Which skills are worth learning?✳

Under 16? You can’t legally work: any employer who hires you faces a fine of $5,000 per month, and those willing to do so are breaking the law

Most kids finish middle school at 15 or 16, right at this age threshold. Hiring anyone under 16 is illegal; each month an employer uses a child worker, they must pay $5,000 in fines. That means any employer willing to hire you is knowingly breaking the law. If anything goes wrong, all medical bills and compensation must come from that employer — if they refuse to pay or go out of business, there’s no one else to cover those costs.

Adding “education’s value” to mortality statistics: each extra year of schooling lowers adult death risk by roughly 1.9%

When data from 603 studies conducted in 59 countries are pooled together, every additional year of schooling is linked to an average 1.9% drop in adult death risk. This effect is strongest for people aged 18–49, where each extra year cuts risk by about 2.9%. For those over 70, the drop is just 0.8%. In other words, the difference between finishing high school and only finishing junior high goes far beyond simple wage gaps.

Before determining whether “education is losing its value,” first look at the national education structure: only 15,467 people per 100,000 have a college education or higher.

According to the 2020 census, only 15,467 people per 100,000 nationwide have completed junior college or higher education — that’s roughly 15%. Meanwhile, 34,507 people per 100,000 have only a junior high school education, making it the most common level. On average, people aged 15 and older have attended school for just 9.91 years, barely enough to finish junior high. Therefore, a junior high education in China is far from “sufficient”; it is the baseline level for most people.

Run the policy check first for “can’t afford it”: most vocational school tuition is already waived, with a grant of 2,300 yuan and student loans up to 20,000 yuan per year

At public vocational schools, tuition is waived for rural students, urban students in agriculture-related majors, and those from low‑income families. Starting spring 2025, all vocational and regular high school students receive a yearly grant of 2,300 yuan. For university students, the grant rises to 3,700 yuan per year, and they can borrow up to 20,000 yuan annually through student loans. If funds are still short, repayment can be deferred or an alternative path chosen — this rarely forces a student to drop out and work immediately.

Failing to get into regular high school doesn’t mean your path ends: vocational schools offer integrated enrollment programs and separate exams, and employers can lower educational requirements for skilled roles.

The 2022 Vocational Education Law states that vocational education holds equal status to general education. Vocational high schools can enroll students through integrated programs with higher vocational institutions, which in turn must reserve enrollment quotas or hold separate exams specifically for vocational high school graduates. Employers cannot set barriers against vocational school graduates based on their educational background. Not making it into a regular high school means switching to an alternative path, not staying at a junior high school level.

Treating “college vs. working” as a calculation problem: the three years of early earnings versus decades of future income

Lay both paths out on paper and compare them. Working early yields only three years of wages from low‑skill jobs — that’s it. Continuing school, on the other hand, adds a steady yearly income boost for the next three or four decades. In Chinese cities, estimates show that each extra year of schooling raises average earnings by 4.0% (1988) up to 10.2% (2001). These figures are a bit dated, so they’re best used only to gauge the order of magnitude.

Keep this baseline in mind: on average, each additional year of formal education yields a roughly 9% annual increase in personal income worldwide.

A comprehensive review of 1,120 estimates from 139 countries shows that, on average, each extra year of schooling raises an individual’s income by about 9% per year. Women consistently see higher returns than men. This figure represents the average effect of “one extra year of formal education,” not the benefit of enrolling in a single course. Treat it as a yardstick: if a program costs 30,000 yuan and lasts half a year, ask whether its expected income boost equals that of one full year of schooling.

Before spending money on a certification, first verify whether it appears in the national list of occupational qualifications or in the roster of evaluation agencies registered with the Ministry of Human Resources and Social Security.

Certain organizations misuse the concept of “new occupations” or “skill levels” to run training programs, charge excessive fees, and issue dubious certificates. Some even impersonate authoritative bodies. The Ministry of Human Resources and Social Security has taken steps to curb such practices. Before enrolling, follow three simple checks: first, confirm whether the occupation is listed in the national list of occupational qualifications; second, verify that the issuing agency is registered with the ministry; third, ensure the certificate number can be found on the official verification platform. If any of these checks fail, the enrollment fee is essentially wasted.

It’s best to use government subsidies for training — don’t jump straight to paying for private courses.

From 2025 through late 2027, the government plans to provide subsidized training to over 30 million participants. Funding comes from multiple sources, including employment assistance funds, unemployment insurance funds, and corporate education budgets. The more in-demand a profession is and the higher its training costs, the greater the subsidy. Employees should ask their employers first; everyone else should consult their local Human Resources and Social Security Bureau’s subsidy list before deciding what to study.

When picking a skill to learn, prioritize those requiring hands-on work and on-the-spot judgment — these are the hardest for automation to replace.

A study evaluating 702 occupations in the U.S. found that roughly 47% of them fall into the high-risk category for automation. Three key factors currently hold machines back: manual dexterity, creative problem-solving ability, and social intelligence. That is why jobs that demand physical work, take place in unstructured environments, and require real-time decision-making are far less likely to be automated, while roles with rigid processes and heavy information processing carry the highest risk. Don’t fixate on that 47% figure — focus instead on these three core barriers.

Check first whether a skill has an official grading system and if those grades can translate into pay raises

The government is promoting a new eight-tier vocational skill grading system, which requires employers to link both job pay and skill levels to each other, with extra emphasis on hiring and promoting workers at the senior levels. Eligible employees can even have their skill levels recognized directly, regardless of their years of service or prior qualifications. However, this is still a policy direction rather than actual, guaranteed pay increases — implementation varies widely across regions and companies. The only reliable way to gauge this is to see if a skill has an official national standard and a formal grading pathway; only then can it realistically lead to pay raises based on skill level.

Don’t take “high wages reported online” as proof of labor shortages — check local lists of in‑demand occupations and subsidy programs instead.

The amount of subsidies each region offers is essentially the government’s way of “pricing” which trades are in short supply; those amounts are set based on urgency and training costs. Differences in subsidy levels for the same trade across cities directly reflect how badly that trade is needed locally. Figures you see online about monthly earnings for a given trade usually come from job boards or personal reports — they highlight only the top earners and ignore working hours, commuting, occupational hazards, and seasonal fluctuations.

For the same money and time, prioritize short-duration programs that lead directly to a job

Government policies encourage “quick-to-implement” training programs that lead to short-term employment, as well as micro-specializations offered by vocational colleges. When choosing a program, there’s only one key criterion: does it provide a specific job opening you can apply to after completion? Programs that require a year or more of study and then demand additional “experience building” after graduation actually cost far more than their listed price. It’s essential to set clear exit terms upfront: specify at what point of completion you can stop the program if you haven’t secured a job yet.

After studying, close the book and test yourself — don’t just reread it

Rereading a passage makes it feel very familiar, but that doesn’t mean you’ve actually retained it. Closing the book and testing yourself yields noticeably better results. A meta-analysis of 222 classroom studies involving nearly 48,500 students showed that adding brief self-tests raised grades by a moderate amount. In a separate review that ranked ten common study methods, self-testing earned the top rating. Approximately 50,000 students participated in these studies overall.

Spread the same study time across several days instead of cramming it all at once

Spending the same amount of time on studying across multiple days helps you retain information far better than cramming it all at once. This conclusion is supported by 317 separate experiments, all of which reached the same result. There is also a practical rule to keep in mind: the longer you want to remember something, the longer the interval between review sessions should be. If a test is tomorrow, review the material that day; if it’s a month away, schedule a review session a few days later.

Don’t treat highlighting, rereading, or summarizing as primary study methods

Students most often rely on highlighting and rereading, yet a review that evaluated ten different study methods ranked both of them at the very bottom. Summarizing, keyword mnemonics, and visualizing texts also ended up at the lowest tier. That doesn’t mean they’re useless; it simply means they only work under very specific conditions. Spending the same amount of time on self-testing would be far more beneficial.

Mixing different types of exercises — don’t do twenty of the same kind in a row

Doing many similar problems back‑to‑back makes them feel easy at the moment, but later you may struggle to recall which method to apply. Mixing several types of exercises yields better long‑term results — the effect size is moderate. This technique works best with visual material such as drawings and categorization tasks; it has a smaller impact on math problems, and for vocabulary memorization it’s actually better to practice in blocks.

Don’t pick learning methods based on “I’m a visual learner and he’s an auditory learner”

The idea that “everyone has their own learning style and teaching should be tailored to that style” is widely believed. To prove this claim, researchers would need to group students by learning style, randomly assign each group to different teaching methods, and then test them all with the same exam. Yet almost no studies have produced results that support this notion. People certainly have preferences, but there is no evidence that teaching to those preferences actually improves learning outcomes.

Test yourself using questions based on how you’ll use the material later; don’t just recite it verbatim

Testing yourself after closing the book offers benefits beyond simply memorizing a few words. A meta-analysis of 192 comparisons involving over 10,000 participants shows this method works reasonably well for different question types, application problems, and reasoning tasks. However, it’s largely ineffective in two cases: when the question–answer pairing is reversed, and when the content has only ever appeared in the book without ever being tested.

评职称先弄清自己在哪个系列、哪一级，再按单位性质找申报渠道

English translation pending · Chinese original

职称全国分 27 个系列，每个系列分初级、中级、高级。一般要一级一级往上报。在单位上班的，由单位审核、公示后往上报。在民营企业的，可以去职称申报受理服务点，或者找人才中介服务机构、行业协会申报。自由职业者报当地人社部门组织的评审。

会计这类初级、中级职称靠全国统考拿，先对照学历和工作年限去报名

English translation pending · Chinese original

有些系列的初级、中级职称是考出来的，考过就算，不再另外评审。会计就是这样：助理会计师要高中以上学历，会计师要本科干满 4 年会计、大专满 5 年。已经拿到某些职业资格的，也可能直接算作有相应职称。

别找中介代评、别买论文代写、别在材料上造假：查实就撤销职称，记诚信档案 3 年

English translation pending · Chinese original

用假材料、代写论文、剽窃成果拿到的职称，查实就撤销。失信记录要记 3 年，还进全国信用平台。申报时签的承诺书不实，3 年内不能再申报。人社部门还专门查处中介的虚假宣传、假冒评审和假证。

拿到职称不等于涨工资：先问清单位是按岗位比例评聘，还是评了不一定聘

English translation pending · Chinese original

职称是聘用、考核、晋升的依据，但评上和聘上是两回事。事业单位一般在岗位比例内评，评上多半能聘。不实行岗位管理的单位可以评聘分开，评上了单位不一定给对应的岗位和工资。换单位、换地方后，职称可能要重新评或者确认。