GCSE提升一档 终身收入差多少 英国DfE研究
英国教育部（DfE）在 2021 年 6 月发布了一份研究报告，标题是 GCSE attainment and lifetime earnings。它用行政数据把 GCSE 成绩和一个人一生的收入直接连了起来，是目前少数能按学科逐档估算回报的官方研究。报告的核心结论是，GCSE 总成绩提高一档，折现到 16 岁的终身收入平均增加 8,500 英镑；换成不折现的口径，一档大约值 23,000 英镑。学科之间的差别很大，数学一档折现后约 14,500 英镑，英语约 7,300 英镑，音乐约 5,500 英镑。这些数字常被家长当成「多考一档就多赚一笔」的证据，报告本身却写得很克制，正文里专门列了三条限制条件。这篇文章把原始口径、学科差异和限制条件拆开讲，最后说清楚一档之差在卷面上通常差在哪里。
The Department for Education published a research report in June 2021 called GCSE attainment and lifetime earnings, drawing on administrative records that link exam results to tax data. It is one of the few official studies that estimates returns grade by grade and subject by subject; the headline result is that a one-grade improvement across a pupil's GCSEs is associated with £8,500 more in the present value of lifetime earnings, or about £23,000 without discounting. Maths carries the largest per-grade return at £14,500, followed by English at £7,300 and Music at £5,500. The authors are careful about what these numbers can support, and the report lists three caveats that news coverage often drops; this article keeps them in.
一、数据来源：LEO 把成绩和税务记录连起来
报告使用的数据来自英国教育部的 LEO 数据集，全称是 Longitudinal Educational Outcomes。它把学生的学校记录、考试成绩和税务收入记录连在一起，覆盖英格兰公立学校系统的全体学生。研究者盯住的是 2001 年到 2004 年参加 GCSE 的那一批人，因为到分析时他们的收入记录已经跟到三十多岁。收入要跟到退休才谈得上终身，行政数据看不到那么远，所以研究者用英国劳动力调查的数据把每个人的收入轨迹补齐到退休年龄。这种「成绩单加税单」的数据在英国是最近十几年才建起来的。它带来的最大变化是样本量足够大，可以按学科和档位切开看，而不是只统计五门 A* 到 C 这类门槛指标。报告的 PDF 全文和补充数据表都可以在 gov.uk 官网的研究出版页面免费下载。
The analysis uses the Longitudinal Educational Outcomes dataset, which links school records, exam results and tax records for every pupil in the state school system in England. The sample concentrates on pupils who sat their GCSEs between 2001 and 2004, because their earnings can be observed into their thirties. Because administrative data stop there, the authors use the UK Labour Force Survey to project each individual's earnings path up to retirement age; the full report and its supplementary tables are published openly on gov.uk. The real gain from this dataset is sample size, which lets the researchers split returns by subject and by grade instead of reporting a single threshold measure such as five A* to C passes.
二、终身收入的基数：130 万英镑与 51.5 万现值
报告先算了一个总量口径。2001 年到 2004 年参加 GCSE 的平均学生，一生不折现的总收入约为 130 万英镑。按英国财政部的折现规则折回 16 岁，这个数字变成 51.5 万英镑。折现的道理不难理解，四十岁赚到的 100 英镑和十六岁赚到的 100 英镑不是同一笔钱，越晚到手的钱折回今天越轻。折现率用的是财政部绿皮书的常规口径，前三十年 3.5%，之后 3%。后面所有的「一档值多少钱」都是折现后的口径，这一点不统一就会让数字互相打架。把两个总量放在一起看，才能理解报告为什么反复强调现值这个词。
Start with the baseline. For the average pupil who sat GCSEs between 2001 and 2004, undiscounted lifetime earnings come to roughly £1.3 million, a total that falls to about £515,000 when expressed as a present value at age 16. Discounting reflects the fact that money arriving at forty is worth less today than money arriving at sixteen; the calculations follow the Treasury Green Book, with a 3.5% rate for the first thirty years and 3% afterwards. Every per-grade figure quoted later sits on the same present-value basis, so the two totals and the grade returns should be read together. The gap between the two totals is the price of time, not evidence that a different data source was used.
三、一档成绩的价格：23,000 英镑与 8,500 英镑
整体看，GCSE 总成绩提高一档，折现后的终身收入平均增加 8,500 英镑。同一套模型换成不折现的口径，一档的价钱是 23,000 英镑，大约相当于 2019 年英国全职员工平均年薪的四分之三。报告还给了分性别的数字，女生一档约 20,000 英镑，男生约 24,000 英镑。如果提升幅度不是一档而是一个标准差，也就是 11.2 档，折现后的增量约为 96,000 英镑。这个量级接近平均折现终身收入的五分之一，也占到折现终身收入标准差的一半上下。数字听着很大，比较对象却要说清楚，报告比的是成绩相差一档、其他条件大体可比的两组学生，不是同一个人的两条平行人生。所以它更像一个政策评估用的平均价格，而不是某个学生未来工资的预测值。
A one-grade improvement across a pupil's GCSEs is associated with £8,500 more in present-value lifetime earnings. Without discounting, the same grade is worth about £23,000, which the report notes is roughly three quarters of average full-time annual pay in the UK in 2019; broken down by sex, the figure is around £20,000 for female pupils and £24,000 for male pupils. A larger improvement of one standard deviation, equivalent to 11.2 grades, comes with an increase of about £96,000 in present value, close to a fifth of average discounted lifetime earnings. These are group averages built for policy appraisal, not forecasts of what any individual will earn.
四、学科之间：数学 14,500 与英语 7,300
学科差异是这份报告最有信息量的部分。数学的一档折现后约 14,500 英镑，英语约 7,300 英镑，音乐约 5,500 英镑，德语更低一些。读到这里容易得出一个错误结论，以为某些科目天生更值钱。报告专门提醒，不同学科的学生不是随机分组，选科和成绩、家庭背景、学校资源都纠缠在一起。三门科学（Triple Sciences）的每档边际回报偏低，很可能是因为选这些科目的学生本来就在成绩与预期收入的高端，边际收益递减。换句话说，学科之间的数字可以比较，却不能当成选科的排序表。对已经定了科目的学生来说，这份数据的用法是在自己的科目组合里找回报最高的那几档。
Subject-level variation is where the report earns its keep. A one-grade improvement is worth about £14,500 in Maths, £7,300 in English, £5,500 in Music, and less than that in German. It is tempting to read those gaps as a ranking of subjects, but the authors warn against it, because pupils who choose different subjects are not random samples; subject choice runs together with prior attainment, family background and school resources. Triple Science shows a relatively low return per grade, most likely because the pupils taking it already sit near the top of the earnings distribution, where marginal gains are smaller. The numbers are useful for comparing grades inside a subject, not for ranking subjects against each other.
五、回报集中在哪几档：D 到 C 与 C 到 B
边际回报不是一条平线，它随档位起伏。多数科目里，从 D 到 C 和从 C 到 B 这两段的回报最大。C 以下的档位回报通常更小，有些科目在统计上与零没有区别。顶尖档位的情况分科目，A 与 A* 在部分科目里回报稳健，在另一些科目里测不出显著差别。男生在高档位上的回报普遍比女生大。这条规律解释了为什么英格兰的教育政策一直围着及格线做文章，16 岁那一年的分档位置关系到后面的升学路径。对成绩处在中间段的学生，把 D 提到 C、把 C 提到 B 的性价比，通常高于把已经很强的科目再往上推一档。
Marginal returns are not flat across the grade range. In most subjects the largest gains come from moving from D to C and from C to B, while grades below C bring smaller returns that are sometimes statistically indistinguishable from zero; returns at the very top are robust in some subjects and not measurable in others. Male pupils tend to gain more than female pupils at the top grades. This pattern is one reason English policy keeps returning to the pass threshold at 16, since a pupil's position at that point shapes the routes open afterwards. For a pupil sitting in the middle of the range, lifting a D to a C or a C to a B is usually worth more than pushing an already strong subject one grade higher.
六、9 到 1 分制下这些档位在哪：4 分标准合格与 5 分强合格
报告用的是旧制的分档语言，今天英格兰的 GCSE 已经是 9 到 1 分制。第一批改用 9 到 1 分制的科目是英语语言、英语文学和数学，时间是 2017 年，其余科目在随后几年陆续换完。官方口径里，4 分是标准合格，大致对齐旧制的 C；5 分在英语和数学上被教育部认定为强合格。英语或数学不到 4 分的学生，16 岁之后按现行资助条件仍然必须继续修读，这条要求早于新的分制，至今有效。需要说明的是，官方只公布了少数参照点，4 分的下限对齐 C 的下限，7 分的下限对齐 A 的下限，中间各档没有逐档的官方对应关系。分数线每年由考试局结合数据与考官判断重新划定，目的不是放水或压分，而是让不同年份拿到同一档的难度大体一致，这套机制由 Ofqual 监管。把报告的结论落到今天的分数上，比较稳的说法是把 D 到 C 这一段理解成及格线附近的那一档提升。
The report speaks in the older letter grades, while GCSEs in England have been graded 9 to 1 since 2017. English language, English literature and maths were the first subjects awarded on the new scale, with the rest following over the next few years. Grade 4 is the standard pass and lines up broadly with the old grade C, while grade 5 is recognised by the Department for Education as a strong pass in English and maths; pupils who finish key stage 4 below grade 4 in these two subjects must carry on studying them post-16 under the current condition of funding. Official documentation publishes only a few anchor points, which is why the middle of the scale has no one-to-one mapping; boundaries are reset each year by the exam boards using statistics and examiner judgement, under Ofqual's regulation. Transposing the report's findings onto today's numbers is safest when the D to C step is read as the gain around the pass line.
七、性别与家庭背景：男性边际回报高 18%
报告在分组比较里发现了两个稳定的差距。男性的边际回报平均比女性高 18%，非免费校餐学生的回报比免费校餐学生高 9%。女性的终身收入整体更低，成绩提升带来的绝对回报也更小。作者用令人沮丧来形容这个结果，同时提醒读者报告没有估算相对回报，如果按女性的平均收入做分母，差距可能没有绝对数看起来那么大。免费校餐一组的结论同样只停留在绝对数上。这两个差距说明一档成绩的经济价值不是一个固定常数，它和性别、家庭背景绑在一起。做个人决策时，把它们当成整体趋势就好，不必据此判断某个学生值不值得投入。
Two stable gaps show up in the subgroup analysis. Male pupils enjoy marginal returns about 18% larger than female pupils, and pupils not eligible for free school meals see returns about 9% larger than those who are eligible; women also have lower lifetime earnings overall, which pulls their absolute returns down. The authors call this a discouraging finding while noting that they did not estimate proportional returns, so the gap in absolute terms may overstate the gap in relative terms. The same caution applies to the free school meal comparison. What these splits show is that the cash value of one grade is not a constant; it moves with sex and family background. For an individual decision, they are worth reading as group trends rather than as a verdict on any particular pupil.
八、从 GCSE 到学位：一档成绩会被放大
GCSE 不是终点，成绩会顺着升学和就业路径一路放大。英国财政研究所（IFS）受教育部委托，在 2026 年 6 月发布了新一轮学位回报估算。研究者追踪 2002 年参加 GCSE 的队列，上过大学的人终身收入比条件相近但没上大学的人高约 40%，按今天的价格接近 32 万英镑。扣掉背景特征与先前成绩的差异后，调整后的净回报约为 18 万英镑，其中大约 43% 通过税收和贷款偿还回流到财政，个人平均多拿约 10 万英镑。报告也提到分布很宽，约四分之一的毕业生终身财务状况可能比没上大学更差；医学与经济专业的回报最高，平均毕业生终身税后收入比没上大学的高出 40 万英镑以上。一组补充结论更有意思，先前成绩偏低但上了大学的人，终身税后收入仍比同档没上大学的人高约 53,000 英镑。这些估算发布在 ifs.org.uk，与 GCSE 那份研究用的是同一类行政记录，可以相互参照。
GCSEs are not the end of the line, since results feed through into further study and work. The Institute for Fiscal Studies, commissioned by the Department for Education, published fresh estimates of degree returns in June 2026; tracking the cohort that sat GCSEs in 2002, it finds that those who went to university earn around 40% more over their lifetimes than similar peers for whom university was a realistic option, close to £320,000 at current prices. After adjusting for background and prior attainment, the net return is about £180,000, of which roughly 43% flows back to the exchequer through tax and loan repayments, leaving the average graduate around £100,000 better off. The distribution is wide, with about a quarter of graduates expected to end up financially worse off than if they had not gone, while medicine and economics top the table at more than £400,000. One supplementary finding stands out: graduates with weaker GCSEs still earn about £53,000 more in lifetime take-home pay than similar peers who skipped university.
九、三条限制：这份估算不能当因果读
报告自己在正文里列了三条限制，读数字之前应该先看这三条。一是数据无法测量所有同时影响成绩与收入的变量，如果存在未观测的特质，估算就会带上偏差。二是无法区分人力资本与信号，也就是分不清收入增长来自真的学到技能，还是来自文凭本身打开了升学和就业的门。另外，终身收入只算个人回报，不含教育对他人生产率、福利和整体经济的影响。还要记住样本是近二十年前参加 GCSE 的人，劳动力市场对某类技能的评价一直在变。报告的结论是，这些估算适合用在政策评估的成本收益分析里，不适合拿来预测某个学生未来能赚多少。一份 2021 年发布的报告放在 2026 年读，最稳妥的用法是看方向和量级，而不是盯住小数点。
The report lists three caveats in its own pages, and they belong next to the numbers. First, the data cannot capture every trait that shapes both grades and earnings, so unobserved characteristics may leave the estimates biased. Second, the method cannot separate human capital from signalling, which means we cannot tell whether higher pay reflects skills genuinely learned or simply the credential that opens doors to study and work. Third, lifetime earnings measure private returns only and exclude any effect on other people's productivity, welfare or the wider economy; on top of that, the sample sat its GCSEs almost twenty years ago, and labour markets keep revaluing particular skills. The authors' own conclusion is that these figures suit cost-benefit appraisal of policy, not forecasts of an individual's future pay.
十、一档之差落在卷面上：错题与时间分配
把研究放回考场，一档的差距很少来自最后一道大题。很多学生以为一档的提升靠冲刺难题，实际丢分最多的是基础题的稳定性，尤其是算术环节和单位换算。从 D 到 C、从 C 到 B 这两段回报最高，试卷上对应的也多是中档题，这类题的评分标准对步骤完整度要求明确。阅卷时最常看到的情况是过程写了一半就跳到答案，中间那一步的分数拿不到。非计算器卷的差别更明显，分数与百分数的常见等值、根式化简、估算复核，这些动作熟练与否直接决定准确率。常见的错误是把时间平均分给整张卷子，结果长题没写完，前面的基础题也没来得及检查。可行的做法是把每次模考的错题按原因归类，分清是知识漏洞、计算失误还是读题偏差，再决定下一阶段补哪一类。一整轮复习下来，能稳定拿回的分通常就集中在这几类丢分上。
Back in the exam hall, the gap between one grade and the next rarely comes from the final long question. Many students assume the jump is earned by chasing hard problems, yet most marks are lost on routine items, particularly arithmetic and unit conversions. The D to C and C to B steps, which carry the largest returns, sit mostly on mid-tier questions where the mark scheme rewards complete working; what examiners see most often is a method started and abandoned halfway, which forfeits the middle marks. On non-calculator papers the difference is sharper, since quick recall of fraction and percentage equivalents, surd simplification and estimation checks decides accuracy. A common mistake is to spread time evenly across the paper, so the long questions remain unfinished and the earlier ones go unchecked. Grouping mock-exam errors by cause, whether a knowledge gap, a slip in calculation or a misreading, points to what to revise next; across a whole revision cycle most recoverable marks cluster in exactly those categories.
海外直订Higher Mathematics for AQA GCSE 高等数学AQA GCSE¥645.0📱 去京东购买京东精选需要一对一辅导？Need One-on-One Tutoring?
A-Level / GCSE / IB / KS2-KS3 全科辅导
16621398022 同微信
Visit aleveler.com
或截图后用「扫一扫 → 相册」选择本图