Docstoc

Exploring the Relationship Between Educational Credentials and the

Document Sample
Exploring the Relationship Between Educational Credentials and the Powered By Docstoc
					    Exploring the Relationship Between Educational Credentials and the Earnings of
                                     Immigrants



     A Paper for the Canadian Sociology and Anthropology Association, London, May 2005




                                  Maria Adamuti-Trache
                             The University of British Columbia

                                                &

                                        Robert Sweet
                                     Lakehead University



                                            Abstract

The study is part of a larger project that uses the 2002 Ethnic Diversity Survey (EDS) to assess
the relative effects of human capital, social structure, and social capital on the economic
integration of immigrants. We limit our analysis to an examination of the relationship between
immigrants‘ educational credentials and earnings. Three aspects of an individual‘s credentials
are considered: country of origin, level of education, and field of study. We extend previous
research that looked at the constraining effects of social structure -- specifically, gender and
visible minority status on immigrant economic integration – on the ability of immigrants to
negotiate the value of their educational credentials in the Canadian labour market.




                                         April … 2005

              Draft copy only: Do not quote without permission (maria.trache@ubc.ca)
                                          INTRODUCTION

Immigration has become an important source of highly skilled labour for the knowledge-based
sectors of Canada‘s economy. Immigration policy since the 1990s has emphasized newcomers‘
education in choosing among candidates (up to 25% of the point-based selection scale). This
resulted in a significantly higher level of educational attainment among recent immigrants as
compared to those from previous immigration waves. ―In 2001, 46% of immigrants aged 25 to
54 who arrived from 1996 to 2000 held at least a bachelor's degree, compared with only 23% of
the same age group who arrived from 1986 to 1990‖ (Statistics Canada, 2003a, p. 86).

A selective immigration system, then, responds to Canada‘s need for highly skilled workers
educated to the post-secondary level. However, from a policy perspective, it not only is
necessary to find and attract well-educated immigrants it also is necessary to integrate them
quickly and efficiently into the workforce in order that their skills may be effectively utilized in
advancing the goals of the economy and society. The task of integration has proven more
difficult than many expected. For instance, in 2001, the employment rate of 25- to 54-year-olds
was only 69% among recent university-educated immigrants (i.e., those who immigrated from
1996 to 2000) as compared to 90% for native-born (Statistics Canada, 2003a).

Recent immigrants with university degrees earned about 31% less than their Canadian-born
counterparts, whether or not they worked in highly or lower skilled jobs. For instance, in
management, ―men aged 25 to 54 who immigrated during the 1990s, and held a university
degree, earned between 50 and 60 cents for every dollar earned by their Canadian born
counterparts‖ (Statistics Canada, 2003b, p. 13). Employment discrepancies and earning gaps
experienced by immigrants who arrived in the latest decade are maintained over longer periods
of time as compared to previous waves of immigration. Whether or not this is related to
changing demands of the Canadian labour market or to an increased competition with
Canadian-born highly educated population is subject to debate. The assessment of the
Conference Board of Canada concludes that ―Immigrants have lower incomes due to transition
difficulties; insufficient working knowledge of English and French; inadequate recognition of their
educational credentials; and, possibly, discrimination‖ (The Conference Board of Canada, 2004,
p. 15). In this paper we argue that even in the case of immigrants who have overcome the
transition period and are active in the labour market, income levels are largely controlled by
social structures, language skills advantage and educational credentials; and there are few best
scenarios ensuring that earnings of immigrants become comparable to those of Canadian-born.


                                                                                                       1
Background
Previous research, directed toward understanding the earnings gap between immigrants and
native-born Canadians, has adopted different theoretical stances. Broadly, these comprise
human capital theory, social structural theory, and social capital theory.


Human capital theory assumes that investment in education is rewarded by increased earnings
and improved working conditions. The demands for a highly-educated workforce have raised
the education premium, and university education receives particular recognition in the labour
market (Lavoie and Roy, 1998; Baldwin and Beakstead, 2003). Between 1991 and 2000,
university graduates registered a 3-6% lower unemployment rate and up to 50% higher average
earnings as compared to all educational levels groups (Statistics Canada, 2003c, 2003d). And,
in general, higher levels of education are associated with enhanced earnings (Finnie, 2001;
Allen, Harris and Butlin, 2003; Heisz, 2003; Morissette, Ostrowsky and Picot, 2004). Human
capital theory represents a useful framework within which to examine the effects of educational
credentials, work experience, and language skills on labour market success. It is, however,
unable to provide a comprehensive account of earnings inequalities in general and immigrant
and native-born earnings differentials in particular (Li, 2003). Other social and structural
characteristics of the labour market impinge on the relationship between educational credentials
and earnings (Beach & Worswick, 1993; Kunz, Milan & Schetagne, 2000). Hiebert (1997), for
example, argues that labour markets are never neutral and the value of an individual‘s
credential is distorted by gender, social class, race, and nativity. While social structures
constrain and limit the value of immigrant educational credentials there remains some scope for
individual agency in the process of negotiating the worth of one‘s educational credentials. Social
context in this instance involves dimensions of identity, attachment and trust among immigrant
groups and between immigrants and members of the host society (Kunz, 2003). Anisef, Sweet
& Frempong (2002) note the possession of cultural capital is equally important. This includes a
familiarity with local labour market conditions and the understanding of social codes and modes
of interacting in achieving job search and career advancement goals.




                                                                                                  2
                                      LITERATURE REVIEW


Previous research on the experiences of immigrants in the Canadian labour market clearly
indicates the presence of social structural barriers. Despite a system of immigrant selection and
entry based on educational qualifications and workplace skill the integration of immigrants into
the workforce is impeded by discrimination and systemic barriers, especially gender. The
complex intersection of social structural features and their impact on immigrant integration in the
1980‘s and 1990‘s was documented by Boyd (1992) and Davies and Guppy (1998). The basis
for the continuing decline in earnings of immigrants through that decade (and to the present)
has been examined in several studies.


Immigrants enter a labour market that is already gendered and systemically biased (racist).
Access to and success in highly-rewarded occupations is overtly or covertly controlled by social
structural features (Finnie, 2001; Anisef, Sweet & Frempong, 2003.) Jobs that are worth more in
the current labour market are more likely to be open to privileged social groups. Beakstaed and
Vinodrai (2003) classify knowledge-based occupations into several professional, management
and technical groups, for which they show that the rate of growth doubled between 1971 and
1996 and wages were more competitive. Comparisons across various occupations and over
time exhibit pronounced gender differences, women‘s participation in knowledge-based
occupations showing a growth factor of 1.3 as compared to 2.3 for their male counterparts.
Findings also confirm that knowledge workers are characterized by higher levels of education.
By 1996, more than 90% of knowledge occupations required some post-secondary education
and, more than 50% of knowledge workers had university degrees as compared to 34% in 1971
(p. 20).


It is argued that immigrants‘ earnings in the Canadian labour market are lower than native-born
Canadians because of the lower market value attached to their educational qualifications, and
not necessarily to a lack of formal recognition of credentials. Li (2001) explores the market worth
of foreign credentials and whether immigrants with Canadian degrees have earning parity with
Canadian-born degree holders. He uses 1996 Census data and compares 16 groups of
subjects differentiated by gender (2), visible minority status (2), and nativity/age at immigration
(4). Immigrants are divided in 3 groups based on age at immigration that is related to education:
Canadian education (immigrated between ages 0-12), mixed education (immigrated between
ages 13-24), foreign education (Immigrated age 25 and over). His study suggests that gender,



                                                                                                      3
race, immigrant status, and type of credentials create multiple sources of inequality. These
combine to marginalize some groups more than others: immigrant women of visible minority
origin are particularly disadvantaged.


While social structures obviously constrain immigrant integration, employers do take into
account the human capital and personal characteristics of their (potential) employees. Studies
of credential recognition and skill utilization indicate that job entry and career advancement
depend to a significant extent on individual differences in language, work experience, and
educational credentials. However, on average, immigrant graduates are less well remunerated
than native-born graduates who possess equivalent educational qualifications (Picot and Hou,
2003) and are unlikely to have their earnings converging to the income levels of native-born
Canadians (Frenette and Morissette, 2003). Attempts to explain these disparities have met with
limited success for reasons associated with the specification of the credentials themselves and
with the inability to identify the social structural factors and individual differences that qualify this
relationship. Work experience gained in the Canadian labour market and official language
proficiency (English or French) are individual factors that make credentials more attractive.


1. Language competence and work experience
Research includes language proficiency as a predictor of immigrant earnings, mainly in
conjunction with work experience and educational credentials. Even if language barriers are
definitely an essential component of social and economic integration, there is no clear evidence
pointing to its supremacy on educational credentials. The immigration policy of 1990 has
assigned up to 24% of the point-based selection scale to language proficiency. First, it has to be
acknowledged that when combined with higher levels of education, knowledge of English
language becomes implicit, since many educated immigrants possess foreign language skills
and the ability to improve quickly. Second, language proficiency applies to the primary applicant.
That allows candidates to select the family with the best chance of meeting the Skilled Worker
Selection Grid criteria. The changing characteristics of recent immigrants (i.e., shift in source
origins and home language) that are less likely of being English-speaking groups is a factor
Picot (2004) identifies as explaining the earnings gap between recent immigrants and
Canadian-born and the rise of low-income rates among immigrants over the last two decades


The immigration policy of 1990 has assigned up to 21% of the point-based selection scale to at
least 4 years of work experience in sectors recognized by the National Occupational



                                                                                                        4
Classification (NOC). Yet, research speaks about skill-underutilization and discounting the prior
work experience of immigrants as a major shortcoming of their effective economic integration.


Reitz (2001) looks at how immigrant disadvantages are decomposed into portions related to skill
utilization (assuming equal skill quality), and pay equity (assuming equal skill quality and
utilization). Reitz uses census Canada 1996 data to estimate regressions of annual earnings for
immigrants and native-born, men and women, considering work experience, as well as
language knowledge, minority origin, place of residence, years since immigration. Next, he
develops models in which occupational categories are considered that allow identifying
immigrant disadvantages within occupational levels which are related to pay inequity. The
comparison between the two sets of models is used to identify the disadvantage component
associated with inequality in access to occupations. Reitz concludes that immigrants receive a
smaller earnings premium for work experience, compared to the native-born, and that
immigrants‘ country of origin has an impact on earnings.


Aydemir and Skuterud (2004) use the microdata files of the Canadian Censuses between 1981
and 2001 and select equal size samples of immigrants and Canadian-born males, 18 to 54 year
old, who worked full-year, full-time in the income reference year to predict weekly wages. Years
of labour market experience (age minus total years of schooling minus 6) are considered
together with years of schooling, cohort period, years since immigration. This allows analysis of
differential returns of earnings for immigrants and Canadian-born. Their first finding is that about
one-third of the overall deterioration in the entry earnings of immigrant cohorts is explained by
the irrelevance of foreign labour market experience that is observed especially among
immigrants from non-traditional source countries. Another one-third of the deterioration is due to
immigrant regions of origin which may actually result in weaker language proficiency.


2. Educational Credentials
Educational credentials that combine highest level of education with the field of study -- and
thus occupation in which immigrants are expected to work -- are the most relevant points to
successfully meet the immigration policy criteria, With the intent of reviewing policy issues
related to immigrants‘ skills, Reitz (2005) looks at the economic, social and political implications
of the education-work relationship and the institutional change needed to reverse the decline in
immigrant employment outcomes. Census 1996 data reveal earnings differences between
Canadian-born and recent immigrants that the author associates with skills underutilization.



                                                                                                       5
Socio-economic and political implications are obvious: low earnings keep many immigrant
families at dangerous levels of poverty, minority groups being even more vulnerable ―Although
educational credentials among recent immigrants have been higher on average than those of
Canada‘s native-born workforce and are rising, and despite the fact that recent immigrants‘
levels of fluency in one official language have not changed, the trends in immigrants‘
employment and earnings are downward‖ (p. 3). The author concludes that the problem is not
the skill levels, but the way they are utilized in the Canadian workplaces. Since these trends
differ by occupation (i.e., immigrants‘ relative success in the professions as compared to
managerial occupations), it suggests that the skill-assessment process is affected by
institutional procedures. The more rigorous credential-assessment practiced in the professions
is advantageous for immigrants. The author recommends institutional change that challenges
institutional complexity (i.e., employers, licensing bodies, unions, post-secondary institutions,
credential assessment providers, etc), timing (i.e., employers are less pressured to consider
change) and racial attitudes. The proposed institutional change is meant to increase the
capacity to utilize immigrants‘ skills in the context of expanding knowledge occupations, since
the non-recognition of foreign education and work experience is estimated to cost the Canadian
economy about $2 billion annually (Reitz, 2005, p. 18).


To adequately discuss the education-work relationship, it is essential to define education
credentials (as assessed by potential employers) beyond merely the counting of ‗years of
schooling‘ (Ferrer and Riddell, 2004). In the case of post-secondary graduates, at least 3
dimensions are analyzed in the literature: origin, or the country in which the degree was
obtained; educational level, typically involving a distinction between college and university
credential or, among university graduates, a distinction between baccalaureate and master‘s
degrees; and field of study, where liberal arts and vocational or professional fields typically are
contrasted.


2.1 Origin of education
A challenging aspect of the socio-economic integration of immigrants is the efficient utilization of
their knowledge and skills in ways useful to the individual and the Canadian employer.
Thompson (2000) advances the hypothesis that if immigrants cannot find employment
appropriate to their education, skills and experience this can be explained either by a lower level
or quality of their education or by the partial compensation that employers give to their
credentials. Since data do not allow one to distinguish among these causes, the study aims at



                                                                                                      6
identifying fields in which ―occupational skill differentials are most acute‖ (p. 9) looking at the
immigrants who came to Canada in early 1990s – a group that has experienced harsh
conditions in finding employment. The substantial increase of the highly educated immigrants
between 1990 and 1999 (45% as compared to 20%) and the change in the composition of
immigrants by country of origin are determinant elements of the socio-economic integration of
immigrants in the 1990s. Thompson‘s findings confirm that immigrant occupational skill
outcomes vary by region with preferential employment to Canadian-born, Northern Europe and
United States immigrants. Being a member of a visible minority group, having a lower
educational attainment, being a recent immigrant would decrease the probability of finding
employment in a highly skilled occupation.


2.2 Educational level
Sweetman and McBride (2004) use census data focusing on post-secondary field of study, and
sex, level of education, and whether the education was completed in Canada or elsewhere in
predicting earnings. The sample includes all permanent residents between the ages of 25 and
65 residing in one of the 10 provinces, not attending school and who have more than 6 months
since immigration. To estimate earnings, all individuals who worked and earn some money in
the reference year are included to better reflect the relevant population, not only those with full-
time full-year jobs. Three educational levels (college or trades certificate, Bachelor‘s or Master's
degree are considered in the analysis. While earnings differences generally paralleled level of
education, these in some cases depended on field of study


2.3 Field of study (FOS)
Existing studies frequently have defined credentials in terms of their level, typically years of
education. Attempts at describing the nature of the credential have distinguished vocational
from academic, college from university, or baccalaureate from post-graduate. Relatively few
studies have considered field of study (FOS) although this represents an important indicator of
the type of skill possessed by an individual and is closely associated with specific occupations
or occupational categories (Boothby, 2000; Finnie, 2001; Anisef, Sweet & Frempong, 2003).


The effect of field of study on immigrant earnings deserves attention because it gives
information on the type of skills immigrants possess and allows one to understand which skills
are easily transferable in the Canadian labour market; or which type of knowledge favour an
easy adjustment to the new Canadian economy. In Sweetman & McBride‘s (2004) study The



                                                                                                       7
distribution by field of study shows that the female and also male Canadian-born groups are
very likely to have a teaching degree while immigrants are much more likely to be in engineering
and applied sciences, or in math, physical sciences and medicine. ―In general, female
immigrants are more likely to enter ―traditionally male‖, and higher paying, disciplines that are
science or math related than are Canadian-born females‖ (p. 48). Field of study is found to
explain at most 14% of the variance in earnings between Canadian-born and each of the two
immigrant groups, within each degree level.




                                                      METHOD


Purpose of the Study
The broad purpose of this study is to provide a more accurate assessment of the relationship
between immigrant post-secondary credentials and labour market outcomes. The study
employs data obtained from the 2002 Ethnic Diversity Survey (EDS) which sample some
43,000 respondents to the long form of the 2001 Census, to extend and elaborate previous
immigrant integration research that examined the relationship between educational credentials
and earnings. Unlike the Census or General Social Survey, the EDS data makes it possible to
determine credential origin and to specify ethnicity1. At the same time, important social
structural variations within the immigrant community can be considered in qualifying the link
between academic credentials and labour market outcomes. For the purpose of this paper,
fields of study are classified in liberal arts and applied following Lin, Sweet and Anisef‘s (2003)
scheme. Respondents who possess a university degree (obtained in Canada or elsewhere)
comprise the research sample. The natural logarithm (log) of earnings8 of immigrant and non-
immigrant groups will be modeled using ordinary least squares (OLS) regression. The analysis
is guided by the following questions:


Immigrant and Non-Immigrant Comparisons
            What effect does social structure – gender and visible minority status -- have on the
             earnings of immigrants and non-immigrants?




1
 In this iteration of the paper we limit ourselves to the ‘visible-minority’ distinction. Future work will differentiate
among ethnic groups.


                                                                                                                           8
             How do social structure and individual differences – language competence, age, age
              of arrival and time since immigration -- qualify the relationship between educational
              (university) credentials and the earnings of immigrants and non-immigrants?


Within-Immigrant Comparisons
             Are there differences by gender and visible minority status in immigrants‘ earnings
              based on educational credential features of: field of study, level of (university)
              education, and whether the degree was obtained in Canada?


Objectives
We begin the analysis by constructing separate profiles of respondents who posses university
degrees. These profiles are based on gender, visible minority and immigrant vs. non-immigrant
variables -- the individual differences of primary interest in the study. This leads to 8 main
groups for which other background, educational and employment information will be displayed.
We next model the (log) earnings of immigrants and native-born using OLS regression, to
assess the relative importance of structural, individual and educational factors in determining
earnings for those respondents who earned income in 2001 through employment. Finally, we
examine differences in earnings of immigrants in relation to their credential origin, level, and
field of study. These differences are further differentiated in terms of gender and visible-minority
status.


Data
The data employed in the study were drawn from the 2002 Ethnic Diversity Survey (EDS)
conducted by Statistics Canada in 2001. As a supplement, EDS contains data from the 2001
Census that add information on respondents‘ education and employment in 2000. The specific
Census data used in this paper is respondents‘ post-secondary program of study that allows the
identification of minor and major fields of study according to Statistics Canada classification.
Based on the minor fields of study, we aggregate academic programs of university degree
holders into two categories: liberal arts and applied. Even if an analysis at the level of major
fields of study would be desirable, low counts for some of the design groups may introduce
large errors, so detailed FOS analysis is postponed for further investigation.




                                                                                                      9
Sample
For the purpose of this paper, a working sample was defined based on the following criteria:
       -   respondents who were Canadian born or landed immigrant
       -   respondents who possess a university degree (based on EDS survey).
       -   respondents who clearly declared their major field of study (Census survey).
       -   respondents between 25-64 years of age who are expected to be active in the labour
           market
       -   respondents who declared employment as their main source of income.


This selection ensures that the analysis focuses on respondents who tested the value of their
credentials in the labour market. By applying these criteria, the EDS sample of 42,476
respondents is reduced to a research sample of about 6,150 cases of whom about 88%
reported an income in 2001. This research sample represents about 60% of all respondents in
the EDS sample that possess university degrees. Rescaled weights are computed from the
cross-sectional survey weights and will correctly estimate population proportions, while keeping
the research sample size.


Some demographic characteristics of the research sample are as follows:
       -   The proportion of women in the research sample is 48%.
       -   The sample contains 28% immigrants, as compared to 21% in the EDS sample. This
           difference is mainly due to the larger proportion of university graduates within
           immigrant group, a difference that is mainly due to the visible minority group.
       -   Average age is 41 years.
       -   Nineteen per cent of respondents declare to belong to a visible minority group. This
           proportion is much higher than the proportion of visible minorities in the EDS sample
           (12%).
       -   The sample is diverse in terms of the highest educational level attained. About 82%
           of respondents have bachelors‘ degree, 16% of respondents posses a masters
           degree, 2% have doctoral degrees. Eighty one per cent of university graduates
           received their education in Canada.
       -   Comparing home and work most used languages (i.e., English, French, both or
           others), we found that 37% of the respondents have a language disadvantage, using
           mostly different languages at home and work.




                                                                                                10
Variables
The variables selected to build profiles and examine the basis for employment and income are
shown in the Appendix Table. Immigrant status, visible minority and gender are selected as
design variables. Age is a control variable that will be interpreted in relation to respondents‘
work experience. Age at arrival and time since immigration will be accounted for the immigrant
groups, although these variables might be masked by the choice if the sample that includes
individuals who already ―succeeded‖ in the labour market by obtaining income based on
employment.


In this paper we construct a ‗language disadvantage index‘ that describes whether language at
work is different than language used at home. We hypothesize that this index measures the
level of social integration of respondents who, irrespective of their background, may choose
consciously to speak at home the official language spoken at work. This choice reflects both
comfort in speaking the language and openness toward integration. For those less proficient in
speaking language, the index would indicate an inability to economic integration by maintaining
barriers of communication, and possible bias of employers in offering hiring or promotion.


The effect of credentials (focal variables) is measured by origin of education (Canada and
elsewhere), level of education (undergraduate and graduate), and field of study (liberal arts vs.
applied). Among individual variables, special attention is given to the matching of language
used at home and at work that creates language advantages or disadvantages in workplace.




                                                                                                   11
                                                  RESULTS


We will first compare the Canadian-born and the immigrant groups, and continue with an
analysis within the immigrant group with focus on gender and visible minority factors. Counts
are rounded to the nearest tens and proportions to the nearest unit. Means (and standard
deviations) are rounded to the nearest unit, while measures of earnings are rounded to the
nearest tens.


Profiles of Respondents by immigrant status, gender and visible minority
In Table 1 we present descriptive statistics of main control variables across the design groups
defined by immigrant status, visible minority and gender. The average age of the research
sample is 41, with the Canadian-born visible minority groups younger than 35 and the immigrant
non-minority groups about 45 years old. Age at arrival and time since immigration are relevant
variables for immigrants. We notice that the non-minority immigrants in this sample arrived
about 20 years before the time of the survey, so this group is largely represented by the
immigration cohorts in the 80s, while the visible minority immigrants belong to cohorts arriving to
Canada in the 85s. Criteria used for sample selection are such that, on average, all immigrant
groups in this study had sufficient time since arrival 2 to ensure settlement and integration in the
labour market.


EDS data allow to construct a variable that measures the mismatch between the language
spoken most often at home and work that shows the language disadvantage experienced by
some respondents in the labour market. The proportion of those who use different languages at
home and work are quite high, ranging from 18% for the Canadian born non-minority males to
85% for immigrant visible minority females.


The educational variables selected for this analysis are: origin of highest university degree, level
of highest degree and type of degree (liberal arts or applied). Canadian credentials are
expected to be better negotiated in the labour market, which appear to benefit over 95% of the
Canadian born respondents. While over 45% of the non visible minority immigrants possess
Canadian education, only 37% of visible minority immigrants have this advantage. This is
particularly intriguing since the immigrant non-minority group is older. The highest level of
education brings an advantage to immigrants, especially those in the non-minority group (about

2
    Typical settlement time? After 10 years (?) ….. REFERENCE


                                                                                                   12
   18% graduate degree holders), followed by the minority group with 23-27% graduate degree
   holders. The distribution of respondents across the type of programs is more uniform, with large
   proportions of degrees in applied fields for all groups.


            Table 1: Distribution of respondents by immigrant status, gender and visible minority
          Immigrant status                  Canadian Born                             Immigrant
           Visible Minority            No                  Yes                 No                  Yes
                    Gender      Male      Female      Male     Female    Male     Female      Male     Female
       Research sample (N)      2140       2130        70        70      400        320       580       420
INDIVIDUAL
Age                              Mean      41        40        35        32        47        44        42        41
                                  (SD)    (10)      (10)       (9)       (7)      (10)       (9)      (10)      (10)
Language disadvantage               N     370       540        40        40       250       200       490       360
                                   (%)    (18)      (25)      (52)      (56)      (64)      (64)      (83)      (85)
Age at arrival                   Mean      NA        NA        NA        NA        22        22        27        25
                                  (SD)                                            (14)      (13)      (12)      (11)
Time since immigration           Mean      NA        NA        NA        NA        24        22        14        16
                                  (SD)                                            (16)      (14)      (11)      (11)
EDUCATION
Origin: Canadian Education
                                    N     2060      2090       60        60       190       150       220       150
                                   (%)    (96)      (98)      (90)      (93)      (48)      (46)      (38)      (37)
Highest level
         Undergraduate              N     1760      1840       60        70       290       230       450       340
                                   (%)    (82)      (87)      (88)      (92)      (72)      (73)      (77)      (81)

         Graduate                  N      390       280        10        10       110        90       140        80
                                  (%)     (18)      (14)      (12)       (8)      (28)      (27)      (23)      (18)
Program type
                 Applied            N     1540      1530       50        40       300       210       460       290
                                   (%)    (72)      (72)      (69)      (56)      (74)      (64)      (79)      (68)

                 Liberal arts       N      600       600        20        30       100       110       120       140
                                   (%)     (28)      (28)      (31)      (44)      (26)      (36)      (21)      (32)
Earnings                        Mean      74630     49340     69660     43420     76070     48090     54130     39620
                                (SD)     (61170)   (32460)   (60970)   (23280)   (54460)   (43090)   (42450)   (27970)
                                 N        1840      1860        60        70       350       280       520       340




   The last item in Table 1 shows the labour market outcome of interest in this paper, the total
   income obtained through employment. For some groups (e.g., male and female non visible
   minority) there are practically no differences in earnings between Canadian-born and
   immigrants. This finding has to be interpreted with caution since our sample selection eliminated
   the disadvantaged groups of immigrants who were not yet successful to negotiate their
   credentials in the labour market. However, when the visible minority status is considered,
   especially combined with gender, there is a clear drop in earnings. The group that fares the best


                                                                                                                   13
in the labour market appears to be that of male non-minority immigrants, while the most
disadvantaged group is that of female visible minority immigrants.


Earnings of immigrants vs. Canadian born
We performed OLS regression analysis to predict earnings (natural log) by the set of predictors
introduced previously (Table 1) to assess the effect of credentials when controlling for age,
gender, immigrant and visible minority status, language disadvantage. Table 2 shows results for
the analysis of earnings predicted for the entire sample of active labour force that reports
income resulted from employment.


Model I summary shows that 12% of the variability in outcomes is explained by individual
factors, and all are statistically significant. Older respondents have earnings advantage, while
gender, immigrant and minority status, and implicitly language disadvantage have a negative
effect on earnings. Age itself was considered separately and accounts for only 3% of earnings
variation that suggests that the design variables considered in this analysis (gender, immigrant
and minority status) are essential in predicting earnings.

                           Table 2. Regression model entire sample (N=5320)
                                         (unstandardized coefficients)
Variables in the equation                                   Model I             Model II
(Constant)                                               10.552 [.046]        10.430 [.047]
Age                                                       .013 ** [.001]      .012 ** [.001]
Immigrant status (No=0; Yes=1)                            -.080 * [.031]       .025 [.035]
Visible minority (No=0; Yes=1)                           -.116 ** [.034]      -.094 * [.034]
Gender (Male=0; Female=1)                                -.394 ** [.020]      -.389 ** [.020]
Language disadvantage (No=0; Yes=1)                      -.183 ** [.024]      -.166 ** [.024]
Origin of education (Canada=0; Foreign=1)                                     -.263 ** [.034]
Highest level of education
                    (Undergrad = 0; Graduate=1)                               .193 ** [.026]
Program type
                     (Liberal = 0; Applied = 1)                               .139 ** [.022]
                               Model summaries            R2adj = 0.123       R2adj = 0.145
                                  & ANOVA tests
                                                          F= 150.4 **          F= 113.5 **
* p<0.05 **p<0.01




                                                                                                   14
The full model introduces also the set of credentials predictors. As a result, the proportion of
variability in outcomes is slightly increased to 15%. As expected, origin of education matters
indicating that foreign education is worth less in the labour market. Graduate degrees are better
rewarded in the labour market and this gives some advantage to immigrant groups who possess
large proportions of master and doctoral degrees. Degrees in applied fields of study have a
better income return, which advantages male immigrants who are more likely to have
credentials in these fields. As a result, we notice that in Model II, the immigrant status factor
becomes statistically insignificant in predicting earnings when credential factors are accounted.



Labour market outcomes of immigrants
In this section we will examine in greater detail the earnings of immigrants. Table 1 revealed
that differences in earnings are more pronounced within immigrant groups when gender and
minority status are considered compared to the Canadian born. Also, the regression analysis
findings (Table 2) point to the social structural factors of gender and visible-minority status as
significant determinants of earnings. Among educational variables, all three dimensions of
credentials proved to be significant. These are elaborated in the next section. An additional
regression analysis for immigrants only was performed (although not reported) to assess
whether age at arrival and time since immigration are relevant factors. We did not find any
significant effect, which is consistent with the way sample was selected in this study: only those
who have in fact negotiated employment in the labour market and for whom the time since
immigration becomes less relevant. However, our findings show that even when employment
negotiation was overcome, other social structural features prove to be important in obtaining a
competitive income, as will be shown by the earnings level described in this section.


Origin of education
Table 3 shows earnings by minority status and gender for Canadian and foreign educated
immigrants. Canadian education results in higher income for all groups, the less advantaged
being visible minority women. While women educated in Canada have comparable earnings
irrespective of minority status, differences in earnings are very large for men.


The most advantaged group in the labour market are Canadian educated men from non-
minority group, and the less advantaged are foreign educated female from a visible minority
group. Foreign education appears to penalize more the visible minority groups, women


                                                                                                     15
especially. The earnings ratios in Table 3 suggest that gender in interaction with visible minority
status produces stronger effects on earnings than origin of education within immigrant groups.


                                 Table 3: Earnings by Origin of education
                        Visible Minority                  No                                  Yes
                                Gender         Male              Female               Male           Female
Canadian Educated                Mean          80040              49670              65600            46410
                                (SD)          (58140)            (38620)            (40620)          (33290)
                                  N             160                130                 190             140
Foreign Educated                 Mean          72680              46790              47430            35020
                                (SD)          (51010)            (46550)            (42120)          (22650)
                                  N             190                150                 330             200
Earning ratios               Gender        Fem vs. Male  Can-ed: 0.62          Fem vs. Male  Can-ed: 0.70
                                                         Foreign-ed: 0.64                    Foreign-ed: 0.74
                                           Minority vs. Non-minority  Male Can-ed: 0.82
                       Visible minority                              Female Can-ed: 0.93
                                                                     Male Foreign-ed: 0.65
                                                                  Female Foreign-ed: 0.75
                   Origin of education                    Foreign-ed vs. Can-ed (all): 0.81




Highest level of education
Analysis of earnings by level of university education (Table 4) shows that as expected, graduate
education results in higher income for all groups, the less advantaged being visible minority
women. This group has the lowest advantage brought by increased educational level, while the
male non-minority group has the highest income advantage. The most advantaged are males
with graduate degrees from non-minority groups. The earning ratios in Table 4 suggest that
gender in interaction with visible minority status produces stronger effects on earnings than level
of education within immigrant groups.


                                  Table 4: Earnings by level of education
                        Visible Minority                  No                                   Yes
                                Gender         Male             Female             Male             Female
Undergraduate                    Mean          72220             45220             51740             38290
                                (SD)          (55630)           (46350)           (36790)           (27760)
                                 N              260                200              400               270
Graduate                         Mean          85940             55040             61670             44960
                                (SD)          (50270)           (33170)           (56250)           (28410)
                                 N              100                 80              130                70
Earning ratios               Gender        Fem vs. Male  Underg: 0.63         Fem vs. Male  Underg: 0.74
                                                          Graduate: 0.64                       Graduate: 0.73
                                           Minority vs. Non-minority  Male undergrad: 0.72
                       Visible minority                              Female undergrad: 0.85
                                                                         Male graduate: 0.72
                                                                       Female graduate: 0.82
                   Level of education                     Undergraduate vs. graduate (all): 0.82



                                                                                                                16
Field of study
In agreement with previous studies (Lin, Sweet & Anisef, 2003) the analysis of earnings (Table
5) shows that applied degrees lead to higher income for all groups. However, the differences
are modest for the non-minority male and female groups and more pronounced for the visible
minority groups.


Earnings for female non-minority are almost similar in either liberal arts or applied education
group, and earnings differences for male non-minority are no more than 7% higher in applied
fields. However, for the visible minority group, possessing an applied degree leads to over 20%
increase in earnings. The closer to parity are female visible minority who possess degrees in
applied fields. Still, belonging to a visible minority group leads to lower earnings, especially for
liberal arts graduates. The earning ratios in Table 5 suggest that gender in interaction with
visible minority status produces stronger effects on earnings than type of field of study within
immigrant groups.


                                    Table 5: Earnings by field of study
                   Visible Minority                      No                                   Yes
                            Gender           Male               Female                  Male         Female
Liberal arts               Mean              72350               47510                 45030          33280
                           (SD)             (47640)             (54650)               (23990)        (19850)
                             N                100                 100                    100           100
Applied                   Mean               77470               48420                 56310          42300
                           (SD)             (56830)             (34950)               (45530)        (30410)
                             N                260                 180                    420           240
Earning ratios          Gender           Fem vs. Male  Liberal: 0.66           Fem vs. Male  Liberal: 0.74
                                                           Applied: 0.63                       Applied: 0.75
                                         Minority vs. Non-minority  Male Liberal: 0.62
                   Visible minority                                  Female Liberal: 0.70
                                                                       Male Applied: 0.73
                                                                    Female Applied: 0.87
                   Field of study                       Liberal arts vs. Applied (all): 0.87


Education Premiums for Immigrants
How large is the education premium for each of the three dimensions analyzed previously? As
noticed in Tables 3-5 the answer is largely determined by social structures. Having a Canadian
education is worth more for visible minority immigrants leading to an increase in income
between 22-28% as compared to non visible minority immigrants (6-10%). Field of study is also
a gateway to better income for visible minority immigrants who get a bonus of 20-21% if they
possess degrees in applied fields as compared to non visible minority immigrants (2-7%). Only
level of education leads to less pronounced differences by visible minority status. The premium


                                                                                                               17
for graduate education indicates a 16-18% increase in income for non visible minority and 15-
16% for visible minority immigrants.




                                  SUMMARY AND DISCUSSION


The main finding of this paper is that individual characteristics (gender, immigrant status and
visible minority) account for the largest proportion of variation in earnings for university degrees
holders. In relation to the above individual characteristics, language proficiency translated in the
advantage of using same language at home at work creates a large disadvantage for
immigrants. When comparing Canadian born with immigrants, the negotiation of credentials in
the labour market is largely determined by whether education was obtained in Canada or
elsewhere. Foreign education is largely penalized in the labour market when controlling for
individual factors. As expected, level of education and field of study create earning advantage
for graduate degree holders in applied fields.


The typical earning ratio of 65 cents earned by women for every dollar earned by men is
observed for immigrants independent of whether education was obtained in Canada or
elsewhere, level of education and field of study. Since clear income disparities can be also
observed for Canadian born women it results that gender inequity in earnings is a persistent
issue that is just amplified for immigrant women. Foreign education is still penalized for visible
minority immigrants. However, minority groups holding graduate and/or applied degrees
approach earning parity faster rather than those with undergraduate liberal arts degrees. This
suggests that Canadian labour market remains gendered and systemically biased by factors
that are not directly related to immigrant status. Gender, race, ethnicity, skin colour may still be
barriers to full economic integration and access to best jobs. The immigrant status often
intersects with these features obviously worsening individuals‘ integration. This requires further
research as to separate the effect of various biases. This also suggests that successful
immigrant economic integration of immigrants may be only marginally improved by corrections
to the system of credential and skills (workplace experience) recognition. However it may be
that the credential negotiation process may be influenced by social/cultural capital and the
subjective perceptions and dispositions that activate those resources.




                                                                                                     18
The declining earnings of immigrants in the last decade are discussed by Worswick (2004) who
argues that since transferability of foreign education into the Canadian labour market is not
operating (either lower or not recognized) and work experience in home country is not valued by
Canadian employers, the Canadian immigration policy that places major emphasis on human
capital needs to be revised. Tolley (2003) analyzes the selection criteria of this policy in relation
to research and remarks that language proficiency appears to be a most reliable determinant of
economic integration. While education is a strong factor that differentiates economic outcomes
within the immigrant groups, foreign education especially from non-traditional source countries
(non-Anglophone) does not bring a large enough return on earnings compared to Canadian-
born. Our findings show that even for immigrants who were successful in negotiating their
credentials in the labour market, earning parity is still far to be achieved especially for women
and visible minorities.


Both education and pre-immigration work experience are not fruitful factors in the absence of
language proficiency. This relation must be enhanced for the high-skilled occupations where the
expected level of language proficiency is higher. Tolley supports a shift in the Immigration policy
from the human capital system toward occupation-based criteria (e.g., Australian system) and
suggests that comparative research of economic outcomes of immigrants selected under
different systems would be illuminating. As Reitz (2001) pointed previously the lack of
recognition of formal credentials by professional associations and employers, non-recognition of
foreign education, discounting of immigrants‘ skills and prior experience create employment
disadvantage and sources of discrimination, and are different from pay which is a source of
discrimination. Immigrants receive lower earnings premiums for education and work experience,
and immigrants from some origins groups are even more disadvantaged. Reitz mentions
prejudices, ignorance, social conformity, bureaucratic practice, lack of information as factors
that impede employers to assess the relevance of immigrants‘ credentials and prior experience
for Canadian workplaces.


While the present study still reveals existing earning inequities for immigrants and within
immigrant groups, the findings show that some privileged groups of immigrants are quite
successful in negotiating their credentials in the Canadian labour market reaching sometime
earnings levels above the Canadian born. First, our sample includes only those who were
successful in finding employment. Second, more research that relate credentials and
occupational status is needed to find whether the negotiation of credentials implies a good job to



                                                                                                    19
education match, and whether immigrants are using in the labour market the education acquired
in Canada or elsewhere. It is not uncommon that immigrants are forced to abandon their
previous credentials and to consider new avenues in the Canadian educational system and the
labour market in order to secure stable employment. Going back to school, working in jobs that
do not match their initial education, doing multiple jobs, changing careers are common ways in
which immigrants earn their incomes. Even if these are typical pathways of economic integration
that are not new in the history of immigration, the fact that highly educated newcomers who
have been accepted to Canada because of their skills and education have to embrace divergent
or cumbersome routes toward employment, leads to an overall economic loss for both
individuals and the Canadian society.




References

Allen, M., Harris, S., Butlin, G. (2003). Finding their way: a profile of young Canadian graduates.
Catalogue no. 81-595-MIE, Ministry of Industry & Statistics Canada

Anisef, P., Sweet, R. & Frempong, G. (2002). Labour Market Outcomes of Immigrant and Racial
Minority University Graduates in Canada. Centre of Excellence for Research on Immigration
and Settlement (CERIS) Working Paper Series: Toronto

Aydemir, A. & Skuterud, M. (2004). Explaining the Deterioration Entry Earnings of Canada’s
Immigrant Cohorts: 1966-2000. Catalogue no. 11F0019MIE. Statistics Canada, Family and
Labour Studies Division.

Baldwin, J. R., & Beakstead, D. (2003). Knowledge workers in Canada's economy, 1971-2001.
Catalogue no. 11-624-MIE — No. 004, Statistics Canada, Micro-economic Analysis Division.

Beach, C. and Worswick, C. (1993). ―Is there a double negative effect on the earnings of
immigrant women?,‖ Canadian Public Policy, 19: 36-53.

Beakstead, D. & Vinodrai, T. (2003). Dimensions of Occupational Changes in Canada’s
Knowledge Based Economy, 1971-1996. Catalogue no. 11-622-MIE — No. 004. Statistics
Canada, Micro-economic Analysis Division.

Boothby, D. (2000). Earnings Differences by Detailed Field of Study of University Graduates,
Research Paper R-00-1-5E. Ottawa. Human Resources Development Canada and Statistics
Canada.

Boyd, M. (1992) Gender, visible minority and immigrant earnings inequality: reassessing an
employment equity premise. In V. Satzewich (ed.), Deconstructing a Nation: Immigration,
Multiculturalism and Racism in 1990s Canada. Halifax: Fernwood Publishing and Saskatoon,



                                                                                                 20
University of Saskatchewan, Department of Sociology, Social Science Research Unit, pp. 279-
321.

Davies, S. and Guppy. N. (1998). Race and Canadian Education. in Vic Satzewich (ed). Racism
and Social Inequality in Canada: Concepts, Controversies and Strategies of
Resistance.Toronto: Thompson Educational Publishing Inc.

Ferrer, A. & Riddell, C. (2004). Education, Credentials and Immigrant Earnings. TARGET
Working Paper 020.

Finnie, R. (2001). Fields of Plenty, Fields of Lean: The Early Labour Market Outcomes of
Canadian University Graduates by Discipline. The Canadian Journal of Higher Education, Vol.
XXXI, No. 1, pp.141-176.

Frenette, M. & Morissette, R. (2003). Will they ever converge? Earnings of immigrant and
Canadian-born workers over the last two decades. Analytical Studies Research Paper No. 215,
Statistics Canada.

Heisz, A. (2003). Relative earnings of British Columbia university graduates. Education
Quarterly Review, vol. 9, no.1, pp. 35-47. Ottawa: Statistics Canada

Hiebert, D. (1997). The Colour of Work: Labour Market Segmentation in Montreal, Toronto, and
Vancouver, 1991, Working Paper No. 97-02. Vancouver: RIIM.

Kunz, J., Milan, A., and Schetagne, S. (2000). Unequal Access: A Canadian Profile of Racial
Differences in Education, Employment and Income. A Report Prepared for the Canadian Race
Relations Foundation. Toronto: Canadian Race Relations Foundation.

Kunz, J. (2003). Being young and visible: labour market access among immigrant and visible
minority youth - final report. Applied Research Branch Paper SP-581-08-03E. Gatineau,
Québec: Human Resources Development Canada,

Lavoie, M. and R. Roy. 1998. Employment in the Knowledge-Based Economy: A Growth
Accounting Exercise for Canada. Applied Research Branch Research Paper R-98-8E. Ottawa:
Human Resources Development Canada.

Li, P.S. (2001). ―The Market Worth of Immigrants‘ Educational Credentials,‖ Canadian Public
Policy 27, No.1.

Li, P.S. (2003). Destination Canada: Immigration Debates and Issues. Toronto: Wall and
Thompson.

Lin, Z., Sweet, R., Anisef, P. (2003). Consequences and Policy Implications for University
Students Who Have Chosen Liberal or Vocational Education in Canada: Labour Market
Outcomes and Employability Skills. Higher Education Policy, 16, pp.55-85.

Morissette, R., Ostrovsky, Y. & Picot, G. (2004). Relative Wage Patterns among the Highly
Educated in a Knowledge-based Economy. Catalogue no. 11F0019MIE — No. 232, Statistics
Canada. Business and Labour Market Analysis Division




                                                                                              21
Picot, G. & Hou, F. (2003). The rise in low-income rates among immigrants in Canada.
Catalogue no. 11F0019MIE. Statistics Canada, Business and Labour Market Analysis Division

Picot, G. (2004). The Deteriorating Economic Welfare of Immigrants and Possible Causes.
Catalogue no. 11F0019MIE, Statistics Canada, Business and Labour Market Analysis Division

Reitz, J. G. (2001). Immigrant Skill Utilization in the Canadian Labour Market: Implications of
Human Capital Research. Journal of International Migration and Integration, 2(3), (pp. 347-378)

Reitz, J. G. (2005). Tapping Immigrants‘ Skills. New Directions for Canadian Policy in the
Knowledge Economy. IRPP (Immigration and Refugee Policy) Choices, 11(1) (pp. 2-18).

Statistics Canada (2003a). The Canadian Labour Market at a Glance. Catalogue no. 71-222-
XIE, Statistics Canada, Labour Statistics Division.

Statistics Canada (2003b). Earnings of Canadians: Making a Living in the New Economy. 2001
Census analysis series, Catalogue no. 96F0030XIE, Statistics Canada.

Statistics Canada (2003c). Education in Canada: Raising the Standard. 2001 Census analysis
series, Catalogue no. 96F0030XIE, Statistics Canada.

Statistics Canada (2003d). Education Indicators in Canada: Report of the Pan-Canadian
Education Indicators Program 2003. Catalogue no. 81-582-XIE, Statistics Canada

Sweetman, A., McBride, S. (2004). Postsecondary Field of Study and the Canadian Labour
Market Outcomes of Immigrants and Non-Immigrants. Ottawa: Human Resources and Skills
Development Canada.

The Conference Board of Canada (2004). Performance and Potential 2004-2005: Key Findings.
How Can Canada Prosper in Tomorrow’s World? URL: www.conferenceboard.ca

Thompson, E. N. (2000). Immigrant Occupational Skill Outcomes and the Role of Region-of-
Origin-Specific Human Capital. Applied Research Branch Research Paper W-00-8E. Ottawa:
Human Resources Development Canada.

Tolley, E. (2003). The Skilled Worker Class. Selection Criteria in the Immigration and Refugee
Protection Act. Metropolis Policy Brief no. 1 (pp. 1-8).

Worswick, C. (2004). Immigrants‘ Declining Earnings: Reasons and Remedies. C. D. Howe
Institute Backgrounder. URL: www.cdhowe.org




                                                                                                 22
Appendix
Variables list
Variable Name                Variable Name                    Survey Questions               Derived variables
   (paper)                   (EDS/Census)
ECONOMIC
Employment-          EC_Q044 (EDS)                     For how many weeks during past        continuity
weeks                C_WEEKS (Census)                  months were you employed?             (Full year=52 weeks;
                                                                                             Yes=0, No=1)
Employment-          EC_Q048 (EDS)                     How many hours per week did you       stab
hours per week       C_HOURS (Census                   usually work?                         (FT=30 or more;
                                                                                             FT=0, PT=1)
Total income         ECQ310 (EDS)                      Best estimate of personal income      lntotinc
                     C_TOTINC (Census)                 before taxes and all sources          (natural log of income)
Source of income     ECQ300 (EDS)                      What was your main source of          source
                                                       personal income in the past 12        (Empl=1; Retirement
                                                       months?                               /savings=2; Other=3)
INDIVIDUAL
Sex                  SEX                               Respondent‘s sex                      gender (M=0, F=1)
Age                  AGE                               Respondent‘s age                           Continuous
Immigrant status     IMMSTAT                           In what country were you born         immigrant_r
                                                       (BK_010)                              (No=0, Yes=1)
Age at arrival       IMMAGE                                                                       Continuous
Year of arrival      IMMY2                             In what year did you first come to    timeimm
                                                       Canada to live (BK_040)               (time since arrival)=
                                                                                             2001-IMMY2
Visible minority     VISMINC                           Are you… ? BK(Q110)                   minority_2
                                                                                             (No=0, Yes=1)
LANGUAGE USE
Language at          LGFOS                             What language do you speak most       Eng (No=0, Yes=1)
home                                                   often at home (LG_Q110)               French (No=0, Yes=1)
                                                                                             both (No=0, Yes=1)
                                                                                             other (No=0, Yes=1)
Language at work     LGWMOS                            In the past 12 months what            Eng (No=0, Yes=1)
                                                       language did you mostly speak at      French (No=0, Yes=1)
                                                       work (EC_Q100)                        both (No=0, Yes=1)
                                                                                             other (No=0, Yes=1)
Language             Based on language home and                                              langdif (same=0;
disadvantage         language work                                                           different=1)
EDUCATION
Educational level    HLOSUN1 (EDS)                     Highest Level of Schooling            bach (No=0, Yes=1)
                     HLORS (Census)                                                          master (No=0, Yes=1)
                                                                                             phd (No=0, Yes=1)
Where was            HLOSREGD                          In what country did you attain this   educcan (Canada=0,
education attained                                     education (EC_Q020)                   Foreign=1)
Major field of       C_DGMFSR                                                                code_EDS (units)
study                                                                                        minor_fs
                                                                                             MFS (10 categories ->
                                                                                             10 variables)
Program type         Based on LSA classification and                                         prg_type
                     use of code_EDS (units FOS)                                             (Applied=0, Liberal=1)




                                                                                                                       23

				
DOCUMENT INFO