Chinese in South Africa: The 1904 Gold-Mine Labourers Who Built the Rand

Chinese in South Africa: a migration the diaspora story skips

Most histories of the Chinese abroad in the gold-rush age stop at the harbour. They follow the men who reached San Francisco, the diggings of Victoria, the cold gullies of Otago, the sandbars of the Fraser River — the familiar ports and fields of the great nineteenth-century migrations. Between 1904 and 1910, a migration of comparable size crossed two more oceans to a shore almost none of those histories visit. Britain shipped roughly 63,000 Chinese contract labourers (契约华工, qìyuē huágōng) to the Witwatersrand (威特沃特斯兰德, Wēitèwòtèsīlándé), the gold reef known simply as the Rand, in the Transvaal (德兰士瓦, Délánshìwǎ) — the largest importation of Chinese labour ever made on the African continent.

What set the Rand apart from every earlier migration was its instrument. This was not the private credit-ticket trade that carried men to Cuba and Peru, the system the site has already traced elsewhere under the name of the coolie trade. The Rand ran on a signed government-to-government treaty: the 保工章程 (bǎogōng zhāngchéng), the Sino-British Labour Convention of 1904. Britain did not deal with brokers on this one. It dealt with the Qing state.

Two more things make the episode its own story. The men came overwhelmingly from the north — Shandong (山东, Shāndōng) and Zhili (直隶, Zhílì, modern Hebei) — moving the opposite way from the Cantonese migrations that dominate the diaspora record, though a smaller, documented Cantonese stream travelled with them. And the scheme was built to return them. A compulsory-return clause meant that everyone involved, at the time, assumed it would leave nothing behind.

Most writing on “Chinese in Africa” gives this a paragraph. This is the full arc: recruitment, indenture, resistance, a political storm in Britain, repatriation — and the community that remained anyway.

Why the Rand ran on imported labour

The scheme was not a Chinese push. It was a British demand, and it began with a war. Between 1899 and 1902 the South African War — the Anglo-Boer War — wrecked the labour supply that had kept the pre-war Rand running. Black African mineworkers had been driven off the workings by the fighting and the scorched-earth sweeps; when the war ended, the mines reopened faster than that workforce could be re-secured. The Transvaal Chamber of Mines faced a simple arithmetic: gold underground, machines ready, and no men to drive them.

The answer the mine-owners reached for was importation — not immigration in the ordinary sense, but a temporary, contracted, and fully returnable workforce, brought in, worked, and sent home. The mines reopened and the men were not there.

The institution that housed them was not new either. The fenced, all-male mine compound had already been built to control Black African labour on the Rand. The Chinese were moved into a system the mines had long used; what made their version distinctive was its completeness — men without families, movement by pass, and a contract that criminalised leaving.

For the mine-owners, the appeal was control. An imported labour force under a fixed term could be housed, fed, and paid on the mine’s own terms, held behind a fence, and dismissed from the country when the term ran out. It would not settle, would not buy land, and would not compete for skilled jobs. On paper the arrangement answered the war’s deficit at the lowest political cost. That the men themselves would not behave like the passive instrument the scheme imagined is the part of the story the plan did not anticipate.

The Rand was, in one useful frame, the sixth great gold field of the age. California, Victoria, Otago, and the Fraser River had drawn Cantonese men across the Pacific; the Rand drew North-China men across the Indian Ocean, on the far side of the world. The field was the last and the least remembered.

The treaty of 1904, and the machine that fed it

The instrument that made all of this legal was signed in London on 13 May 1904. The 保工章程 (bǎogōng zhāngchéng) — the Sino-British Labour Convention, formally the 中英会订保工章程 — bore the signatures of the British Foreign Secretary, Lord Lansdowne, and the Chinese Minister in London, 张德彝 (Zhāng Déyí). It was a real bilateral treaty, and it fixed real terms: a three-year first term, a ten-hour working day, a wage of about one shilling a day, rations, medical care, an age and health inspection on embarkation, and an obligation to return the men to China. Mae Ngai notes that the convention set a minimum age of twenty and required inspection to ensure the labourers were of “sound body and mind.”

Then came the hollowing. 张德彝 pressed five protective conditions on China’s behalf. Two of them went to the heart of the matter: that China be allowed to station a consul to protect the workers, and that “雇主不得责打工人” — that employers must not beat the labourers. Britain refused to write the no-beating prohibition, and a parallel ban on sub-letting, into the treaty itself. The convention fixed only general principles; the specifics of hours, wages, food, and medical care were left to individual contracts, which the colonial authorities could and did alter at will. Chinese accounts describe the promised protections as “徒具空文” — empty paper.

Whether the Rand men were free contract-labourers or disguised coolies is itself a live historiographical question, and it is left open here. The treaty’s protections were strong on paper and hollow in practice, and both readings draw on that same fact.

The machine that fed the pipeline reached across northern China. The principal channel ran through the British-leased territory of 威海卫 (Wēihǎiwèi, Weihaiwei) in Shandong, with the practical embarkation ports at 芝罘 (Zhīfú, better known as Yantai) and 秦皇岛 (Qínhuángdǎo) in Hebei.

This was a North-China corps. But it was not North-China only: a documented Cantonese stream — the 粤工 (Yuègōng) — was recruited through Macau, Canton, and Hong Kong under the 1904 《南非洲英属特兰士瓦招募粤工开矿合同》 printed in the journal 《外交报》. In March 1904 the shipping firm Butterfield & Swire posted recruitment notices in 嘉应 (Jiāyìng) and four townships in 梧州 (Wúzhōu), the paperwork of a southern channel running alongside the northern one.

From Weihaiwei to Durban

The voyage ran from the northern depots — 威海卫, 芝罘, or 秦皇岛 — down to Hong Kong, and then across the Indian Ocean to Durban (德班, Débān) on the Natal coast. The first landings came in mid-1904, and the sources do not agree to the week. One account gives the first 1,054 contract labourers shipped from Hong Kong; another dates the first arrival at the Witwatersrand to 19 June 1904; a third records the SS Tweeddale docking at Durban in July 1904. They are different legs of the same opening season, and the record is best left as a range rather than forced into a single day.

The clearest Chinese voice on the ships belongs to 谢子修 (Xiè Zǐxiū), who had arrived in South Africa in October 1903 and joined the Chinese association (中华会馆, Zhōnghuá huìguǎn) in Johannesburg. He recorded a 1904 voyage in his own hand: a consignment of 1,969 Chinese men. In September 1904 he took a post as a mine inspector at the East Rand Proprietary mine and later wrote 《游历南非洲记》 (Yóulì Nánfēizhōu Jì), an account of more than seven thousand characters sent to the Qing court and now held in the First Historical Archives of China. He explained his motive plainly: “余意尽欲探其内容，非贪其薪俸也” — he wanted to see inside the system; he was not chasing the salary.

The name that startles readers is Herbert Hoover (胡佛, Húfó). The future American president, then an engineer with the London firm Bewick Moreing, was a director of the shell company — the Chinese Engineering and Mining Company — that secured the recruitment concession and ran the 秦皇岛 pipeline that shipped men south from July 1904. Hoover was an officer of the machinery, not its architect, and the per-head commissions reported for him — figures around £1 17s 6d — are second-hand and inconsistent between sources. They belong in the story as a curiosity, not as a ledger.

Life in the compound

What waited at the far end was the compound. Transvaal Ordinance 17 of 1904 made the fenced, all-male labour camp (矿工大院, kuànggōng dàyuàn; also 营区, yíngqū) the legal form of life on the Rand. Movement was controlled by pass. Desertion was a crime, punished by fine, imprisonment, or deportation. Skilled and semi-skilled work was forbidden outright: the Chinese text of the Transvaal recruiting contract read “除粗工以外，不得另作别样工夫” — no work other than unskilled labour.

The compound was all-male by design. No Chinese women were permitted on the Rand, which is why the cohort produced almost no second generation. The historians Melanie Yap and Dianne Leong Man call it “a community of men without women” — a phrase worth keeping attributed, because it names the structural fact precisely. Inside the fence lay crowded barracks, mine dust, dust-disease, accidents, and a discipline that included the whip. An informal economy of opium, gambling, and credit ran alongside the official one; the Qing and British authorities corresponded directly over 华工服食鸦片 — the labourers’ use of opium.

The mortality is the hardest figure to hold. One widely circulated account puts more than 6,000 of the Chinese killed or injured between 1904 and 1906 — roughly a tenth of the corps. That figure rests on Chinese secondary summaries rather than a located primary returns table, and it should be read as a probable, single-source estimate attributed to that account, not as an official count.

The concrete objects tell it better than any adjective: a pass that had to be carried at all times, a gate that was locked at night, a canteen ticket spendable in one place only, a metal token that bought goods at the company store and nowhere else.

One shilling a day

The wage was the scheme’s quiet fault line. The negotiated rate under the convention was about a shilling a day — roughly £1 to £2 a month — plus rations, with part of it deferred. That deferred portion, the 养家银 (yǎngjiāyín), was remitted to families in Shandong and Zhili. Uniquely among the mass Chinese labour migrations, the contract built an official, state-administered remittance channel. It ran apart from the private letter-and-courier system — 侨批 (qiáopī) and 水客 (shuǐkè) — that belongs to the Wuyi diaspora, and it is the reason almost no private remittance correspondence survives from this cohort. The actual volume that moved through the formal channel versus informal ones is thinly evidenced; no reliable total can be given.

Whether the Chinese were paid more or less than the Black African miners beside them is a genuinely contested question, and the sources do not settle it in one direction. The weight of the contemporary Chinese and colonial-press evidence puts the Chinese day-rate below the Rand’s minimum Black African wage — one shilling against roughly one-and-sixpence, and by one 1904 Weihaiwei report 15 to 21 yuan a month against 75 yuan for Black miners. One scholarly reading, associated with the historian Li Anshan, argues the opposite: that the Chinese rate was set above the African rate as a divide-and-rule tactic. That reading stands against the weight of the contemporary evidence and should be carried as a contested book view, not as a settled ranking.

A further comparison, from one documented accounting, sharpens the inequality: Black African miners were assigned 32 inches a day of rock advance and Chinese labourers 36 inches, with make-up work after shift if the quota was unmet. And take-home pay shrank again at the company store. Deductions for food, clothing, tools, and canteen tickets, and wages paid in non-negotiable iron tokens usable only at company shops selling shoddy goods at inflated prices, pulled the real wage down below the printed one.

Resistance on the reef

The men resisted, and the record shows it. From late 1904 and intensely through 1905 and 1906, Chinese labourers struck, slowed their work, refused to renew their contracts, and deserted in waves. They were moved by the quality of the food, by the deductions that ate their pay, by the confinement of the compound, and by physical abuse — the very protections the treaty had promised and never delivered.

The workers’ own words survive in fragments. “我只拿到一先令，我已经干的够多了” — I only get one shilling, and I have done enough.

The numbers are indicative rather than audited, and they come from Chinese secondary sources. One account documents 28 serious disturbances between June 1904 and June 1905, more than 21,000 recorded slowdown incidents in the first year, and more than 1,000 deserters in a year. Some 13,532 labourers were punished or fined in 1905 and 1906, more than a fifth of the corps. These should be read as evidence of a pattern, not as a final tally.

The Qing state was listening. It had appointed a Consul-General, 刘玉麟 (Liú Yùlín), in 1904; he reached South Africa in April 1905 and in June 1905 toured the mines and made the abusive practices public. His exact title and tenure should be treated as provisional, but the 1904–05 consul and inspections are documented. The consulate was an early instance of what came to be called 护侨 (hùqiáo) — the bureaucratic protection of overseas Chinese — and alongside 谢子修 on the ground it gave the men the thinnest sliver of an official ear.

By 1906 the “cheap labour dream” had inverted. In the words of one Chinese account, the Transvaal had become “不仅…一座地狱，而且是一座陪钱的地狱” — not only a hell, but a hell that lost money.

The Chinese Question

The scheme’s fate was decided not on the Rand but in Britain. The slogan was “Chinese slavery in the Transvaal,” and it was driven by white mining unions and by Labour and Liberal politicians who argued that “coolie labour” undercut white wages. In their telling, the Chinese worker was the instrument of a wage cut; the campaign against him was a labour campaign first.

The hinge came in January 1906, when Britain went to the polls. The Liberal Party won a landslide, 29 Labour candidates took seats, and the Chinese-labour question — as much as any single issue — carried the result. It is arguably the only British general election ever decided on a question about Chinese labourers.

The racial pivot deserves to be told honestly. Radicals mobilised working-class racial sentiment; the elite Liberal press took a moralist anti-slavery line. Once in power, the Liberals largely continued the Conservative policy they had denounced. Pressed on the “new slavery” charge, the Prime Minister, Campbell-Bannerman, conceded the phrasing might have been “有点过激” — a bit excessive.

The politics moved fast from there. In early 1906 the Liberal government granted responsible government to the Transvaal, pressed by Jan Smuts — an effective prelude to the creation of the Union of South Africa in 1910. White labour had won the political contest. The discontinuation of Chinese indentured labour was, by then, the issue of the day, and the Union of 1910 in part settled it.

One verdict on the scheme’s economics is documented but not universal: on this view it proved costlier and less productive than the mine-owners expected. Some mining records contest the claim. It stands as one reading of the accounting, not the final word.

Repatriation, and a record that says no one stayed

The wind-down was slow, then complete. Recruitment in China halted in November 1906. The law under which the men were held lapsed on 21 March 1908. The Rand still held 41,319 Chinese in 1907, and the repatriation ran from 1907 to 1910. The last batch left the Transvaal in March 1910.

What remained was almost nothing. Official figures suggest only about 1,862 Chinese miners were still on the Rand by 1910. The community memory states the outcome more starkly still: from 1907, one history records, “直到今日为止尚未发现有矿工留居南非的证据” — to this day no evidence has been found of any miner settling in South Africa. That phrasing belongs to the official record, and it should be read as an official-record impression rather than an absolute. A small number did stay on, and merged into the free Chinese community that already existed.

Why so few remained was built into the scheme from the start. The compulsory-return clause — 资遣回华 (zīqiǎn huíhuá), the funded return to China — combined with the all-male design and the practice of deporting offenders left almost no room for settlement. This was a labour rotation, not a migration of families. The men came on three-year terms, worked, and were sent back, in a process so orderly and so total that it looked, to those who counted it, like something that had never happened at all.

The contrast with the California or Australian gold fields is exact. There, Chinese men also arrived alone, but no treaty compelled them out, and enough of them stayed, married, and built communities to leave a permanent mark. On the Rand the state removed the men it had imported; the return clause was not a formality but a design choice, and it worked.

The community that did not disappear

Here is the counterintuitive turn, and it is worth stating plainly: it is largely a myth that the contracted miners became the forefathers of South Africa’s Chinese population. The enduring Chinese South African community descends mainly from the free Chinese merchants and artisans who were already there before 1904 — in Cape Town and Kimberley, in the 1880s and 1890s — not from the repatriated corps. Those men were rotated and returned, and they left behind little durable institutional life. A small miner residue may have stayed on and merged into that free community, but it was not the community’s foundation.

Exclusion ran through all of it. The Cape Colony passed anti-Chinese legislation in 1904 that stripped Chinese residents of the franchise, on top of the entry restrictions the Orange Free State and the Transvaal had imposed in the 1890s. The scheme launched under conditions of racial exclusion, and those conditions long outlasted it.

Under apartheid the position of the community stayed strange. The Population Registration Act of 1950 classified Chinese South Africans as “Coloured” (有色人, yǒusèrén) — neither White nor Native. Yap and Leong Man call it, in the title of their history, a matter of confusion; the group was barred from white public schools, pushed into Chinese-medium education and business niches, and occasionally granted “honorary white” (荣誉白人, róngyù báirén) status. The community’s own summary of its position — not white enough under apartheid, not black enough under the ANC — captured the bind.

The legal payoff came after 1994. The community had joined the anti-apartheid struggle, and it won recognition that it had a place in the post-apartheid dispensation. The anchor event is the judgment of the Pretoria High Court in Chinese Association of South Africa and Others v Minister of Labour and Others, [2008] ZAGPHC 174, delivered on 18 June 2008, case number 59251/2007. The court held that Chinese South Africans who were citizens before 1994, and their descendants, fall within the definition of “black people” — as Coloureds — for the purposes of Broad-Based Black Economic Empowerment (黑人经济振兴, hēirén jīngjì zhènxīng). The ruling ended an eight-year legal battle and benefited roughly 10,000 to 20,000 citizens. The 18th of June is now a community commemoration day. Community accounts sometimes add a further appeal step in 2015; that step could not be verified against a law report, so this article rests on the 2008 judgment itself. What the ruling did and did not do also matters: it applied to pre-1994 Chinese South African citizens and their descendants, not to every Chinese person in South Africa, including those who have arrived since 1994.

The successor economy kept the community visible — a niche set of laundries, market gardens, cafés, general-dealer shops, and the numbers game known as fah fee, distinct from the American hand-laundry-to-restaurant arc.

The men of 1904 were sent home so completely that the record said they had left nothing. A Chinese South Africa survived them anyway. It was misclassified by a state that could not quite place it. On a June day in 2008 a court confirmed that it had always been there. That is the chapter the record tried to file under “no evidence,” and it has a name.

Continue Reading

• The coolie ships of the Americas: Cuba, Peru and Jamaica
• Chinese railroad workers and the exclusion era
• The Hakka diaspora beyond Guangdong
• Fujian and Chaozhou: the other Chinese diaspora

Stay Connected

Follow RootsOfChina on Facebook and Instagram for new article updates, behind-the-scenes from the qiaoxiang, and stories from the Chinese diaspora. New diaspora histories land every week — join the newsletter to have the next one sent to you.

Sources

• SAHistory.org.za, “Arrival of Chinese labourers in South Africa”
• UK Parliament, Historic Hansard, “Chinese Labour in the Transvaal” (9 March 1904)
• 中国大百科全书, “南非金矿契约华工 / Indentured Chinese Labour in South African Gold Mines”
• 清史研究, 《南非契约华工的历史原因分析》
• 张芝联, 《1904–1910年英属南非德兰斯瓦尔招用华工事件的真相》, 北京大学学报 1956
• 三联生活周刊, “契约华工在南非”
• 界面新闻 / Paul Bailey, “从山东到索姆河”
• China News, feature on Yap & Leong Man, Colour, Confusion and Concessions
• 《契约华工在南非(1904-1910)》 (wage comparison; 1s vs 1s 6d)
• 南非契约华工的历史原因 (the 63,811 port-by-port export tally)
• 民族复兴网 (63,811; first 1,054 from Hong Kong; Hoover/CEMC)
• 开平市政府, “游历南非记” (谢子修《游历南非洲记》)
• 海外网 / 人民日报海外版, “护行天下——华侨护照展” (刘玉麟 appointment)
• 威海市档案馆 (Weihai Municipal Archives), “一战华工营”
• theheritageportal.co.za (Kathy Munro), “Ordinances of the Transvaal 1903, 1904…”
• Lexology, “Broad-Based Black Economic Empowerment” ([2008] ZAGPHC 174, case 59251/2007)
• RSM South Africa, “B-BBEE Frequently Asked Questions”
• Chinese Association of South Africa and Others v Minister of Labour and Others [2008] ZAGPHC 174 (18 June 2008)
• Books: 李安山《非洲华侨华人史》(2000); 李安山《非洲华人社会经济史》(2019); 陈翰笙《华工出国史料汇编·第九辑 非洲华工》(1984); 方积根《非洲华侨史资料选辑》(1986); Mae Ngai, The Chinese Question (W. W. Norton, 2021); Rachel K. Bright, Chinese Labour in South Africa, 1902–10 (Palgrave Macmillan, 2013); Melanie Yap & Dianne Leong Man, Colour, Confusion and Concessions (HKU Press, 1996).