Page 48 - IJPS-9-1

International Journal of
Population Studies Transportation assimilation in Hong Kong

Table 3. (Continued)
Model 1 Model 2 Model 3
Year FE Yes Yes Yes
Place of residence FE Yes Yes Yes
Place of work FE Yes Yes Yes
Place of residence FE×Place of work FE Yes Yes Yes
Place of residence FE×Year FE Yes Yes Yes
Place of work FE×Year FE Yes Yes Yes
R-squared 0.077 0.077 0.077
Observations 169,766 169,766 169,766
Note: FE: Fixed effect; +p<0.1, *p<0.05, **p<0.01, ***p<0.001 (two-tailed tests); The results for occupation and industry are not displayed.
in the likelihood of minibus usage. Figure 3 presents the In sum, our findings support Hypotheses 1, 2, 4, and 5
link between the duration of residence in Hong Kong and partly support Hypothesis 7. Hypotheses 3, 6, and 8
and the predicted probability of taking the minibus by are not confirmed. Table 5 summarizes the results of our
ethnic group. As shown in Figure 3, South Asians have the eight hypotheses.
lowest initial probability of minibus taking, while South-
east Asians have the highest. As the duration of living in 4. Discussion
Hong Kong increased, South-east Asians became even less In general, we confirmed our expectation that immigrants
likely to take the minibus, which shows an opposite trend in Hong Kong had been increasingly picking up the locals’
in minibus usage compared to any other ethnic group. In transportation behaviors during the years spent at the
addition, we found no substantial inter-ethnic difference in destination; in other words, they were more likely to ride
arrival age on minibus taking. The three-way interactions minibuses as one of the top three modes of transportation
show that while age at migration reduces the effect of to work. However, the exact pace of this assimilation
duration of residence in Hong Kong among Chinese who may vary strongly on individual-level factors, which in
immigrated in or after 1997, the moderating effect of general can be grouped into two major categories: (1) The
age at migration is significantly less pronounced among characteristics that result in entry differences which then
Chinese who immigrated before 1997. We also employed continue with the length of stay in Hong Kong, and (2)
a logit model to replicate Model 3 (Table S1), which reveals those that we can only determine the average effects along
similar results to those of the LPM. with the length of stay.
Model 4 tests how Cantonese ability, district-level Age at migration and ethnicity belong to the first
ethnic density, logged personal income, and gender category, whose effects were influential for immigrants’
moderate the effect of duration of residence in Hong minibus taking both at the beginning and in the long
Kong on minibus usage. We found that only logged run. Younger immigrants already had initial advantages
personal income is an effective moderator. The association in minibus taking over older immigrants. In addition,
between the duration of residence in Hong Kong and for certain ethnic groups, for example, recent Chinese
the likelihood of taking the minibus is significantly less immigrants in our case, the advantage of arriving in Hong
pronounced in people with higher incomes, which may Kong at a younger age further strengthened as time passed
result from the fact that rich immigrants tended to swap since younger immigrants demonstrated a faster speed of
public transportation for private cars. According to the transportation adaptation. This also explains why we often
2016 census, among the immigrants who have stayed in observe immigrants arriving at a younger age are often
Hong Kong for 20 years or more, 19.3% of the 10% richest more absorbable of local knowledge, even if they had the
people (i.e., the people with the top 10% income) chose same length of stay in the destination as those who arrived
private cars as their major transportation mode, while
this proportion was only 2.9% for the rest of people. at an older age.
Figure 4 presents the association between the duration Similarly, the seven ethnic groups began at different
of residence in Hong Kong and the predicted probability levels when they first arrived in Hong Kong, with Chinese
of taking the minibus by income percentile. Model 5 is arriving before 1997 at a more advantaged starting point
a simplified model excluding statistically insignificant than the Chinese arriving later and South Asians being the
interactions. most disadvantaged group. However, throughout the years,

Volume 9 Issue 1 (2023) 42 https://doi.org/10.36922/ijps.0386