1. For example, the system of street offices is set up in the cities with a population of more than 100,000. See the “Regulations on Organizing Urban Street Offices,” approved in the Fourth Meeting of the Standing Committee of the National People's Congress on 31 12 1954Google Scholar. Compared with counties which belong to the rural category, cities are permitted to retain more revenue for management and construction and have greater power to approve economic projects.

2. Adjacent suburban districts and distant suburban districts may contain towns. When a town is located within a city's administrative boundary, the population of the town is statistically included in that city's population. See Table 1.

3. Actually, these criteria were partially revealed by Chandrasekhar, in his book, China's Population: Census and Vital Statistics (Hong Kong: Hong Kong University Press, 1959), p. 46Google Scholar, written following a visit to the higher officials of the State Statistical Bureau in Beijing in 1958.

4. See State Council, “The directive for designation of cities and towns,” in Gazette of the State Council of the People's Republic of China, No. 17 (1955), p. 847Google Scholar. State Council, “The stipulations for the demarcation of urban and rural areas,” in Gazette of the State Council of the People's Republic of China, No. 20 (1955), p. 988.Google Scholar

5. There were four constitutions in the history of the People's Republic of China published respectively in 1954, 1972, 1979 and 1982. Article 53 of the 1954 constitution stated: “The administrative regions of PRC can be divided as follows: (1) The whole nation is divided into provinces, autonomous regions, and cities directly under the central government; (2) Provinces and autonomous regions are divided into autonomous zhou, counties, autonomous counties, and cities; (3) Counties and autonomous counties are divided into townships, minority townships, and towns.” This was revised several times in the later versions of constitutions, with townships changing to communes and communes changing back to townships. Article 30 of Chapter 1 in the 1982 constitution also says that municipalities directly under the central government and large cities are divided into districts and counties.

6. State Statistical Bureau, Instructions for the Key Points on the Demarcation of Urban and Rural AreasGoogle Scholar, SSB Internal Document (1955).

7. According to the 1963 criteria, cities that still had a clustered population of 100,000 or more normally could retain their designated city status. Settlements with less than 100,000 persons did not retain their city status unless they were politically and economically important enough. All other cities that did not meet the above conditions lost their city status. For details, see the associated documents listed in Table.

8. See Xuwei, Hu, “Dui woguo chengzhenhua shuiping de pouxi” (“Analysis of the urbanization level in China”), Chengshi guihua (Urban Planning), No. 2 (1983), p. 26Google Scholar. Also see Ma, Laurence J. C. and Cui, Gonghao, “Administrative changes and urban population China”, Annals of the Association of American Geographers, Vol. 77, No. 3 (1987), p. 376Google Scholar, and Zhou, Yixing, “New trends of Chinese urbanization in the 1980s,”Google Scholar in Yeung, Yue-man (ed.) Zhongguo chengshi yu quyu fazhan: zhanwang ershiyi shiji (Urban and Regional Development in China: Prospects in the 21st Century) (Hong Kong: Hong Kong Institute Asia-Pacific Studies, The Chinese University of Hong Kong, 1993), pp. 105–131.Google Scholar

9. Central Committee of the Chinese Communist Party, State Council, “Directive for adjusting criteria on designation of cities/towns status and reduction of suburban areas cities,” issued on 17 12 1963.Google Scholar

10. This administrative arrangement was extended to a town level when villages were put under the leadership of towns. The problems which arose were similar to those in counties under the jurisdiction of cities.

11. Ziyang, Zhao, “Current economic situations and the direction of future economic development”, BSC, No. 26 (1981), pp. 816–850.Google Scholar

12. Ma, Laurence J. C. and Cui, Gonghao, “Administrative changes and urban population in China,” pp. 373–395Google Scholar; Shanxin, Pu, Zhongguo xingzhen quhua gailun (Overview of China's Administrative Divisions) (Beijing: Knowledge Publishing House, 1995)Google Scholar. Also, Zhou, Y. X. and Shi, Y. L., “Toward establishing the concept of physical urban area in China,” The Journal of Chinese Geography, Vol. 5, No. 4 (1995), pp. 1–15.Google Scholar

13. In 1979, the Zhejiang provincial government approached the State Council for permission to include some counties as the parts of cities. From the local experience, the Zhejiang government argued that a county under the jurisdiction of a city could (1) improve the co-ordination between urban and rural economic activities; (2) facilitate rural growth by letting rural areas access urban financial sources; (3) reduce administrative expense by reducing administrative staff. In 1981, Zhejiang's request was approved by the State Council. Based on the Zhejiang case, the Ministry of Civil Affairs and Ministry of Labour and Personnel issued a directive in 1983 to guide the reforms of administrative structure and counties under the jurisdiction of cities. See Ministry of Civil Affairs and Ministry of Labour and Personnel, The Requested Report to the State Council Concerning Several Major Questions of Reform on the Structure of Administrative Setup, issued on 18 05 1983Google Scholar. After that, most provinces with the permission of State Council restructured their spatial administrative systems where all counties were placed under the administration of cities. In 1986, State Council issued another document to formalize the conditions of counties under the jurisdiction of cities. See State Council, Circular for Approving the Report Submitted by Ministry of Civil Affairs on Adjusting the Criteria on City Designation and the Criteria on Counties under Jurisdiction of Cities, issued on 19 04 1986Google Scholar. Up until the late 1980s, all provinces had some of their cities undergo similar administrative adjustments. There were 30% of the counties under the administration of cities nationwide, and 90% of cities had brought at least one county under their control.

14. Detailed examination of the political and economic impacts of cities leading counties can be found in Zhao, X. B., Systemic Restructuring for Regional Development in China, (Ph.D. dissertation, University of Manchester, 1992)Google Scholar, but is beyond the scope of this paper.

15. The criteria prior to 1993 were criticized for having two major drawbacks. First, the urban features of a settlement were not evaluated thoroughly enough. By only using a proportion of non-agricultural population and a gross value of industrial output to assess standards, the characteristics and performance of urban economies were not sufficiently taken into account. Moreover, urban infrastructural environment was not given any consideration at all. As a result, many counties, although economically dominated by agriculture and less developed, were conferred a city status. Secondly, there were no criteria for promoting cities from the county-level to the prefecture-level. In practice such promotion was mainly based on the personal judgement of officials who were in charge of examination of promotion applications, rather than on standardized criteria. Realising that the prior criteria were too loose to guide the designation process, on 8 February 1993, the Ministry of Civil Affairs presented a report to State Council requesting a revision of city designation criteria. The State Council approved the report and circulated it nationwide for implementation on 17 May of the same year. Source: authors' interview with officials from Ministry of Civil Affairs in June 1996, Beijing.

16. Urban household registration was carried out in late 1951, with the application of the Ministry of Public Security's interim regulations on urban household administration issued on 16 July 1951. At that time, urban hukou holders were people who were registered to reside in cities and towns. Urban hukou was not related to the government food supply and job assignment and thus was not subject to very strict control. Rural populations were not required to participate in hukou registration before 1955.

17. State Council, “The directive for designation of cities and towns,” p. 847Google Scholar; State Council, “The stipulations for the demarcation of urban and rural areas,” p. 988Google Scholar; State Council, “Directive concerning the establishment of regular household registration system,” Gazette of the State Council of the People's Republic of China, No. 10 (1955), p. 539Google Scholar; State Council, “Regulations on household registration in the People's Republic of China,” Gazette of the State Council of the People's Republic of China, No. 2 (1958), p. 45.Google Scholar

18. For discussion of the concept of agricultural and non-agricultural population based on household registration, see Kirkby, R. J. R., Urbanization in China: Town and Country in a Developing Economy 1949–2000 AD. (London: Croom Helm, 1985)Google Scholar; Ma, Laurence and Cui, Gonghao, “Administrative changes and urban population in China,” pp. 373–395Google Scholar; Chan, , Wing, Kam and Yuen, Tsui Kai, “Agricultural” and “Non-agricultural” population statistics of the People's Republic of China: Definitions, Findings and Comparisons, Occasional Paper No. 1, (Hong Kong: Department of Geography and Geology, University of Hong Kong, 1992)Google Scholar. Also Zhou, Yixing and Young, Qi, “Woguo chengzhen dengji tixi biandong de huigu jiqi shengqu diyu leixing” (“Review of change in the urban system and the provincial patterns”), Dili xuebao (Geographical Journal), No. 41 (1986), pp. 97–109.Google Scholar

19. Please note that much of the published data on population, with the exception of the census (including that from the State Statistical Bureau and many academic institutes), are originally from the Ministry of Public Security, no matter who is the publisher. One must be careful to check the source of the data. In many cases the source is mentioned in the preface or postscript and does not exactly follow the specific tables.

20. Central Committee of the Chinese Communist Party and State Council, “Directive for adjusting criteria on designation of cities/towns status and reduction of suburban areas of cities.”Google Scholar

21. See State Council, “Procedures for the third national population census,” Gazette of the State Council of the People's Republic of China, No. 5 (1982), p. 154.Google Scholar

22. In practice, how and to what extent the second sense of the NPCT indicator is applied is still an open question. In our field study and interviews with the statistical officials, we were informed that, although now required by the Central Government, this new concept of non-agricultural population had not been fully applied to date. Due to the lack of effective mechanisms and sufficient resources for the registration of immigrated labour, it is estimated that only half of the non-agricultural population, defined by occupation, can be registered in designated towns. In cities, the concept of non-agricultural population defined by the hukou status is still exercised. Indeed two definitions are actually interchangeably applied to the data of non-agricultural population in the official statistics without giving detailed explanations. In cities, the “hukou” definition is applied. In towns the “occupation” definition is practiced, although there are omissions. This situation illustrates the need for a statistical adjustment, as will be discussed later, to give a more precise estimation of urban population.

23. For example, the figures of the NPCT from the Ministry of Public Security's source refer to the hukou concept, whereas the statistics from the State Statistical Bureau may refer to the hukou concept in some cases and the occupational concept in other cases. In the census of 1982 and of 1990, the non-agricultural population of cities and towns referred to hukou status.

24. The concept of de jure population in the 1990 census was defined as one whose hukou was registered in the enumeration place, or one who was living in the enumeration place for one or more years; or one who was absent in one's hukou registration place for one or more years. See State Council, “Procedures for the fourth national population census,” Gazette of the State Council of the People's Republic of China, No. 22 (1989) p. 809.Google Scholar

25. Ibid.

26. In China, cities or towns with a population of less than 50,000 are not legally required to set up residents' committees. See “Regulations on organizing urban street offices,” approved in the Fourth Meeting on 31 12 1954Google Scholar, by the Standing Committee of National People's Congress. Therefore, the 1990 second-set urban population definition excluded the people of cities and towns with populations of less than 50,000. Also, according to the “Interim regulations for the administration of urban temporary population,” temporary urban populations were required to register in urban areas and obtain certificates of temporary residence. However, it was never clear whether or not they were to be included in the population lists of the residents' committee where they lived. See Ministry of Public Security, “Interim regulations for the administration of urban temporary population,” Gazette of the State Council of the People's Republic of China, No. 26 (1985), p. 908Google Scholar. Also, much of this temporary population did not register at all. According to our field surveys of temporary population in several settlements in Guangzhou city and the Zhujiang Delta region, only 20% to 30% of the temporary population had registered and obtained certificates of temporary residence.

27. Macura, Milos, “The influence of the definition of the urban place on the size of the urban population,” in Gibbs, Jack P. (ed.), Urban Research Methods (D. Van Nostrand, 1961), pp. 21–31.Google Scholar

28. Petersen, William and Petersen, Renee, Dictionary of Demography: Terms, Concepts, and Institutions (Greenwood Press, 1986).Google Scholar

29. Smailes, A. E., “The definition and measurement of urbanization,” in Fones, R. (ed.), Essays on World Urbanization (London: George Philip and Son, 1975), pp. 1–18.Google Scholar

30. Tan, K. C., “Revitalized small towns in China,” Geographical Review, No. 76 (1986), pp. 138–148CrossRefGoogle Scholar. Tan, K. C., “Small towns in Chinese urbanization,” Geographical Review, No. 7 (1986), pp 265–275CrossRefGoogle Scholar. Lee, Yok-shiu, “Small towns and China's urbanization level,” The China Quarterly, No. 120 (1989), pp 771–786.CrossRefGoogle Scholar

31. Ma, Laurence J. C. and Cui, Gonghao, “Administrative changes and urban population in China,” pp. 373–395Google Scholar. Wu, Xiaoying, “Rural to urban migration in the People's Republic of China,” The China Quarterly, No. 139 (1994), pp. 669–698.CrossRefGoogle Scholar

32. Wu, Xiaoying, “Rural to urban migration in the People's Republic of China,” pp. 669–698.Google Scholar

33. Ma, Laurence and Cui, Gonghao, “Administrative changes and urban population in China,” pp. 373–395.Google Scholar

34. In many cases this is a major reason for local leaders to push for city designation.

35. Since 1980, all large and some medium-size cities have been allowed to use 5% of the total net profit tax paid by the city's industrial and commercial enterprises for city development and construction projects. Detailed discussion on cities' benefits over other settlements was presented by Pu Shanxin, an official of the Ministry of Civil Affairs, at the seminar on administrative districting and urban development in China on 17 April 1996 the Centre of Urban Planning and Environment Management, the University of Hong Kong.

36. Ma, Laurence J. C. and Cui, Gonghao, “Administrative changes and urban population in China,” pp. 373–395.Google Scholar

37. Kirkby, , Urbanization in China.Google Scholar

38. For example, as Kirkby, (Urbanization in China)Google Scholar pointed out, the objective of Beijing's further northward extension in the early 1960s was direct control over several large reservoirs. Also, the territories almost 100 kilometres to the south of the city-proper were deemed part of the heavy-industrial and armaments centre of Chongqing in order to ensure supplies of non-ferrous ores.

39. Zhao, X. B., Systemic Restructuring for Regional Development in ChinaGoogle Scholar; Zhao, X. B. and Zhang, L. (forthcoming), “Fiscal decentralization and regionalism in China,” International Regional Science Review.Google Scholar

40. Solinger, D., “Uncertain paternalism: tensions in recent regional restructuring in China,” International Regional Science Review, Vol. 11, No. 1 (1987), pp. 23–43.CrossRefGoogle Scholar

41. A review of the government documents on the administrative divisions published in the Gazette of the State Council reveals that there is no nationwide criteria to determine a municipal jurisdiction. The adjustment of an urban administrative boundary is normally first appealed for by the local authorities based on their own practical reasons. After a series of administrative procedures, the appeal will finally be ruled on by the State Council based on individual circumstances.

42. Kirkby, , Urbanization in ChinaGoogle Scholar. Ma, Laurence J. C. and Cui, Gonghao, “Administrative changes and urban population in China,” pp. 373–395Google Scholar. Chan, Kam Wing, Cities with Invisible Walls: Reinterpreting Urbanization in Post-1949 China (Hong Kong: Oxford University Press, 1994).Google Scholar

43. Kirkby, , Urbanization in ChinaGoogle Scholar; Chan, Kam Wing, “Rural-urban migration in China, 1950–82: estimates and analyses,” Urban Geography, Vol. 9, No. 1 (1988), pp. 53–84CrossRefGoogle Scholar; Chan, Kam Wing, “Urbanization and rural-urban migration in China since 1982,” Modern China, Vol. 20, No. 3 (1994), pp. 243–281CrossRefGoogle Scholar; Wong, Y., “Zhongguo chengshihua de zhenshi shuiping” (“The true level of Chinese urbanization”), Guoji jingji pinglun (Review of International Economics), Nos. 3–4 (1996), pp. 57–61Google Scholar; Chan, Kam Wing and Xieqiang, Xu, “Urban population growth and urbanization in China since 1949: reconstructing a baseline,” The China Quarterly, No. 104 (1985), pp. 583–613.CrossRefGoogle Scholar

44. Chan, Kam Wing, Cities with Invisible WallsGoogle Scholar; Wu, Xiaoying, “Rural to urban migration in the People's Republic of China,” pp. 669–698.Google Scholar

45. Ma, Laurence J. C. and Cui, Gonghao, “Administrative changes and urban population in China,” pp. 373–395Google Scholar. Wu, Xiaoying, “Rural to urban migration in the People's Republic of China,” pp. 669–698.Google Scholar

46. Wu, Xiaoying, “Rural to urban migration in the People's Republic of China,” pp. 669–698.Google Scholar

47. Kirkby, , Urbanization in ChinaGoogle Scholar; Chan, Kam Wing, “Rural-urban migration in China, 1950–82.”Google Scholar

48. The Section 7 of the “Regulation on household registration in the People's Republic of China” stipulates that within one month of the birth whoever registers this birth must register with the mother's household registration unit, where her permanent address is. If the baby is with his/her mother at a temporary residence, then, on returning to his/her mother's permanent residence, the baby must be registered. If the baby is born at the mother's temporary residence, the registration of birth can wait until the mother's return to her permanent abode. See State Council, “Regulations on household registration in the People's Republic of China,” p. 45.Google Scholar

49. After 1984, peasants who opened their own business in market towns and took responsibility for their own grain could obtain non-agricultural hukou in those towns. See the circular of the State Council concerning the matters of peasants settling down in market towns, issued on 13 October 1984.

50. The basic functions and features of the hukou system remained intact after 1980. Since hukou status is politically delineated and is an important way to maintain social control, there is no incentive for the state to fundamentally change it. The classification of citizens as agricultural/non-agricultural based on the hukou registration still exists. Persons holding different hukou still have different political and economic rights and obligations. Transfer of hukou status from agricultural to non-agricultural is still subject to policy and quota controls. Change of permanent residence outside one's residence jurisdiction still requires the certificate of movement. The hukou system per se is not as relaxed as is conventionally perceived. For example, the policy which has allowed peasants to move into small market towns since 1984 is subject to many preconditions, such as seizing qualifications for state food supply, financial subsidy and job allocation. In 1985 the Ministry of Public Security issued another hukou regulation, which stipulated the compulsory registration of temporary population in all urban areas. This was not previously required. On 29 October 1988 the General Office of State Council issued the circular on stopping the overt “sale” of urban hukou in certain cities and counties. When necessary, the government enforces the hukou registration and takes action to send back any “illegal” migrants. One typical example was the recent crackdown on movement of illegal migrants and strict control over in-migration to Shenzhen, as Hong Kong's return to China approached.

51. Chan, Kam Wing and Xieqiang, Xu, “Urban population growth and urbanization in China since 1949,” pp. 583–613.Google Scholar

52. Normally one could assume that the later one is more accurate, as is often claimed in the State Statistical Bureau's publications.

53. We understand that in some urban settlements administrative boundaries are overbounded. This problem will be controlled by some adjustments, as discussed below.

54. It should be pointed out that group A did exclude some “illegal” residents who violated the hukou regulations. As indicated in the pre-census checks on hukou registration, there were some “illegal” urban residents who had managed to escape detection and to survive in urban places. This group of persons were not systemically enumerated. But the Chinese hukou system was closely linked to the citizens' rights and the welfare distribution before the reforms and is still closely linked to some citizens' rights (e.g. the birth quota under the family planning policy). Although it is not perfect, the reliability of the non-agricultural population indicator based on the hukou system is much better than other data sources.

55. Unauthorized migration refers to migration without official permission, such as without moving certificate.

56. In a de facto sense, the effectiveness of the hukou system to regulate migration has been weakened since the late 1970s, not because of the relaxation of the hukou system per se, but because of the abandonment or reform of other incorporated institutions, such as the rationing system and the employment system. Without the co-operative functions of other social and economic institutions, it is hard for the hukou system alone to block migration effectively.

57. Shanxin, Pu, Zhongguo xingzhen quhua gailun (Overview of China's Administrative Divisions) (Beijing: Knowledge Publishing House, 1995)Google Scholar. Wong, Y., “The true level of Chinese urbanization,” pp. 57–61.Google Scholar

58. See footnote 26 for discussion on the census undercount.

59. For example, the temporary population needs accommodation. Although some individuals may rent private accommodation in the town, those “long-term” temporary residents often form stable settlements which are not much different, except for size, from the wellknown “Zhejiang” or “Xinjiang” or “Anhui” ethic villages in the fringe areas of the big cities. According to normal practice in townships, any temporary population living in the town must pay various local governmental fees/taxes, including garbage collection fees, environmental and security fees, and town development fees. These fees, collected from the temporary population, can be used by local governments as town expenditure. Because of the fiscal particularities, township mayors, the tax officials and officers in charge of the temporary population, pay close attention to in-migrants and regularly check the number of the “longterm” temporary residents.

60. The Township Information Database contains the basic information for each township, including the township mayors' names, addresses and contact telephone numbers. The Database was compiled and administered by the Ministry of Civil Affairs.

61. For each township, at least two phone calls were made directly to the township mayors' offices. The first call explained the purpose of the survey and the questions to be asked. The mayors or their associates were asked to prepare and be ready to answer the questions in the next calls, which were made one or two days later. Normally three to five calls were needed to complete a survey. This data was verified before the report was done. In answering the questions, the mayors or their associates were asked to double-check their report figures. In the survey, 10% of townships provided data in the first call as the mayors knew the latest figures. 90% of townships provided the data in the second, third or fourth calls, after a careful check of their records and changes. 35% of the phone calls were directly answered by the mayors, 40% by mayors' secretaries, and 25% by the administrative officers who were in charge of the administration of temporary population. In the latter two cases, both the secretaries and officers were authorized by their mayors to check the data and answer the questions.

62. The experts interviewed include, among others, Professor Zhou Yixing from Beijing University, Professor Gu Chaolin from the Institute of Geography of the Chinese Academy of Sciences, Professor Pu Shanxin from the Ministry of Civil Affairs, planners from China Academy of Urban Planning and Design, as well as some township mayors/chiefs from Guangdong Province and Hunan Province.

63. As shown in Table 6, China's urban population is constituted by parts A, B, C and D, with A (non-agricultural population in city and town proper based on household classification) as a core. The data on the non-agricultural population of each individual city and town were obtained from the Ministry of Public Security. The data were first grouped according to the designated administrative status of cities and towns and then aggregated into provincial and finally regional and national levels. According to the selected K₃ values (Table 9), China's urban population was readjusted by Equation (2) (UP = A X (1 + K₃)). China's urbanization levels were therefore readjusted according to the newly calculated urban population.

64. SSB's 1995 by-census results, see State Statistical Bureau, 1995 1% Sample of Population Census (Beijing: Zhongguo tongji chubanshe, 1996).Google Scholar