EVOPEDIA

← all systems · see it in The System

Imperial examination (China)

The imperial examination was a civil service examination system in Imperial China administered for the purpose of selecting candidates for the state bureaucracy.

Properties published scienceour proposal

Sourced fields, and our own readings.

aliasesThe imperial examination system
born atbetween 587 and 622 (622: opened to most men); examination of nominees from the Han
born precisiondecades
boundaryInside: the examination ladder and its machinery: registration and the door (eligibility, identity, searches), the tiers of examinations from county to palace, the sitting (cells, sessions, set forms such as the eight-legged essay), the marking under anonymity (sealing, recopying, several examiners), the ranking against quotas, and the degrees and pass lists it issues. The throne's examining role is inside (decreeing the rules, curriculum and quotas, the palace examination it supervises, inspecting lists and ordering retests: board L7); the Ministry of Rites that held the examinations is inside. Outside: the candidates as people (walked as Human), their families and lineages, schools and books that prepare them, the rest of the throne and the ministries that appoint the passers (the state, Imperial China, the bureaucracy), and the canon itself (Confucian classics).
end precisionday
ended at2 September 1905 (China); Korea 1894; Vietnam 1919 (or 1913)
nested levelA part of the imperial Chinese state (Imperial China's examination channel and the bureaucracy's door), built from candidates, examiners, the canon and quotas; one kind of examination (sys:examination), copied into Korea, Vietnam, Heian Japan and Ryukyu and echoed in 19th-century Western civil services.
scalesocial

Measures (46) published science

imperial examination lifetime587/622 to 1905 (about 1,300 years)
examination cycle (clock)every three years for the regular tiers (Song on); Joseon regular gwageo triennial
age at passing (settling time)average jinshi candidate in his thirties (China); Joseon 28.1 (regular), 31.2 (special) years of age
time for Keju to dominate the top (lag)about a century: chief ministers with a Keju start 22% (618-649) to 61.9% (705-755), 85% in the 9th century years
abolition to revolution (lag)1905 to 1911 years
Tang chief ministers with a Keju startunder 22% (mid-7th century) to 85% (9th century) percent
Tang male elites with an examination credential3.8% (before 649) to 17.6% (9th century) percent
office-rank return to passing (late Tang)two full ranks of the nine-rank system
Tang epitaph sample3,640 epitaphs of males (618-907)
father's rank effect on son's rank (Tang)0.28 non-passers; 0.08 (not significant) passers points per point
Tang passers from prominent aristocratic branches38% (epitaph sample) vs 76% (Sun 1980, list of 301) / 80% (Kung 2022) percent
Tang pass rates by categoryclassics tests 10-20%; jinshi 1-2%; 6,585 jinshi in the dynasty percent
Song adult males sitting the prefectural exambelow 0.5% (11th century) to above 2% (13th century), several regional series (Chaffee) percent
Song officials by mode of entry (1213)40.8% jinshi; 52.5% protection (yin), ranks six to nine percent
Ming jinshi per yearabout 89 (89 regular exams in 277 years; over 24,000 in total) jinshi per year
Qing jinshi26,747 in 112 examinations, 1644-1905; average 238 per exam degrees
licentiates per headone per 2,800 (about 1400) to one per 540 (1700) persons per licentiate
Qing end-to-end success oddsslightly better than 1 in 6,000 (0.01%) probability
Qing provincial pass rate0.33-0.5% percent
late Qing pass rate1% of two or three million annual applicants percent
shengyuan to jinshi1 in 3,000 ratio
father-in-office effect on top-tier odds+108% odds (2.14% vs 1.38% probability) odds ratio
ministers with an examination background0 before the 6th century to above 80% in the Ming percent
ministers with a son in officeabout 49% before 1200; 5.6% after the 16th century percent
Qing local officials by mode of entry (1764)Keju 72.5%; purchase 22.4%; yin 1% percent
ruler duration with Kejutext: Keju +4.8 years, parliament +4.9; Table 2 column (1): Keju_dummy +6.044 (SE 1.540), Parliament +4.947 (SE 1.906); column (2): Parliament +4.786 with Keju_scale +0.141 per graduate years of reign
Qing jinshi reaching rank threeabout 12.6% of 24,269 percent
exam-rank effect on top office (Qing)+3 percentage points (24%) per 10 percentile points probability
examiner-tie effect on top office (Qing)+17% per SD of ties to powerful examiners probability
jinshi son of a jinshi fatherabout 7% (Tongcheng), matching Ho's 7.1% percent
jinshi sons staying in the top 2 percent1 in 3 share
Ming jinshi from registered scholar families7.1% of 22,577 (Ho 1962) percent
intergenerational rank coefficient (Tongcheng)fell about 20 points from 1700 to about 0.45 in the 1800s beta
juren with a degree-holding ancestor (five generations)87% percent
juren and jinshi with a degree-holding father or grandfather50-60% early Qing; 60-70% mid-late Qing percent
Joseon literary examination passersabout 14,600 (three-fifths through special exams) persons
Joseon examinations held163 regular, 573 special examinations
Joseon second-stage pass rate (1617)under 2% (about 2,000 applicants, 33 places) percent
Joseon medical licensing exams233 sittings, 1397-1894 examinations
Vietnamese national-level passers2,888 from more than 180 examinations (1075-1919) persons
Vietnamese national selection rate (1463)1% percent
long-run schooling legacy (Vietnam)+0.12 years of schooling and +0.2% literacy per extra laureate in the district
long-run schooling legacy (China)+0.7 years of schooling today per extra jinshi per 10,000 people in Ming-Qing
per-stage yield (Payment to Entropy reading)Qing chain: provincial 0.33-0.5%, end-to-end about 0.01%; Tang jinshi 1-2% usable out / sent in
Song degree-holders without an official ancestor (three generations)about 57% (1148 and 1256 lists); most had some relative in the bureaucracy percent
cells in the largest examination hall20,644 (Jiangnan, Nanjing; encyclopedia, citation needed) cells

Open contentions (10) open contention

Where sources or readings disagree. Both stay on the map.

Cross-scale patterns seen here (19) our proposal

Shapes that recur when the same kind of problem is solved at different scales, and how each shows in Imperial examination (China). Each links to every other system that shows it.

Valence: what it trades, and with whom (15) published science

Each exchange with a neighbour, read from both sides: who gains, who pays, and what carries it.

Partners and parts published science

Built from / builds (10)

Sits inside / holds (3)

expressed by (14)

floor filled by (17)

Kinds (12)

wired (1)

Notes published science

Layers

The ways it takes in its world, where they meet, and how it acts back, each from L1 (where it meets the world) up.

imperial examination: from the empire's students to the throne's appointments (reader's view)

  1. L1: The examination's world: the empire's educated men (licentiates one per 2,800 people around 1400, one per 540 by 1700), their families and lineages paying for books, tutors and travel, the schools and printed books that prepare them, and the offices waiting at the top
  2. L2: The door: registration and self-nomination, eligibility (sumptuary bars lifted in the later Ming; women excluded), quotas on entry by prefecture, guards verifying identity and searching for hidden texts
  3. L3: The ladder: county and prefectural entry exams (tongsheng, shengyuan), the triennial provincial exam (juren), metropolitan (gongshi) and palace (jinshi) exams, on the canon fixed by the throne (Zhu Xi's Four Books from 1313/1370)
  4. L4: The sitting: each candidate in a numbered cell, no outside communication, three sessions over days; answers in the set forms (the eight-legged essay)
  5. L5: Marking under anonymity: names hidden (702), papers anonymous (992-1032), recopied three times so calligraphy cannot be known, two examiners and a determining official; examiners barred from their home prefecture; the examiner-examinee bond formed here
  6. L6: Ranking against the quota: pass lists in ranked tiers (the top three; zhuangyuan first), degrees with legal and social privileges, quotas by prefecture, province and (Ming) north-south-centre; rulers inspect the list and order retests
  7. L7: Appointment and the state's purpose: degree holders become candidates for office (Hanlin Academy, provincial posts, chief ministers); the throne holds local elites' loyalty and a larger, more replaceable selectorate; the ruler's power and stability

the boardroom (shared floors)

  1. L5: Marking under anonymity: sealing office, copy office (three copies), two examiners and a determining official; luck as important as talent for one child (Elman, via Hao)
  2. L6: Ranking against quotas into tiers and degrees with privileges; the list read by the court and the public; regional quotas (Ming north-south-centre); retests and revocations
  3. L7: The throne's use and rule of the examination: appointment of passers, the size of the selectorate, the loyalty of local elites, the rules and curriculum decreed, reform and, in 1905, abolition

way in: Candidates arriving to sit

  1. L1: Educated men coming forward: more than 400,000 registering for Song prefectural exams by the mid-13th century; two or three million applicants a year in the late Qing
  2. L2: Registration and the gate: self-nomination, eligibility, identity verified, searched for hidden texts
  3. L3: Climbing the tiers: entry exams (shengyuan), provincial (juren), metropolitan and palace (jinshi); only provincial passers may sit the national level (Vietnam the same)
  4. L4: In the numbered cell with no outside contact, writing the paper

way in: Cheating and favour-seeking

  1. L1: Hidden copies of the classics (a handkerchief of 10,000 characters), bribes, officials' sons seeking favour
  2. L2: Smuggled past the gate (texts on underwear)
  3. L3: Pre-exam influence on examiners by circulating essays in the capital (forbidden in the Song)
  4. L4: Used in the cell; or identity of the writer signalled to the reader

way in: Examiners and their networks

  1. L1: Officials serving as examiners with gate supervisors, registrars, sealers and copyists
  2. L2: Appointed per examination; barred from their home prefecture (1037)
  3. L3: Reading only sealed, numbered, recopied papers
  4. L4: The examiner-examinee bond formed by the cohort an examiner passes: master and disciple

way in: Families and lineages investing

  1. L1: Families and lineages with land and money; wealthy non-scholar families also investing
  2. L2: Lineage 'land for education' and funds; the most promising sons chosen and tutored (China and Korea)
  3. L3: Tutors, lineage schools and as many attempts as possible for the chosen
  4. L4: The son at the sitting: a father in office raises the odds of a top-three rank by 108% (whole sample), though not of passing in the Tang epitaphs

way in: Outside pressure and reform

  1. L1: Western technology, military defeats in the 1890s, pressure for a national school system
  2. L2: Officials' proposals from the 1830s-1840s to add Western technology
  3. L3: Proposals to add mathematics and science (1861-1874); Qing rulers keep science out
  4. L4: Hundred Days' Reform (1898) and the New Policies after the Boxer Rebellion
  5. L5: Reformers memorialise the throne; key sponsors Yuan Shikai, Yinchang, Zhang Zhidong
  6. L6: Seven controversies on reform and abolition across 1,300 years (content and format; examination or recommendation; examination or schools)

way in: Public outcry over a list

  1. L1: Candidates and the public reading a published list
  2. L2: Outcry over officials' sons (zidi) on the list; protests (Yangzhou 1711)
  3. L3: The emperor scrutinises the newly announced list
  4. L4: Retest ordered; names struck from the list
  5. L5: Degrees revoked; examiners demoted (789) or punished

way in: Regions and their claims

  1. L1: Regions of unequal wealth and population density; density raises top-tier odds (+37% per national SD)
  2. L2: Claims and complaints: the 1397 all-southern list; memorials on north and south
  3. L3: Quotas allotted to prefectures and provinces (Song on; 1712 provincial reform; Guangxu remote-region quotas)
  4. L4: Each province's examination run under its quota; candidates travel long distances to venues

way in: The canon and what is asked

  1. L1: The Confucian classics and their commentaries, printed and sold; block printing's ink (pine) and paper (bamboo, mulberry)
  2. L2: The canon fixed by the throne: Zhu Xi's scholarship as standard (Yuan), the Four Books (Ming, 1370)
  3. L3: Schools and academies teaching to the examination; the curriculum narrowing over time toward memorisation
  4. L4: The question set and the set form in three sessions (Ming-Qing): the Four Books and Classics (eight-legged essays), then a discursive essay, judgments and documentary forms, then five essays on Classics, history and current affairs; exact quotation required

way out: Degrees as status

  1. L1: Families and lineages raised (and their property protected); failed candidates keep local standing as teachers, patrons and managers
  2. L6: Degree conferred with legal privileges (shengyuan: no statute labour, access to local government, limited immunity from corporal punishment); the pass list published

way out: Degrees into office

  1. L1: Officials posted across the empire: 72.5% of mid-rank local officials with a Keju background by 1764
  2. L6: Honours graduates appointed to the Hanlin Academy; Hanlin and provincial posts lead to chief minister (56% of the Keju effect)
  3. L7: Passers named candidates for office; in the Tang a separate vetting by the Ministry of Personnel, with delays

way out: Rules decreed back onto the sitting

  1. L2: Gates, searches and cells enforced at the next sitting
  2. L7: Decrees: anonymity (702, 992-1032), recopying (1015, 1037), examiners barred from home prefectures, quotas, retests, curriculum

way out: What is tested shapes what is learned

  1. L1: A society that values education: more schooling and literacy today where more passed (China, Vietnam)
  2. L3: Schools become testing centres; boys trained to write for the exam
  3. L7: The canon and the form decreed (Zhu Xi; the eight-legged essay)

Sources (16) published science

Imperial examination (Wikipedia, via Wikimedia Enterprise snapshot)
Wang EH, Yang C (2025) The Political Economy of China's Imperial Examination System. Cambridge Elements, Cambridge University Press. doi:10.1017/9781009347600 (open access; full text read through the paper portal) (behind its links)
Eight-legged essay (Wikipedia; via Wikimedia Enterprise snapshot) (behind its links)
Choi D, Kim C (2026) Meritocracy with Multiple Rules: How Special Civil Service Examinations Reproduced Privilege in Joseon Korea. Social Science History. doi:10.1017/ssh.2026.10135 (full text read) (behind its links)
Bai Y, Jia R (2016) Elite Recruitment and Political Stability: The Impact of the Abolition of China's Civil Service Exam. Econometrica 84(2). doi:10.3982/ecta13448 (abstract only; closed) (behind its links)
Wen F, Wang EH, Hout M (2024) Social mobility in the Tang Dynasty as the Imperial Examination rose and aristocratic family pedigree declined, 618-907 CE. PNAS 121(4): e2305564121. doi:10.1073/pnas.2305564121 (PMC10823257) (behind its links)
Hao Y (2021) Social mobility in China, 1645-2012: A surname study. China Economic Quarterly International. doi:10.1016/j.ceqi.2021.08.002 (full text read) (behind its links)
Bai Y, Jia R (2026) Meritocracy and the Networks It Forged: China's Imperial Examination. NBER Working Paper 35556. doi:10.3386/w35556 (full text read) (behind its links)
Shiue CH (2025) Social Mobility in the Long Run: An Analysis of Tongcheng, China, 1300 to 1900. Journal of Economic History. doi:10.1017/s0022050725000142 (full text read) (behind its links)
Lee NH (2015) Medical licensing examination (uigwa) and the world of the physician officers (uigwan) in Korea's Joseon Dynasty. J Educ Eval Health Prof 12:16. doi:10.3352/jeehp.2015.12.16 (sections read: abstract, the licensing examination and its rosters) (behind its links)
Vu TM, Yamada H (2020) Persistent legacy of the 1075-1919 Vietnamese imperial examinations in contemporary quantity and quality of education. Keio-IES Discussion Paper DP2020-012 (full text read, own paced fetch of the Keio copy; published 2023 in J Comparative Economics 51(1), not read) (behind its links)
Chen Z, Ma C (2023) Quantitative history studies on China: State capacity, institutions, culture and human capital from prehistoric times to the present. Asia-Pacific Economic History Review. doi:10.1111/aehr.12272 (sections read: Confucianism, human capital and long-term persistence) (behind its links)
Jiangnan Examination Hall (Wikipedia, via Wikimedia Enterprise route; sections read: the cell-count line) (behind its links)
Chen T, Kung JK-s, Ma C (2015) Institutions, Culture, and Human Capital in the Long Run: Legacies of China's Imperial Exam System. Slides, Warwick Summer Workshop in Economic Growth (sections read, own paced fetch); the published paper (Long Live Keju!, Economic Journal 2020, doi:10.1093/ej/ueaa043) was not reached; step 3: the portal again gave the Economic Journal record only (DOI), and the OpenAlex search for its work id answered 502 (behind its links)
Yuan J, Zhang A (2025) The imperial examination and the college entrance examination from the perspective of affirmative action: assistance to disadvantaged examinees in China. Humanities and Social Sciences Communications. doi:10.1057/s41599-025-05366-x (sections read: the imperial-examination sections on affirmative action, resource assistance and quotas) (behind its links)
Li M, Liu H, Huang W-C (2022) A Historical Investigation and Classification Analysis of Seven Controversies on the Reform and Abolition of the 1300-year Chinese Imperial Examination System. International Journal of Educational Research. doi:10.1016/j.ijer.2022.101976 (sections read: abstract, the seven controversies, recommendation and academy controversies) (behind its links)