In the late 1950s, the colonial government of Hong Kong decided to conduct a census for the first time since the Second World War. With a territory of just over 1,000 square miles over land and sea, the Crown Colony was one of Britain’s smallest. Its size, however, hid the major complexities of the task. The remotest settlements in the rocky, mountainous archipelago could only be reached on foot or by boat. Meanwhile, estimations put the population at possibly three times its last official count. The 1961 census would be the first in 30 years, due to war, occupation and waves of refugee crises exacerbated by the victory of the communists and the end of the Chinese Civil War in 1949. Hong Kong’s already crowded urban areas swelled with squatter villages, populated by refugees speaking mutually unintelligible languages. 

There were other complications, however, that were not so novel. The ‘floating population’, known as sui seung yan (‘on-water people’) in Cantonese, sometimes in derogatory terms the ‘Tanka’, had frustrated official attempts at enumeration since the census began in 1881. Viewed by the local Cantonese population as a separate ethnic group, the sui seung yan had for generations lived offshore on a variety of small boats and junks. Chiefly involved in fishing, they also served as lightermen and provisioners to civilian and military ships in Hong Kong’s busy port. Often spending their entire lives at sea, they were essentially nomadic, but the colony’s natural advantage of deep harbours meant several anchorages within its bounds served as key points for trade and shelter from the typhoons that hit the south China coast.

Despite their lack of permanent mooring, the sui seung yan formed an important part of Hong Kong’s society and its maritime-dependent economy. Furthermore, as common law ensured any child born in the dominions of the Crown, land or sea, was a British national, the government was forced to view them as part of their jurisdiction. Necessity demanded that any count of the population include them, even if this was easier said than done.

This was because the census as conducted in Britain’s imperial domains was not only useful for practical purposes (such as taxation); it was also symbolically important. For Britain to project the image of control across its global empire, it had to demonstrate total knowledge of its lands and inhabitants. The urge to know and count can be seen from the early attempts to find an efficient method of counting the elusive sui seung yan.

From 1901, for example, the colony’s wide, natural harbour was blocked off by launches and divided into sections, mirroring the districts for the land census, each enumerated by a team of water police, seamen and census officials. Once a boat was counted, it was marked with white paint. Officials noted that, while the system was thorough, it was arduous and expensive due to the number of boats and personnel needed (including replacements for enumerators taken seasick). It also took several days, which removed the element of a census being an accurate count of the population on one fixed date.

Officials also complained of the ignorance of the sui seung yan, many of whom were illiterate. In 1921, enumerators were told not to ask for their nationality (instead assigning it depending on birthplace) as, the census report noted, ‘such a question would puzzle extremely the ignorant women of the floating population, who would do all the talking when the enumerator visited their floating home’. Some reported that boat people were confused when it came to their ages, probably due to differences between the Gregorian and Lunar calendars. One official claimed some actively interfered by surreptitiously painting white crosses on boats that had not been counted (for a price).

Such were the issues that the committee for the 1961 census opted to conduct an entirely separate ‘Marine Census’ just for the floating population. This included a pilot census, which partnered with UN advisers and representatives of the sui seung yan themselves. Apart from the use of new red census labels in place of white paint, the officials introduced aerial photography from helicopters or higher ground. With this, boats could be counted, identified and compared against numbers taken from enumerators on the water. Census Commissioner K.M.A. Barnett pronounced it an ‘almost infallible’ safeguard for census accuracy.

Aerial photography was used again in the 1966 by-census and the next decennial census in 1971. In addition to the astonishing visual impact of surviving photographs, the impact of technology in making the sui seung yan and their boats look contained and manageable is palpable – especially compared to the fears of prewar officials that boats could easily slip away unseen and uncounted.

What is notable is that officials across the decades all commented on the willing co-operation of the boat people. It is telling that there are virtually no recorded incidents of outright refusal to co-operate, in comparison to some more belligerent villages, whose representatives had to be convinced by senior officials. Though it is difficult to ascertain solid reasoning for this, we know that radio broadcasts in 1961 and 1971 often extolled the importance of the census in helping the government plan social policy. The published report of the 1961 census indicated the boat people as the community most in need of provisions for primary education. True enough, in the following decades, the Education Department moved to establish government schools accessible to boat children in areas near typhoon shelters. It also funded private initiatives, such as the Po Kwong School, itself located aboard a boat.

But there may be other, deeper reasons behind the sui seung yan’s agreeing to take part in the colonial census. Historically a denigrated population in China, they were barred from posts in the imperial bureaucracy and certain gentry functions. Historians have highlighted this as an explanation for their willingness to act as informers and smugglers for the British during the First Opium War. Their small, inconspicuous boats made them a crucial early ally in the capture of Hong Kong itself.

At the earliest public land auctions held after the establishment of the colony, entrepreneurs from the community, such as Loo Aqui, were some of the first to invest their money in the foundation of the city. In the final prewar census of 1931, boat people were significantly more likely (17.2 per cent) than the urban land population (2.3 per cent) to assert that they were British nationals rather than Chinese, second only to rural landholders in the New Territories (35.4 per cent).

From reports it is obvious that the relationship between census officials and the sui seung yan cannot simply be characterised as one of oppression or collaboration. The tensions around the census instead indicate the nuanced nature of power and autonomy between the colonial state and a semi-nomadic population. It reminds us of the many complexities that existed within colonialism, especially in the choices of indigenous peoples to participate or withdraw. 

Phyllis Chan is a PhD student based at the University of Bristol researching the history of Hong Kong.