Are these observations reflected in actual enrollment figures, graduation rates, and career data? Or do they stem more from headlines and anecdotal experiences than from statistically significant trends? This article aims to examine the validity of these observations using publicly available data on computer science enrollment and career choices.
Between 2000 and 2020, computer science departments at universities in the United States saw significant swings in enrollment:
Before we look at the latest data, it’s important to note that the overall trend over the past two decades has been one of growth. According to the U.S. National Center for Education Statistics (NCES), the number of undergraduate computer science degrees conferred in the United States grew from around 40,000 in 2010 to over 80,000 in 2020 11. This doubling reflected the immense demand for software engineers across industries, from finance to healthcare to e-commerce.
Starting around 2020, large-language model (LLM) technologies—exemplified by OpenAI’s GPT series—began demonstrating an ability to generate code snippets, debug code, and even build rudimentary applications. By late 2022 and into 2023, public-facing AI coding tools (e.g., GitHub Copilot, ChatGPT’s “Code Interpreter” capabilities, etc.) became more widespread. These developments fueled speculation that the coding job market might tighten, with AI taking over routine tasks and reducing the demand for human programmers.
While experts disagree on how profoundly AI coding will impact long-term job prospects, some students appear to be interpreting the technology’s progress as a threat to the viability of a career in software engineering. Anecdotally, high school career counselors report that students ask if “computers will be programming themselves” by the time they graduate college. Additionally, social media platforms host lively discussions among college students about whether they should switch from CS to data science, design, or business-oriented fields such as product management.
But do these anecdotal trends show up in actual enrollment numbers?
To analyze enrollment trends, we can look at preliminary data from multiple sources:
Below is a summary of available data on undergraduate CS enrollments from 2019 to 2023. While 2024 numbers are not yet fully compiled, preliminary insights from some institutions suggest similar patterns continuing.
Year | Total CS Undergrad Enrollment (Approx.) | Year-over-Year Change | Source |
---|---|---|---|
2019 | 650,000 | - | NSCRC |
2020 | 680,000 | +4.6% | NSCRC |
2021 | 720,000 | +5.9% | CRA Taulbee Survey |
2022 | 760,000 | +5.6% | CRA Taulbee Survey |
2023* | 775,000 – 790,000 (est.) | +2% to +4% (est.) | Preliminary data from select universities 22 |
*Data for 2023 is estimated, based on early reports.
Key Observations:
According to the National Student Clearinghouse Research Center 22, overall undergraduate enrollment in U.S. higher education declined by about 1–2% annually over the last few years, largely due to the COVID-19 pandemic’s effects. However, CS seems to have weathered these headwinds relatively well, continuing to post at least modest gains. The data suggests that while the pandemic and other factors have impacted higher education broadly, computer science remains a sought-after major.
A distinction must be made between enrollment (the number of students choosing CS as their declared major) and retention (the rate at which students remain in the major until graduation). Retention rates can illuminate whether students are switching out of CS after facing challenging coursework or concerns about future job prospects.
According to the Computing Research Association (CRA) Taulbee Survey 33, the average retention rate for CS majors from freshman to senior year in 2022 was approximately 65%—only slightly lower than the 68% reported in 2019. This dip might be influenced by a variety of factors, including pandemic-related disruptions and the highly competitive nature of introductory CS courses at many universities.
However, these retention rates do not indicate a mass exodus specifically attributed to AI coding fears. In fact, exit surveys from some universities (e.g., the University of Illinois Urbana-Champaign’s CS department) indicate that common reasons for switching out include the rigor of math prerequisites and personal preference changes rather than fear of AI automating coding roles.
When students do switch out of CS, they often move to adjacent fields:
Again, these switches are longstanding patterns; they predate the sudden popularity of AI coding tools. Surveys from 2022–2023 have not demonstrated an unusual spike in switches specifically due to AI.
Product management has become one of the fastest-growing fields in the tech sector. Product managers (PMs) bridge the gap between technical teams, designers, and business stakeholders. They are responsible for product strategy, roadmap planning, and ensuring the final product aligns with market needs. The role often commands competitive salaries and allows for broader decision-making responsibilities compared to entry-level engineering positions.
A common claim is that new CS graduates, rather than taking software engineering jobs, now lean toward product management. Data from LinkedIn’s Workforce Insights 44 indicates that the number of entry-level product management roles has grown notably since 2018. However, it remains a relatively small fraction compared to the volume of entry-level software engineering roles.
Below is a simplified comparison of entry-level job postings for two popular roles—Software Engineer and Product Manager—on LinkedIn in the U.S. from 2019 to 2023.
Year | Entry-Level SW Engineer Postings (Approx.) | Entry-Level PM Postings (Approx.) | Source |
---|---|---|---|
2019 | 40,000 | 3,500 | LinkedIn Workforce Insights |
2020 | 38,000 | 4,200 | LinkedIn Workforce Insights |
2021 | 42,000 | 5,500 | LinkedIn Workforce Insights |
2022 | 45,000 | 7,000 | LinkedIn Workforce Insights |
2023 | 47,000 | 8,500 | LinkedIn Workforce Insights |
Although product management roles are increasing, they still lag behind software engineering in overall numbers. Moreover, product management positions often prefer or require candidates with experience in engineering or another specialized area, meaning that fresh CS graduates without significant internship or co-op experiences may find it more challenging to break directly into a PM role. In other words, PM roles are growing, but they’re not supplanting software engineering as the primary career track for CS graduates.
Could AI coding tools nudge some CS students toward PM roles? It’s plausible that some students, witnessing how AI can automate parts of the coding workflow, might sense better job security in areas requiring managerial, interpersonal, and strategic skills—traits that are harder for AI to replicate. Yet, direct evidence of a widespread migration from software engineering to product management specifically “because of AI” is limited at this time.
The U.S. Bureau of Labor Statistics (BLS) projects a 25% growth in employment of software developers, quality assurance analysts, and testers from 2021 to 2031, much faster than the average for all occupations 55. These projections incorporate the assumption that automation, including AI, will eliminate certain tasks but also create new roles or expand demand in other areas (e.g., AI supervision, specialized software engineering for advanced systems).
Thus, the net effect on CS job openings is not clearly negative. Companies are likely to require a mix of AI-savvy developers, data scientists, and product managers to realize the full potential of these tools.
Now, let’s return to the initial observation:
“With the recent advancements in generative AI capabilities, specifically in AI being able to code and develop software, high school graduates are hesitating to join computer science courses for college, college students are dropping out or switching majors away from computer science, and college graduates are preferring product management careers over software engineering careers.”
If the data does not strongly confirm the observation, why does the perception exist? Several factors might contribute:
While it is understandable that the rising capabilities of generative AI might cause some individuals to question the long-term stability of software engineering careers, publicly available enrollment and career data do not currently support the claim of a mass decline in computer science interest or an exodus from software engineering to product management.
If anything, the data suggests a healthy, evolving job market where AI tools may automate certain coding tasks but simultaneously create or expand roles requiring deeper skill sets in advanced software engineering, AI model design, data science, and product management. Students who embrace both foundational computing knowledge and agile adaptation to new AI workflows may find themselves well-positioned, rather than replaced, by the next wave of technological disruption.
Ultimately, while the perception of a decline in CS interest or a shift to product management may hold true in isolated pockets and in popular discourse, the broader national trend does not (as of yet) confirm a wholesale retreat from computer science or traditional software engineering career paths. The future likely lies in a synergy between human developers and AI coding tools, rather than a zero-sum replacement scenario.
National Center for Education Statistics (NCES):
- Undergraduate Degree Fields
- Historical data on computer science degrees conferred.
National Student Clearinghouse Research Center:
- Enrollment Data
- Preliminary reports on 2022 and 2023 undergraduate enrollment trends.
Computing Research Association (CRA) Taulbee Survey:
- Taulbee Survey
- Data on CS enrollments, retention, and degrees awarded in North America.
LinkedIn Workforce Insights:
- Job Posting Analytics
- Aggregated data on job postings for software engineering and product management roles.
U.S. Bureau of Labor Statistics (BLS):
- Occupational Outlook Handbook
- Projections for software developer and related technology roles for 2021–2031.
Disclaimer: The data presented above is drawn from publicly available reports and surveys as cited. Variations at specific institutions or regions may differ, and ongoing updates (especially post-2023) may offer new insights. Readers are encouraged to consult the latest data from official sources or university reports to get the most current picture.
To learn more about how tech can influence your career, business or education subscribe to Real Cyber Report Newsletter.