Twentieth-century China witnessed devastating wars, radical political revolutions, and, recently, very rapid economic development that lifted hundreds of millions from dire poverty. The economic development profile of this century of revolutions has not been described with consistent and reliable data, leaving open debates on China’s transition to modern economic development. My Ph.D. project identifies jiapu (Chinese lineage genealogies) as a novel, reliable, and consistent source which allows me to reconstruct the changing occupational structure of twentieth-century China.

Jiapu record individual-level information of descent groups with a common ancestor together with biographies detailing the life course of some members. The value of this source for studies of historical demography has been recognised, but the occupational information has not been identified and exploited systematically. My Ph.D. project created, from scratch, the novel Yangtze Jiapu Dataset (YJD) with 208,130 occupational observations from 828 counties in the Yangtze Valley, where more than 40% of the national population lived. Comparisons with censuses suggest that the YJD broadly represent the regional population.

My project marks the first use of jiapu data to quantitatively estimate the occupational structure of the Yangtze Valley from ca. 1920 to the 1982 census, the first census with reliable occupational data. These new estimates and census data suggest that a substantial shift of the labour force out of agriculture only occurred in the three decades from 1990 onwards with the primary sector share decreasing from 74.6% in 1990 to 19.1% in 2020, which is much later than other changes characteristic of modern economic development, such as the increase in per capita income and the emergence of modern factories with high productivity. The occupational structure series also suggest continuities in the increase in industrial employment across the divides of 1949 when the People’s Republic of China was established and 1978 when the ‘reform and opening-up’ started, which were regarded as political and economic turning points. Comparing the occupational structure and GDP structure data shows that neither the period 1950–1978 nor the period 1978–2020 witnessed the structural change that Kuznets observed in many earlier-developed countries, which he suggested was a concomitant of modern economic growth.

The YJD also reveals patterns of by-employment and women’s work. Theories of the division of labour, proto-industrialisation, involution, industrious revolution, and economic resilience suggest that by-employment undergoes a continuous decline, or no substantial change, or a rise and then a decline in economic development, but these hypotheses have not been supported with longitudinal empirical data. The YJD, for the first time, provides direct evidence of the prevalence of by-employment over time, which suggests two surges of by-employment among men born between 1821 and 1990, reflecting influences of specialisation, structural change, and institutional changes relating to land and labour. My Ph.D. project also exploited the microdata of the farm household survey led by John Lossing Buck in the period 1929–1933, which suggests that the male by-employment pattern of the Yangtze Valley was similar to that of Yamanashi, Japan, in 1879, conforming to the Asian pattern proposed by Saito and Settsu, in which the majority of cases were of peasant farmers having subsidiary employment in industry, commerce or transport. The by-employment pattern of women in the Yangtze Valley is slightly different in that women devoted more of their working time to domestic industries than to agricultural production.

The YJD documents the significant expansion in the diversity of women’s work from mostly agriculture and textiles to all occupational categories in the twentieth century, with an increase in the female-to-male ratio for most occupational groups. Data for 42,666 couples show that married women in different sectors and born in different periods had diverse experiences of engaging in the same occupation as their husbands, due to differences in the prevalence of family-based business, marital matching strategies, and political movements. My Ph.D. project also investigated the female labour force participation rate (FLFPR) by combining Buck’s survey 1929–1933, population registers in the 1940s, and censuses in the period 1982–2020. Strikingly diverging from the U-shaped profile found in the industrialisation of Britain and the US, the FLFPR in the Yangtze Valley followed an inverse U-shaped profile from the 1930s to 2020, showing the influences of collective institutions, education, and foot-binding. The geographical variations of the prevalence of foot-binding were not recorded completely in archives, but the ‘embodied’ nature of foot-binding suggests that foot-binding still impacted the participation in market production of women with bound feet in the 1980s. The 1982 FLFPR data, therefore, were used to estimate the geography of foot-binding in the 1930s.

Future research that integrates the macro occupational structures, which this project illuminates, and the micro experiences of people’s working lives, as recorded in jiapu biographies which have yet to be exploited, will achieve a more comprehensive understanding of the economic and social development of twentieth-century China.