By junior year, Li Jialu had stopped pretending his advertising major was enough. Since December, he’s spent most nights teaching himself to code, convinced it will matter more than his major ever could.

During a summer internship at a tech company in Shanghai, the gap was clear: students with technical backgrounds had more skills, more opportunities — and often twice the pay. “Most liberal arts internships pay around 120 to 180 yuan ($17 to $25) a day,” he tells Sixth Tone. “Some advertising firms offer as little as 100.”

Li’s coding skills quickly proved useful. In his marketing role, data analysis and programming helped him move faster and think with more structure. “These advantages from mathematics and computer science are something many liberal arts students lack,” he says.

He calls it the “tragic fact” of being a liberal arts student, a fear shared by classmates who began interning in their first year. That anxiety is only heightened by China’s most competitive job market on record: 12.22 million graduates in 2025, with liberal arts students trailing STEM peers in both hiring rates and pay.

A 2024 survey by job platform Zhaopin found that only 43.9% of liberal arts majors had secured offers by mid-April, nearly six points behind their peers in technical fields.

Policy is shifting, too: this year, China’s Ministry of Education added 29 majors to its undergraduate catalog — the majority STEM-focused — while top universities cut enrollment and eliminated programs like public administration, musicology, and broadcasting.

On Xiaohongshu, known in English as RedNote, the tag “switching to coding” has drawn more than 220 million views and 1.4 million discussions. Posts range from frustration over low liberal arts salaries to anxious questions like, “Is 30 too old to start coding?” Users trade free resources, while veteran programmers offer advice.

As the pressure pushes students beyond their majors, a booming market of online classes now offers coding, data analysis, and AI tools to anyone willing to log on at night, creating a new path for liberal arts students trying to catch up.

Universities and policymakers are reinforcing the shift, adding STEM-heavy programs and trimming humanities enrollment to better match what they call “national development needs.”

The other track

For all his late-night classes on coding, Li still isn’t sure he’s learning the right things. Online tutorials are endless, he says, but it’s hard to measure progress.

“Self-study lacks feedback,” he says. “It’s hard to know if you’re improving or just going in circles.”

It’s why some students, like Chang He, a doctoral student in Beijing, have taken a more drastic route.

After finishing his journalism degree in 2021, Chang decided a few bylines weren’t enough to build a career. He enrolled in a second bachelor’s in data science, moved on to a master’s, and is now pursuing a Ph.D. in computer science and technology.

He believes it is the only realistic way for liberal arts graduates to break into tech. “Even a master’s in Arabic might not open many doors,” he says. “But a second degree in data science gave me a way into an entirely new industry.”

Chang admits his foundation is weaker, sometimes leaving him stuck on code his classmates breeze through. But, he says, “Graduate studies are about specialization, not total mastery. Different strengths are normal.” AI has also eased the shift. When he hits unfamiliar code, he feeds it into a model to break down and analyze, a shortcut that makes the transition manageable.

Employers see the imbalance just as clearly. “Most openings are in technical and R&D roles,” says Sun Jie, an HR manager at a major tech firm. “Support functions like administration have far fewer spots. And with graduates vastly outnumbering roles, it’s a serious supply-demand mismatch.”

For liberal arts graduates, she argues, technical literacy can make the difference. “Many roles aren’t strictly ‘arts’ or ‘STEM.’ Companies want versatile talent who can drive growth,” she says, pointing to product managers who combine data analysis with user insight as a prime example.

Another veteran HR director, Guo Changyuan, puts it more bluntly: At her tech firm, engineering degrees come first. When those aren’t available, she says, a university’s reputation often matters more than the student’s major.

“For technical roles, STEM graduates usually have a stronger foundation,” she says. “But in the end, we look at logical thinking, communication, and overall competency. Salary isn’t about whether you studied arts or STEM, it’s about what level you qualify for.”

Guo believes liberal arts graduates with STEM training are at a disadvantage when competing head-on with STEM majors for technical jobs. Instead, she says, they should target hybrid roles that combine technical literacy with broader skills, where their blended background can offer an edge.

That’s the path Li now has in mind: a career in product management. To prepare, he plans to study applied psychology alongside his advertising and coding skills. “Knowing how to create content, connect with users, and develop products gives me an edge that neither pure tech nor pure humanities can offer,” he says.

System update

As more liberal arts students turn to technical skills, universities are trying to catch up. Under a national push called the “New Liberal Arts” initiative, schools are integrating coding and data training into once purely humanistic fields.

The journalism school at Huazhong University of Science and Technology, in central China’s Hubei province, for instance, has added classes in Python and data mining for students with no programming background. Using visual tools and project-based learning, the courses promise to make non-technical majors more employable in the age of AI.

And at Shanghai’s Fudan University, associate professor and career advisor Cai Xiaoyue says dual-degree programs now let liberal arts students tack on STEM training without abandoning their original field.

She argues that while coding and data analysis strengthen job prospects, communication and teamwork skills remain just as essential. “In the age of AI and digital transformation, liberal arts education must evolve to not only understand AI but also embody and ultimately transcend it,” she says.

Lu Wenbo, a career advisor at Wuhan University of Science and Technology, in Hubei province, concurs. He argues that universities need to adapt faster to market signals, scaling back programs with poor job outcomes and aligning new ones with industry needs. Liberal arts fields evolve quickly, he asserts, making curriculum updates difficult, but he stresses their role remains essential, if less visible.

Lu also urges students to think beyond rigid academic tracks. “Undergraduate education focuses on foundational learning and well-rounded development,” he says. “Specialization comes later. Students should confidently explore beyond their major.”

At his university in Shanghai, Li says, liberal arts students now take Python in their first year and later move on to data mining and visualization. The goal is to bridge the gap, but he finds the courses too abstract. “The curriculum is disconnected from real-world industry needs,” he says. “There’s too much theory and too little practical application.”

That disconnect forces even seasoned professionals to backfill technical skills on their own, a sign of how quickly expectations have shifted.

Wen Xin, an HR manager with four years of experience, began teaching herself data analysis earlier this year to better handle payroll reports and turnover trends. She started with Excel but found it too limited and slow. On the advice of friends in tech, she switched to Python and SQL.

At first, the syntax and symbols overwhelmed her. So she flipped the approach: rather than follow a course, she let real problems lead the way. “Start with problems, not systematic learning,” she says. Step by step, she began automating the tasks that once ate up hours of her day.

Now, Wen is charting two possible paths. One leads to a full transition into tech, where she hopes to become a data analyst. The other keeps her in HR, but in a more specialized role focused on people analytics.

In either case, she says, her background in the humanities is an asset. “I’m more attuned to human factors and communication,” Wen says. “When I design an HR system, I want it to feel human. That’s where liberal arts graduates tend to be more sensitive.”

As a recruiter, Wen believes empathy is often undervalued — but increasingly essential. “STEM graduates might pick up data faster,” she says, “but empathy is what builds company culture. What organizations truly need are problem-solving abilities, not a specific academic major.”

Additional reporting: Feng Jingyi; editor: Apurva.

(Header image: John Lund/Getty Creative/VCG)