• Open Access

Under-use of migrants' employment skills linked to poorer mental health


Correspondence to: Dr Alison Reid, Epidemiology Group, Western Australian Institute for Medical Research, M519, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009; e-mail: alison.reid@uwa.edu.au


Objective: Since the mid-1990s Australia's immigration program has focused on encouraging skilled migration. This study investigated skill usage in three longitudinal studies of immigrants to Australia and examined if there is an association with mental health status.

Methods: Three Longitudinal Surveys of Immigrants to Australia (LSIA), with multiple data collection waves, were conducted between 1994 and 2006. Some 5,192 primary applicants participated in LSIA1, 3,124 in LSIA2 and 9,865 in LSIA3. Data collected included demographics and employment history in all surveys as well as mental health in LSIA1 and 2.

Results: Among migrants in LSIA 1, 49% reported working in jobs in which they used their skills sometimes, rarely or never, 3½ years after immigrating. This was not solely explained by English language proficiency as 47% of migrants who reported speaking English well or very well did not use their qualifications in their job. Migrants who did not use their job qualifications at wave three had a worse GHQ-12 score at wave three after adjusting for age, sex, country of birth and highest educational qualification. There was no difference in wave one or wave two GHQ-12 score between those who did or did not use their job qualifications at wave three. The pattern was similar for those migrants in LSIA 2.

Conclusions and Implications: There is a large under-utilisation of employment skills in the migrant population in Australia up to 3½ years after immigrating. This is associated with poorer mental health.

Skilled migration has become an increasingly important component of the migrant intake of Organisation for Economic Co-operation and Development (OECD) countries. Between 1995 and 2000 the overall migrant workforce in OECD countries increased by 4% per year, while the highly educated migrant workforce increased by 35% and 14% in the United Kingdom (UK) and the United States (US) respectively.1 In Australia in 2007/08, skilled migrants made up the largest proportion of all permanent migrants, with two skilled visas being granted for every family visa granted.2 Skilled migrants to Australia came from more than 180 countries with the majority coming from the UK, India, China and South Africa.2 Similarly, in Canada 50% of all new permanent residents in 2005 were skilled migrants.3

Skilled migrants are selected based on criteria such as age, receiving country language ability, qualifications and work experience, as they are expected to fill gaps in the receiving countries labour force. However, there is some evidence to suggest that in their new country skilled migrants are more likely than native-born workers to work in jobs for which they are over-qualified. For example, university-educated migrants to Canada who arrived between 1995 and 1999 tended to work in low-education jobs in 2001 compared with Canadian-born workers.4 This situation was not limited to migrants who had arrived in the previous five years. Twice the number of established (arrived in Canada between 1991 and 1994) male university-educated migrants worked in jobs with low educational requirements compared with Canadian born university-educated males in 2006.5 Among migrants to Australia between 1997 and 2007 who were employed in 2007, 23% of professional migrants and 30% of technicians and trade workers (37% of all skilled work visas granted) had changed their major occupation group from their final job in their home country.6 More specifically, 22% of primary applicants were working as professionals in Australia in 2007 compared with 26% who were working as professionals before their arrival in Australia.7

The term under-employment is used where a job is found to be deficient in relation to the employment experiences of others, or in terms of a person's own education, work history and expectations. A job can be deficient in the number of hours worked, the wages that are earned or the skills that are used.8 In this study, under-employment means the disparity between skills obtained by the worker and skills used in their job. The mismatch between the demands and conditions of the job and the skills and needs of the worker result in the job being a cause of stress.9 There are many reasons why a migrant may not use their skills in their job in their new country and be at risk of under-employment. These include non-recognition of skills by officiating bodies, non-recognition of qualifications, non-recognition of foreign work experience by employers, poor dominant language skills, lack of connections and networks and possible discrimination by employers.10,11

The primary pathway by which under-employment affects health is as a stressor,8 and the association with adverse mental health effects is the same as that for unemployment, e.g. increasing rates of depression, alcohol abuse, suicide and mental hospitalisation.12 However, in general, there has been little research examining the relationship between under-employment and health. What research has been done has mostly been limited to cross-sectional studies, which makes the causal pathway difficult to tease out. For example, was the person working in a lower-skilled job because they were ill or was illness caused by working in a lower-skilled job? Of the few longitudinal studies that have examined this issue, both physical and mental health impacts from under-employment have been reported. There was a significant risk of decline in the self-reported health of university-educated workers working in jobs for which they were overqualified,13 while workers who were under-employed two years after they finished high school reported more symptoms of depression than did their peers.14 This issue has not been well studied among migrants with only two studies reporting poorer self-assessed health among recent migrants to Canada who worked in jobs for which they were overqualified.15,16

Since the mid-1990s Australia's immigration program has focused on encouraging skilled migration. This study aims to investigate the use of skills and qualifications in migrant workers in three longitudinal studies of immigrants to Australia to determine if there is an association with mental health status.


Three longitudinal surveys of immigrants to Australia (LSIA) were conducted among migrants who arrived in Australia between 1993 and 2005 by the Department of Immigration and Citizenship (DIAC), formerly known as the Department of Immigration and Indigenous Affairs (DIMIA). The characteristics of the three surveys are described in Table 1.17 Information from participants for LSIA 1 and LSIA 2 was collected by in-depth interviews while LSIA 3 used a written questionnaire for the first wave and a telephone interview for the second. More than 40% of the interviews of the first two surveys were conducted in a language other than English, using either family or friends as interpreters or accredited interpreters or bilingual interviewers. LSIA 3 interviewees who had difficulty with the English questionnaire had access to an interpreter service. For the second wave of LSIA 3, those who had had difficulty with English in the first wave had a bilingual phone interview.

Table 1.  Description of three Longitudinal Surveys of Immigrants to Australia.
Survey characteristicsLSIA 1LSIA 2LSIA 3
Years conducted1994–19992000–20022004–2006
Number of waves322
Time span of waves6, 18, 42 months after migration6, 18 months after migration6, 18 months after migration (or being granted an onshore visa)
How surveys conductedIn-depth interviewIn-depth interviewWritten questionnaire (wave 1) computer assisted telephone interviewing (wave 2)
Who interviewedPrimary Applicant and any accompanying spousePrimary Applicant and any accompanying spousePrimary Applicant
Visa grantedOffshoreOffshoreOffshore and Onshore
Number of initial participants5,1923,1249,877
Year of migration1993–19951999–20002004–2005

LSIA 1 and 2 were conducted on offshore migrants (those who applied for their visas while they were living overseas) while LSIA 3 had many more migrants who had been granted their visa onshore (they had applied for their visa after they were resident in Australia, e.g. overseas students). LSIA 1 and 2 included humanitarian migrants in their cohorts but LSIA 3 did not.

This study included only primary applicants, that is those who made the application to come to Australia, aged 15 years and older and who completed all waves of the survey.

Variables of interest

Questions asking how often the respondent used their skills in their current main job were asked at each wave of each LSIA survey. Responses were coded as very often, often, sometimes, rarely and never. Under-employment was determined as reporting using those skills only sometimes, rarely or never in the main job in the final wave of each survey.

Mental health was measured using the General Health Questionnaire-12 (GHQ-12)18 in LSIA 1 and 2. The GHQ was developed as a screening tool to detect those likely to have or to be at risk of developing psychiatric disorders. It measures the common mental disorders of depression, anxiety, somatic symptoms and social withdrawal.19 The GHQ-12 was not collected in LSIA 3. It was scored 0, 1, 2, 3, which has been suggested as a more appropriate method of scoring the GHQ-12 for use in occupational studies, than the 0,0,1,1 scoring method.20 Summing the possible answers gave a range of 0–36 where a higher score equated to poorer mental health.

Other variables examined included country of birth, visa category, which was aggregated to family, skilled and humanitarian, sex and age at last birthday. The primary applicant was asked to select between very good, good, fair, poor or very poor as a response to the question: “Over the last month would you say your health has generally been?” to ascertain general health status. Highest formal qualification was used as a measure of socioeconomic status and was aggregated to bachelor degree or higher, technical or professional or trade qualifications.

Statistical analysis

Chi-square tests were used to examine the relationship between the use of skills and survey participant characteristics. T-tests and one-way analysis of variance tests examined the relationship between use of skills, participant characteristics and GHQ-12 score. For LSIA 1 and 2 linear regression was used to relate use of skills and qualifications in the job at the last wave of the survey to GHQ-12 score at the last wave of the survey. Final models also adjusted for sex, current age, country of birth, past GHQ-12 scores, self-reported health status in final wave and highest qualification to account for heterogeneity among the migrants. The level of statistical significance was set at p<0.05 and data analysis was undertaken using Stata 10.0.21

The Department of Immigration and Citizenship, which provided the cohort data, permit only weighted data to be published. Therefore, probability weights, (provided by the survey owner and that account for the probability of being recruited into the survey) were applied to the data using the survey suite of (svy) commands in Stata 10.0. Tables 2–5 show weighted values that are representative of the whole migrant population that arrived during each survey period and who could have been recruited into the survey.

Table 2.  Characteristics, qualifications and use of skills in job at final wave among primary applicant migrants in each Longitudinal Survey of Immigrants to Australia (who completed all survey waves). a
Primary applicant characteristicsLSIA 1LSIA 2LSIA 3
  1. a Results weighted to reflect the whole population of migrants who arrived in Australia and who could have been recruited into the survey

  2. b Approximately half of the family stream migrants surveyed in Wave One were surveyed in Wave Two

  3. c Final wave question for LSIA 1 & 2, First wave question for LSIA 3

Number (%) who participated in all waves361826495183
Mean age (years) at Wave 120.820.033.5
Male %585150
Visa Categories %   
Speak Englishc%   
Native English speaker323843
Very well131620
Not well21198
Not at all441
Highest formal qualification %   
Bachelor degree or higher244360
None or Missing423016
Use of job skills at final wave %   
Very Often/Often516460
Reason job skills not used at final wave %   
Insufficient English to apply skills1121
Qualification not relevant to job899588
Qualification not recognised213
Lack of local work experiencen/an/a1
Don't know117
Table 3.  Use of skills in job at final wave and English language proficiency among primary applicants to Australia.a
 LSIA 1 (wave 3)
Use of skills at work %
LSIA 2 (wave 2)
Use of skills at work %
LSIA 3 (wave 2)
Use of skills at work %
  1. a Results weighted to reflect the whole population of migrants who arrived in Australia and who could have been recruited into the survey

  2. b p<0.05 (p value calculated on sample rather than weighted population using a Chi-squared test)

Spoken English proficiency %Very often/oftenSometimes/rarely/neverVery often/oftenSometimes/rarely/neverOftenSometimes/rarely/never
Very well/well5347b6733b6238b
Not well/not at all366419812971
Table 4.  Mean general health questionnaire scores, by visa category, use of skills at final wave and English language proficiency.a
 LSIA 1  LSIA 2 
  1. a Results weighted to reflect the whole population of migrants who arrived in Australia and who could have been recruited into the survey.

  2. ns – not statistically significant.

  3. p-values calculated on sample rather than weighted population. T-tests examined use of skills and spoken English proficiency and GHQ-12 score and oneway analysis of variance examined Visa Category and GHQ-12 score.

Visa category
Wave 1
Wave 2
Wave 3
Wave 1
Wave 2
Very Often/Often9.028.698.568.118.17
Table 5.  Linear regression models of GHQ score at final wave of study and use of skills at final wave of study, adjusted for country of birth, age, sex, socio-economic status (highest qualification), past GHQ score and general health status.
 LSIA 1 (Wave 3) LSIA 2 (Wave 2) 
 β95% CIβ95%CI
  1. %p<0.05 %%p<0.01

Use skills–0.33–0.66 – 0.003–0.12–0.56 – 0.33
General health status    
Very Good0.75 – 1.430.55 – 1.44
Good1.09%%1.79 – 3.050.99%%2.69 – 4.66
Fair2.42%%4.13 – 8.093.67%%3.51 – 9.93
Poor/Very Poor6.11%% 6.71%% 
Highest Qualification    
Bachelor degree or higher–0.50 – 0.27–0.76 – 0.23
Technical/Professional–0.11–0.84 – 0.10–0.27–1.02 – 0.47
Trade–0.37 –0.28 
GHQScoreWave 10.12%%0.08 – 0.150.27%%0.22 – 0.31
0.24%%0.20 – 0.28n/a 

Ethics approval for this study was obtained from the University of Western Australia Human Research Ethics Committee.


Seventy per cent of respondents completed all three waves of LSIA 1 and 85% completed both waves of LSIA 2 (Table 2). Only half the family stream migrants in LSIA 3 wave one were re-interviewed in wave 2 so the proportion of respondents who completed both waves was lower at 52%. Slightly more primary applicants were male than female and their mean age increased over the three studies, reflecting the increase in qualification and skills acquisition of the later migrants. Of those whose main language was not English, 42%, 39% and 47% of primary applicants reported speaking English very well or well in LSIA 1, LSIA 2 and LSIA 3 respectively. The proportion of migrants who arrived on family visas declined over time as the proportion of skilled visas increased. Over the period of the three surveys the proportion of migrants arriving in Australia in possession of a bachelor degree or higher increased, while the number migrating with trade certificate skills remained the same. The proportion of migrants reporting that they used their skills or educational qualifications in the workplace at the final wave of each survey very often or often increased between LSIA 1 and LSIA 2, and attenuated slightly for LSIA 3. Forty per cent of primary applicants reported using their skills, sometimes, rarely or never in their job at the final wave of LSIA 3. Qualification not being relevant to job was the reason most commonly reported in the three surveys among those who reported not using their skills or education in their job at the final wave.

The use of skills or educational qualifications in the job at the final wave of each survey and English language proficiency is examined in Table 3. In LSIA 1, half of those who spoke English well or very well reported using their skills often or very often. This proportion increased in LSIA 2 and LSIA 3. In contrast, those who reported speaking English well or very well but who used their skills sometimes, rarely or never in their job at the final wave of each survey decreased from 47% in LSIA 1 to 38% in LSIA 3. This shows a large under-utilisation of skills and qualifications among migrants to Australia that is not related to English language proficiency.

There was a statistically significant difference in GHQ-12 score by visa category with humanitarian visa recipients having the highest GHQ-12 scores (Table 4).GHQ-12 score increased slightly over each wave for family and humanitarian visa holders and remained relatively stable for skilled migrants in LSIA 1. Skilled workers tended to have higher GHQ-12 scores than family visa recipients in both surveys. GHQ-12 score was higher among those who reported using their skills in their job sometimes/rarely or never compared with those who used them often or very often, excepting wave 1 of LSIA 1. Similarly, GHQ-12 score was higher among those who reported not speaking English well or not at all compared with those who spoke it very well or well. GHQ-12 score in wave three of LSIA 1 varied between country of birth groupings, from a low of 7.37 (SD 4.05) among migrants from Oceania to a high of 10.21 (SD 4.34) among migrants from Eastern Europe. The difference between these two groups was statistically significant (p<0.01, data not shown). Although the differences in GHQ-12 score by subcategory are statistically significant, they are only small differences and so are probably not clinically significant.

Linear regression models related final wave GHQ-12 score and use of skills and qualifications in final wave job for LSIA 1 and LSIA 2 (Table 5). After adjusting for physical health status, country of birth, age, sex, socioeconomic status and past GHQ-12 scores, use of skills in the job at 42 months after migration to Australia (LSIA 1) was associated with a reduction in GHQ-12 score (p=0.053). In LSIA 2, use of skills and qualifications in jobs at 18 months after migration to Australia was again associated with a reduction in GHQ-12 score, although the association was not as marked as that in LSIA 1 which had longer follow up. In summary, those who used their skills in their job had a better (i.e. lower) GHQ-12 score compared with those who did not use their skills or qualifications in their job.


The results from three large longitudinal surveys of migrants to Australia show that there is a large under-utilisation of skills and qualifications (under-employment) among recent migrants. Between 36% and 49% of new migrants reported that they do not use their skills or qualifications in their job, up to 42 months after arrival in Australia. Further, this skill and qualification under-utilisation is not limited to those who do not speak English well. Between 33% and 47% of those who reported speaking English well or very well do not use their skills or qualifications in their job in the final wave of each survey. There is some evidence that not using skills and qualifications at work may impact detrimentally on mental health.

The findings of this present study, which used a different metric to measure mental health and a different method to determine under-employment (over qualification), are of a similar magnitude to those reported among recent migrants to Canada.15 Chen et al examined over-qualification and mental health status among economically active new migrants to Canada in the Longitudinal Survey of Immigrants to Canada. If the skill requirement of the recent migrant's current job in Canada was lower than their educational attainment, lower than the skill requirement of their job before they migrated to Canada, or lower than the skill requirement of their expected job in Canada then the migrant was classified as being over-qualified. Mental health status was measured by asking about persistent feelings of sadness, depression and loneliness in the past year as a binary yes/no variable. Mental health declined if they responded no in the first wave and yes in the final wave. Those who were overqualified in terms of educational attainment had an OR=1.45 (95% CI 1.17–1.80) for declining mental health. Those who were over-qualified in terms of experience had an OR=1.23 (95% CI 1.00–1.51) for declining mental health while those who were overqualified in terms of job expectation had an OR=1.38 (95% CI 1.12–1.69) four years after migrating to Canada.

Similar results have been reported from a qualitative cross-sectional study among 22 recent skilled migrant workers to Canada who were unemployed or working outside (below) their field of employment. Participants reported negative impacts on mental and physical health resulting from under/unemployment. The most frequently-mentioned mental health issues included stress, anxiety, depression, unhappiness and worry.16

Self-reported health may not be a useful measure to assess health responses among different ethnic groups because health can be perceived in different ways by different groups.22 This present study used the GHQ-12 to measure common mental disorders. Earlier work has shown that responses to the GHQ differed when issued in various cities around the world. The diversity in how distress was experienced by various ethnic groups showed as dissimilar mean scores.23 In the present study, we reported differences in GHQ-12 score between migrant groups with the highest mean score among those from Eastern Europe and the lowest mean score among migrants from Oceania. This would be an important consideration in determining the threshold between cases and non cases. However, this present study modelled the mean GHQ-12 score for all migrants by various subgroups of visa category status, English language skills and use of skills in their job and adjusted for country of birth, therefore accounting for any difference in GHQ-12 scores between ethnic groups. So for this present study the limitations of using the GHQ among different ethnic groups should not have particularly influenced our results.

The GHQ-12 score of recent migrants to Australia reported in this present study was similar to that of employed workers reported elsewhere, although the score for humanitarian visa category holders was higher. Banks et al. showed a higher GHQ score among unemployed males (mean=13.84) and females (mean=14.25) compared with employed males (mean=8.80) and employed females (mean=8.53).20 Migrants to Australia must undergo (and pass) a medical examination, chest x-ray and an HIV/AIDS test. Therefore, the majority of migrants coming to Australia are healthy. In general, migrants tend to be healthier than the population they go to, although over time their health tends to regress to the health levels of the dominant population.24 Not using their skills at work and its subsequent effect on mental health, may be one component of post migration life that contributes to this health decline.

Immigration has been and continues to be very important for the growth of the Australian economy. Migrants contribute to Australia's economy on the demand side through the expansion of business and government services, as well as their own spending. Migrants contribute to the supply side of the economy through the labour, skills and capital they bring with them to Australia, through the contributions to technology and productive diversity.25 The skilled migration stream intake is anticipated to increase to 125,850 places in 2011/12 with around two-thirds of all visas issued being for the skilled stream.26 Poorer mental health may impact on the contribution made by skilled migrants to Australia and subsequently on the Australian economy.

Strengths of this study include the examination of three cohort studies of migrants to Australia. These surveys were conducted over the period when the distribution of migrant visas to Australia was changing. When LSIA 1 was conducted the majority of migrants came to Australia on a family visa, whereas in 2005/06 when LSIA 3 was conducted, most migrants came to Australia on a skilled visa. The surveys included large numbers of migrants and from many different countries. Attrition was a problem in the surveys at 30% for LSIA 1, 15% for LSIA 2 and 50% for LSIA 3 (where only half the wave 1 responders were interviewed at wave 2). This present study weighted the data using arrival population weights and therefore attrition should not have a major effect on the conclusions drawn.17

This study has shown there is a large under-utilisation of skills among migrant workers to Australia up to 3½ years post-migration. There is some evidence that under-utilisation of skills may be associated with a detrimental impact on mental health. With the emphasis now on recruiting skilled migrants to Australia, consideration should be given to establishing mechanisms that enable the skilled migrant to use their skills in Australian workplaces. Currently, various agencies provide a range of employment assistance schemes to assist new migrants into the workforce. These include employment orientation courses aimed at familiarising the migrant with the Australian workplace, including a component on how to use your overseas skills and experience27 and a Skills Recognition Service provided by the Government of South Australia that provides information on comparative educational assessment and the skills recognition process.28 In terms of practical support, migrants could be offered employment training and/or a period of mentoring and supervision if after one year of arrival in Australia they were unsuccessful in obtaining employment in their field.


Professor Lin Fritschi; National Health and Medical Research Council – Capacity Building Grant #546248; Department of Immigration and Citizenship.