The Impact of Immigration on the
New Zealand Labour Market
Paper presented at ‘Economic Impacts of Immigration and Population
Diversity, International Workshop’, 11-13 April 2012
Michael Tse & Sholeh A. Maani
The University of Auckland
Economics Department
1
Question & Motivation
• A large segment of the NZ population is
foreign-born (almost a quarter).
• A key policy question is whether or not
immigration affects the labour market
opportunities of the existing workforce?
2
The direction of the impact on existing
workers is dependent on a number of
factors. These include:
• Substitutability between immigrants
and natives. Are immigrants and the
native-born with similar educational
qualifications complete substitutes?
3
Elasticity of substitution
• If immigrants and natives are
substitutes, then the inflow of
immigrants would reduce wages in the
labour market (Borjas, 2003; Orrenius & Zavodny,
2007)
• If immigrants complement native
workers, then we would expect positive
changes to earnings from immigration
(Ottaviano & Peri, 2007; Borjas, Grogger & Hanson, 2008)
•
4
Immigrant education and experience
• The value placed on education and
experience acquired abroad is often
less than the value placed on
domestic education and experience
(Lalonde & Topel, 1991; Duleep & Regets, 2002; Akresh,
2006; Antecol, Kuhn & Trejo, 2006)
5
International literature
• Altonji & Card 1991 and Borjas 2003: 10
% point increase in fraction of immigrants
reduces the wages of less skilled by 3-4
%.
• Card 2005, Addison and Worswick 2002 :
Mare’ and Stillman 2009, no significant
adverse effect
6
Modelling approaches of wage
effects for the native-born
• 1. Spatial approach
(Card, 1990, 2001, 2005; Altonji & Card, 1991; Dustmann, Fabbri &
Preston, 2005).
• 2. Factors of production approach
• (Borjas, et al., 1996, 1997; Jaeger, 2007Leamer, 2000; Orrenius &
Zavodny, 2007, Mare’ and Stillman, 2009).
• 3. National level analysis (skill group)
(Borjas, 2003, 2004, 2005; Orrenius & Zavodny, 2007).
7
Data
• New Zealand Income Survey (NZIS), 2002
to 2007
• This is an individual level data released
under the Confidentialised Unit Record
File (CURF) format.
8
Modelling Approach
• National level analysis based on skill and
work experience categories
• Wage effects of immigrant supply shocks
Extensions:
• We add spatial regional controls
• We incorporate ‘effective immigrant
experience’
9
• Immigrant supply shock:
• Pijt immigrant supply shock
• M (Immigrant), N(Native-born)
•i
educational qualification
•j
experience group
• t year
10
• 4 educational categories:
• No schooling
• School qualification (high-school
completion)
• Post-school
• Bachelor or higher degree
11
Model
• Immigrant shock, fixed-effects and
interaction effects on earnings and hours
worked:
• Pijt
•i
•j
•t
immigrant supply shock
educational qualification
experience group
year
12
13
Index of Congruence
a native-born
b immigrant
c occupation (two-digit)
Borjas (2003), Welch
(1999)
14
15
Results
• Immigrant shock, fixed effects and
interaction effects:
• Pijt
•i
•j
•t
immigrant supply shock
educational qualification
experience group
year
16
19
Spatial Correlation
• each cell is now defined as (r, i, j, t). That
is, each cell is determined by a specific
region, education level, experience group,
and year.
20
21
22
Defining Effective Experience
Let X be the effective experience of an
immigrant worker:
• A age
• Am age at migration
• AT age of labour force entry
23
• We estimate the three coefficients above
in a standard immigrant assimilation
regression of the form:
• Ic = 1
• Id = 1
•N
immigrant entered as a child
if entry as adult
native born
24
a = 0.4 experience overseas conversion
m = 0.7 experience after migration
t = 1.1 experience of child immigrants
25
26
Conclusions
We extend the standard national level
approach to incorporate local government
regions in the analysis.
• We defining groups by region-educationexperience, and it has some impact on the
results, but the effect is small.
27
• We adjust for the value firms place on
experience acquired abroad, and ‘effective
experience’ for each worker.
• Based on this experience framework the
estimates of wage effects continue to be
small.
28