low-credential work.
Some graduates use low-credential roles briefly while they build income, finish a portfolio, or relocate. The danger is that the first job can become sticky. The 2024 Strada Institute and Burning Glass Institute report found that 45% of graduates who were underemployed one year after graduation remained underemployed 10 years later, which means early job quality can shape long-term career mobility.
For data science graduates, not all fallback roles carry the same risk. A job as a reporting assistant, operations analyst, marketing analyst, QA analyst, research assistant, or technical support analyst can still create transferable evidence if the graduate uses data tools and documents measurable results. A job with no analytical tasks, no technical systems, and no path to internal transfer is much more dangerous.
Graduates who accept a low-credential role should create an exit plan before the job becomes their default career path. The plan should be specific, time-bound, and tied to evidence that hiring managers can evaluate:
- Define the degree-level target role, such as data analyst, BI analyst, analytics engineer associate, or junior data scientist.
- Identify the missing requirements by comparing 20 to 30 U.S. job postings for that role.
- Build one portfolio project that directly matches the most common missing skill, such as SQL analytics, dashboarding, experimentation, forecasting, or model deployment.
- Use the current job to create data-adjacent accomplishments whenever possible, even if the title is not technical.
- Set a job-search checkpoint within 90 to 180 days instead of waiting for the role to become comfortable.
The key is to avoid treating any job as automatically career-building. A low-credential role helps only if it creates evidence, contacts, domain knowledge, or internal mobility toward degree-level work.
Other Things You Should Know About Data Science
Yes, it can be worth it for students who build applied skills, internships, and a clear portfolio. The degree is riskier when students rely only on coursework and graduate without evidence that they can use SQL, programming, statistics, and business communication in realistic settings.
Good stepping stones include data analyst, BI analyst, reporting analyst, operations analyst, product analyst, research assistant, QA analyst, technical support analyst for data tools, and analytics coordinator. These roles are most useful when they involve databases, dashboards, metrics, experimentation, or process improvement.
Not always. A master's degree may help if the graduate lacks advanced quantitative depth or wants research-heavy roles, but it is often better to first fix practical gaps such as SQL, portfolio quality, internships, cloud exposure, and targeted applications. More education should solve a specific barrier, not simply delay the job search.
A reasonable checkpoint is 90 to 180 days. If the role is not creating data-related accomplishments, internal transfer options, or stronger interview evidence, the graduate should intensify applications, networking, project work, and certification planning rather than waiting indefinitely.
See What Experts Have To Say About Studying Data Science
Read our interview with Data Science experts
- Talent Disrupted | Strada Education Foundation https://www.strada.org/reports/talent-disrupted
- Quantifying the UK Data Skills Gap - Full report https://www.gov.uk/government/publications/quantifying-the-uk-data-skills-gap/quantifying-the-uk-data-skills-gap-full-report
- Half of graduates end up underemployed - what does that mean for colleges? https://www.highereddive.com/news/half-of-graduates-end-up-underemployed-what-does-that-mean-for-colleges/710836/
- The Labor Market for Recent College Graduates https://www.newyorkfed.org/research/college-labor-market
- Educational Attainment by Race and Ethnicity - Race and Ethnicity in Higher Education https://www.equityinhighered.org/indicators/u-s-population-trends-and-educational-attainment/educational-attainment-by-race-and-ethnicity/
- Room for Progress in College Graduates’ Transition to the Labor Market - Public Policy Institute of California https://www.ppic.org/blog/room-for-progress-in-college-graduates-transition-to-the-labor-market/
