Local Versus National Identity in Hong Kong, 1998–2017

ABSTRACT

In understanding why the proportion of Hong Kong people whose local identity overshadows their national identity has been increasing in the past decade, two perspectives have been considered: the birth cohort and the periodic perspectives. However, because previous studies have been based on cross-sectional surveys, they have failed to examine both cohort and period effects on trends in identification in Hong Kong. A cross-sectional survey approach has two limitations: (i) an inability to distinguish the net effects of cohort from that of age, in that there is an exact linear dependency between age and cohort; and (ii) a lack of information on periodic variables. In order to overcome these limitations, this article is based on a pooled dataset of longitudinal surveys combined with official statistics from 1998 to 2017 and employs an age-period-cohort analysis. It is found that both cohort and period effects have contributed to the strengthening of local identity in the past decade. Those born in the 1980s or later are more likely to identify as Hongkongers than as Chinese. An influx of tourists from mainland China and a decrease in satisfaction with the central government have also contributed to the rise of a local over a national identity.

KEY WORDS:

• National identity
• localism
• cross-border integration
• Hong Kong
• age-period-cohort analysis
• multilevel analysis

Notes

1. Between 1997 and 2002 the proportion of people who reported they were satisfied with China’s policy towards Hong Kong was always higher than those who reported they were dissatisfied, while public satisfaction with the local government decreased from 54.3% to 11.7%. In the next several years, following an increase in the public’s satisfaction with the local government, the public’s views of China’s policy towards Hong Kong became more positive – those who expressed satisfaction increased from 30.9% to 53.5%, while those who expressed dissatisfaction decreased from 24.6% to 5.5% (Hong Kong Institute of Asia-Pacific Studies 2019).

2. Cohort effect and age effect are confused because there is an exact linear dependency between age and cohort (cohort = period – age). These two effects refer to different explanatory mechanisms. The age effect describes the change in an individual’s attitudes and behaviours over his/her life cycle. The cohort effect describes distinctive characteristics of people who were born and grew up in the same years, and these characteristics will not change over their life cycle (see Glenn 2003; Yang and Land 2013).

3. In a cross-sectional survey, subjects are observed at one time point. This makes it possible to test the association between an independent and a dependant variable at a specific point in time. For example, in examining whether specific groups in the population have a stronger or weaker local identity than other groups. However, in a cross-sectional survey subjects are not observed at various time points, so information on changes in the socio-economic environment and identity after the handover could not be collected using a cross-sectional survey. Nor could it be used to test whether specific changes in the socio-economic environment are associated with the rise of a local identity.

4. In the 2000s, to attract capital and talent, the Hong Kong government set up some new programmes for mainland residents, such as the Capital Investment Entrant Scheme and the Quality Migrant Admission Scheme.

5. Many mainland pregnant women whose husbands were not Hong Kong permanent residents had made use of the IVS to give birth in Hong Kong. The number of these “doubly non-permanent resident babies” increased from 620 in 2001 to 35,736 in 2011 (Census and Statistics Department 2012, 20). These children also attended schools in Hong Kong. The result was claimed to be shortages of hospital beds for local pregnant women and of places in kindergartens and primary schools for local students.

6. Landlines were used because it is a longitudinal telephone survey research project started in 1998 when mobile phones were less common.

7. The category “more as a Chinese” includes those who answered “half-half.” It is inappropriate for the category “more as a Hongkonger” to include those who answered “half-half,” because the article aims to explain why there is an increasing number of people whose local identity overshadows their national identity and those who answered “half-half” do not have stronger local identity than national identity. Although the survey did not provide the respondents with the option of choosing “half-half,” about 10% of the respondents insisted on giving the answer that their national identity was as strong as their local identity. We replicated the tests with the sample, excluding those who answered “half-half,” and obtained consistent findings. Hence, those who answered “half-half” are included in the analysis.

8. The level of control that Beijing exerts over local newspapers is an indicator of the degree to which Hong Kong is autonomous.