Immigration and Earnings Inequalities in Hong Kong, 19962006

1
Immigration and Earnings Inequalities in Hong
Kong, 1996-2006
By LAM Kit-Chun Department of Economics Hong Kong
Baptist University and LIU Pak-Wai Chinese
University of Hong Kong
2
Objectives

• To study the pattern of inequalities by using
  ratios of percentile earnings, as well as their
  changes from 1996 to 2006
• to study the socio-demographic and economic
  characteristics of working individuals in
  different percentile groups, including proportion
  of immigrants, and identify intercensal changes
  in profile.
• To study the different characteristics of
  immigrants and local people as a possible
  explanation for earnings inequalities

3
Objectives

• To study how different socio-demographic and
  economic variables may affect earnings of
  employees in year 2006, and identify changes in
  pattern over time.
• To study the contribution of different factors to
  total difference in log earnings for various
  groups.

4
Data

• The data we use are the 1/14 sample of the 1996
  and 2006 by-censuses.
• We confine our samples to working population of
  age 18 to 60
• our samples do not include imported domestic
  helpers.

5
Measure of inequalities

• We will focus our attention to three earnings
  ratios
• between the 90th and 10th percentiles (90-10
  earnings ratio),
• between the 90th and 50th percentiles (90-50
  earnings ratio),
• between the 50th and 10th percentiles (50-10
  earnings ratio).

6
Percentile earnings ratios
7
Dummy variables on Immigrant Status

• IMMIGOLD is a dummy variable for old immigrants
  who were born in China, have indicated their
  nationality as Chinese, and have resided in Hong
  Kong for seven years or more at the time of the
  census
• IMMIGNEW refers to new immigrants who were born
  in China, have indicated their nationality as
  Chinese, and have resided in Hong Kong for less
  than 7 years at the time of the census
• IMMIGOTH refers to immigrants who do not belong
  to the IMMIGOLD or the IMMINGNEW categories.
• The omitted dummy variable HKBORN refers to
  workers born in Hong Kong.

8
Schooling dummy variables

• SCH2 is a dummy variable for junior secondary
  education,
• SCH3 for senior secondary education,
• SCHC4 for post-secondary education and SCH5 for
  university education.
• The reference group consists of individuals with
  primary education and below.

9
Experience Gender

• EXP refers to the years of working experience of
  the individual computed as (Age years of
  schooling 6),
• and EXPSQ is its square term.
• MALE is a dummy variable for male individuals

10
Industry dummies

• IND2WRR is a dummy for wholesale and retail trade
  and restaurants and hotels
• IND3FIN for financing, insurance, real estate and
  business services
• IND4SER for community and social Services
• IND5OTH for other industries including
  agriculture, forestry and fishing mining and
  quarrying electricity, gas and water
  construction storage and communication.
• The reference group refers to manufacturing
  industry.

11
Characteristics of different percentile
groups Gender Schooling
12
Characteristics of different percentile
groups Age profile
13
Characteristics of different percentile
groups Industrial composition
14
Characteristics of different percentile
groups Immigrant composition
15
Characteristics of different immigrant groups
16
Characteristics of different immigrant groups
17
Estimates of Earnings Regressions for working
population
18
Contributions to differences in log earnings,
90th 10th percentiles
19
Contributions to differences in log earnings,
50th 10th percentiles
20
(No Transcript)
21
(No Transcript)
22
Conclusion Discussions

• It is found that earnings inequality as measured
  by earnings ratio between the 90-10 percentiles
  increases from 5.2 in year1996 to 6.2 in year
  2006, representing a considerable increase of
  more than 19.
• However, the increase in spread comes largely
  from the increase in spread between the 90
  percentile and 50 percentile (an increase of
  13.6)
• while the increase in inequality between the 50
  percentile and the 10 percentile is relatively
  small.

23
Conclusion Discussions

• The gross earnings disadvantage of new immigrants
  relative to local people is about 62 in year
  2006, which is reduced to about 35 when other
  factors like schooling, working experience and
  gender are held constant.
• At the same time, the average year of schooling
  of immigrants (i.e. around 9.7 years) is lower
  than that of people born in Hong Kong (i.e. 11.9
  years).
• The earnings disadvantage of immigrants together
  with their lower education has resulted in their
  over-representation in the 10th percentile
  earnings group.

24
Conclusion Discussions

• Our findings indicate that the gross effect of
  immigrant status can explain 5.9 of the total
  earnings differential between 90th and 10th
  percentile groups,
• which is further reduced to 3.8 when other
  factors are being held constant.
• How can we explain this in face of the large
  earnings differential between immigrants and
  local people.
• The answer lies in the small share of new
  immigrants in the whole population, which is just
  about 2.7 in both 1996 and 2006.

25
Conclusion Discussions

• A far more important contribution to earnings
  inequalities is years of schooling.
• Holding other factors constant, the contribution
  of schooling to the earnings differential between
  the 90th and 10th percentile groups amounts to
  about 30, and about 23 between the 50th and
  10th percentile groups.
• Note that other than the coefficient effect of
  schooling on earnings, part of the schooling
  effect works through the quantity effect due to
  different schooling profile of immigrants.

26
Conclusion Discussions

• It is also found that the earnings differential
  between local people and immigrants of a
  particular education level is different for
  different education levels.
• Besides, if we take into consideration the
  difference in quality of school which has not
  been measured in this paper, the contribution of
  schooling could be even larger.

27
Policy Implications

• Given the important role of the education gap in
  explaining earnings inequalities, there is a
  strong argument for government to adopt a policy
  to enhance the educational attainment of new
  immigrants for the sake of reducing the earnings
  gap between immigrants and local people.

28
Policy Implications

• First, as far as possible new immigrants should
  be allowed or encouraged to immigrate to Hong
  Kong at a pre-school or early school age, so that
  they can benefit from the Hong Kong school system
  which enables them to have a higher return to
  schooling.

29
Policy Implications

• Second, for new immigrants who have passed the
  normal school age when they immigrate to Hong
  Kong
• government should actively promote its menu of
  subsidized job training, re-training and
  continuing education opportunities that are
  available
• This should enhance the marketable skills of the
  new immigrants and improve their relative
  earnings position.

30
Policy Implications

• Third, our results show that controlling the year
  of schooling, new immigrants suffer from a
  sizeable earnings disadvantage.
• one reason could be the difference in culture and
  the lack of soft skills.
• Women are particularly disadvantaged as they are
  often required to wait many years, often idling
  in the Mainland or in Hong Kong with a two-way
  permit, for final approval to come to Hong Kong
  for reunion with their husbands under the one-way
  permit scheme.
• Early approval of OWP for reunion with spouse
  should alleviate the situation.