Immigrant Satisfaction: What Is It? Does It Matter? Laurence H Lester Ph: Fax:

Transcription

1 Worrkiing Paperr No 154 Immigrant Satisfaction: What Is It? Does It Matter? Paper presentted tto tthe NILS Semiinar Seriies 14 Jully 2005 Laurence H Lester Ph: Fax: laurence.lester@flinders.edu.au

2 1. Immigrant Settlement- Measuring Successful Settlement Introduction This Working Paper examines the relationship between immigrant s self-reported satisfaction and other variables available in the Longitudinal Survey of Immigrants to Australia (LSIA) using (exploratory) cross-tabulation tables. There have been two LSIA, selected from off-shore visaed immigrants (with permanent resident status), and their migrating unit family where appropriate (\DIMIA, 2002 #132}. The first cohort (LSIA1) arrived in Australia between 1 st September 1993 and 31 st August This cohort was interviewed on three occasions, approximately five to six months after arrival (wave 1), after a further 12 months (wave 2) and a further two years later (wave 3). In wave 1, 6,961 immigrants were interviewed comprising of 5,192 Primary Applicants 2 (PA) and 1,769 migrating unit spouse (MU) an accompanying legal spouse, fiancé, or de facto partner. 3 Inter-wave sample attrition reduced the number of immigrants interviewed at subsequent waves as shown in Table 1-1. Table 1-1 LSIA Sample Numbers (% of Initial Sample) Cohort / Wave Primary Applicants Migrating Unit Spouse Total Interviewed Cohort 1 Wave 1 5,192 (100) 1,769 (100) 6,961 (100) Wave 2 4,469 (86.1) 1,530 (86.5) 5,999 (86.1) Wave 3 3,752 (72.3) 1,302 (73.6) 5,054 (72.5) Cohort 2 Wave 1 3,124 (100) 1,057 (100) 4,181 (100) Wave 2 2,649 (84.8) 889 (84.1) 3,538 (84.6) Note: Source DIMIA, Longitudinal Survey of Immigrants to Australia cohorts 1 and 2. In total, cohort 1 of the LSIA was around 7% of the total population of off-shored visaed immigrants, and represented the remaining reference population of around 75,000 PAs who arrived in Australia in the two-year survey period of LSIA1 (DIMIA 2002). 1 This NILS Working Paper is drawn from ongoing research related to my PhD thesis. I thank the ARC (LP ) and DIMIA for funding this research, DIMIA for providing the LSIA data, and my supervisor Professor Sue Richardson for her helpful comments on an early draft. All errors remain my responsibility. A version of this paper was also presented to the UniSA External Scholars Seminar Series, 28 July In 1996, the term Principal Applicant (LSIA1) was replaced the term Primary Applicant (LSIA2). 3 PAs include immigrants 15 years and over at the date of arrival. Those under 15 years were less than 4% of the total (for LSIA2 there were other exclusions but they were less than 1% of the total). 2

3 The second cohort (LSIA2), which arrived between 1 st September 1999 and 31 st August 2000, included 3,124 PAs and 1,057 MUs at wave 1. The second cohort was interviewed twice; approximately six months after arrival (wave 1) and one year later (wave 2). In total, cohort 2 of the LSIA was about 10% of the total population of offshore visaed immigrants, and represented the remaining population of 32,500 PAs who arrived in Australia in the one-year survey period of LSIA2 (DIMIA 2002). To use traditional econometric methods to model successful settlement it is necessary to choose a single representative dependent variable that is explained by multiple independent variables. To this end, this Paper examines immigrants satisfaction with Australia as a potential overall representative measure of successful settlement Immigrants Satisfaction or Subjective Well-Being There has been a tendency in the past for economist to avoid investigation of subjective well-being (SWB) (Clark, Etilé et al. 2004; Gittins 2005), but recently research and analysis has become more prevalent (Winkelmann 2004), based on the premise that Economic things matter only in so far as they make people happier (Oswald 1997, p.1815), and that Happiness is generally considered an ultimate goal of life (Frey & Stutzer 2002, p.402). The general reluctance of economist to consider the evidence provided by surveys of SWB seems to have been a result of the grounding of economics in the pursuit of self-interest by self-determining agents (Smith 1776; Layard 2005). In addition, Ormerod (1999) suggests that an economy, comprising a collection of individuals, may more appropriately compared to a living organism, one that learns and changes as opposed to operating strictly in a predictable manner. Moreover, the action of individuals cannot fully be described by their pursuit of self-interest (by those who possess a full information set), but that individual actions are also shaped by their connection to others in society. Burchardt (2004) notes that there are economists who maintain that questions about satisfaction are not suitable measures of utility as they invite answers that use cognition and comparisons and not simply the individual s mood. There is now however general recognition by economists that economic outcomes can be influenced by sociological and psychological factors, as well as economic factors 3

4 (Mazerolle & Singh 2004), and happiness surveys are accepted as a tool of welfare economics (Layard 2005). The Paper s second aim is to examine the relationship between immigrants reported level of satisfaction with their life in Australia and their labour market status. There is an emphasis on immigrants gaining employment as a measure of successful settlement, by the bureaucratic and political arms of government and in the academic literature. This Paper considers the appropriateness of this view in regard to immigrants who arrived in Australia in the late 1990s and early 2000s (surveyed in the LSIA) and suggests that the emphasis may be misplaced. Notwithstanding that employment is seen as a way by which immigrants get access to economic and social resources (Stier & Levanon 2003, 81), and that employment can be an indictor of assimilation (Miller & Neo 1997; Thapa 2004), it is not the sole criterion on which successful settlement should be assessed. As a first step in the analysis, this Paper conducts exploratory data analyses on immigrants self-reported subjective level of satisfaction with Australia by examining responses to two questions in the LSIA survey: All things considered, how do you feel about your life in Australia? Are you Very satisfied, Satisfied, Neither satisfied nor dissatisfied, Dissatisfied, Very dissatisfied, or No opinion? Do you think your decision to migrate to Australia was the right one? Yes, No, or Note sure/don t know? As noted, there are five possible answers to the first question (and No opinion ), and there are three answers to the second. In keeping with usual practice, it is assumed that the recorded ordinal answers are representative of underlying continuous measures (Clark, Etilé et al. 2005). These questions are viewed as general indicators of overall satisfaction or happiness of immigrants, and analysis of such questions has become relatively common in the economics literature (see for example, Clark, Georgellis et al. 2001; McBride 2001; Ferrer-i-Carbonell 2002a; Graham, Eggers et al. 2004; Borooah 2005) and the reviews of Diener & Biswas-Diener (2002; Frey & Stutzer (2002). It seems unlikely that an immigrant will claim to be very satisfied with life in Australia, or with their 4

5 decision to immigrate to Australia, but be unhappy or dissatisfied in a more general sense. An important issue however is whether the self reported level of satisfaction with life in Australia should be viewed as the only (or major) indicator of successful settlement? According to the data in the LSIA, immigrants were overwhelmingly satisfied with their life in Australia. As shown in Table 1-2, at wave 2 in cohort 1 and 2, the level of satisfaction with life in Australia was very high. 4 Table 1-2 Level of Satisfaction with Life in Australia LSIA (%) C1W2 C2W2 Very Satisfied Satisfied Subtotal% Neither satisfied nor dissatisfied Subtotal% Dissatisfied Very dissatisfied No opinion Total If the answer to this question were accepted as the single indictor of successful settlement it would appear, at first glance, that there was little to study regarding successful settlement as it could be concluded that immigrants in the LSIA were, in their view, unambiguously successful. As discussed below, however, the LSIA data provides several other indicators of successful settlement and it is these, in combination with the satisfaction question that are suggested as indictors of successful settlement in a broader sense. Nonetheless, there is not a commonly accepted definition of what constitutes immigrants successful settlement and hence the analysis can be view as an exploration, providing suggested matters for further consideration. Finally, it need be noted that there will be influences on immigrant s satisfaction that are not or cannot be measured, and several of these factors are more important in describing female or male happiness, and hence it cannot be expected that a complete explanation for individual happiness can be found in the LSIA data. 4 All LSIA data reported in this Paper are derived using weights supplied by DIMIA. 5

6 1.3. Are Survey Data Reliable and Appropriate for Quantitative Evaluation? There is theoretical and practical support for measuring, analysing, and comparing answers to questions relating to satisfaction, and comparing self-reported levels of satisfaction. Moreover, it has been established for some time that single question measures of satisfaction are at least equally satisfactory as multiple question measures (see for example Andrews & Withey 1976). Such questions have been widely used in sociological and psychological research (van Praag & Baarsma 2005), and have now become more prevalent in the economics literature (see for example, With regard to the view of the correctness of the decision to immigrate, this is a straightforward question with an unambiguous meaning and interpretation. It is however necessary to note the caveat that applies to surveys in general. There may be some unmeasurable bias in the responses due to several factors including cultural biases, interactions with the surveyor, memory, difficulties in interpreting questions and answers, self-image, and social desirability bias (Bertrand & Mullainathan 2001). Nonetheless, subjective survey data have been incorporated into economic analysis since the 1970s (see for example, Senik 2005) who refers to over 25 studies using subjective data for economic studies). There are several indirect measures of immigrants feelings and attitudes to Australia in the LSIA that provide support to the analysis of satisfaction, and these measures can also be subjected to statistical analysis (including for example, encouraging others to immigrate to Australia, emigration intentions, and citizenship intentions). These measures are discussed below. There are other measures of immigrants feelings and attitudes to Australia that are available in the LSIA that generally cannot be statistically analysed to the extent as the questions referred to above. For example, immigrants are asked open-ended questions about the things they most like and dislike about Australia. Whereas interpersonal comparisons of levels of satisfaction are generally acceptable this is not the case for answers to these questions (such as Lifestyle, Everything, Nothing ). In addition, as expected, things about Australia that some immigrants like others dislike. Answers to such questions cannot be analyses with statistical rigor. As well as providing guidance for the econometric analysis in the following Paper a number of important findings come out of the exploratory data analysis of satisfaction 6

7 of immigrants in the LSIA. First, whilst immigrants labour market outcomes are important there is an over-emphasis in discussion by government and academia on this factor generally, and certainly from the perspective of immigrants. Significant numbers of immigrants identified themselves as not in the labour force, and this was so for both PAs and MUs in both LSIA1 and LSIA2 and across visa categories. In a significant proportion of cases assessing successful settlement by labour market status is inappropriate, unnecessary, and in some cases not possible. Second, although labour market outcomes are emphasised by government and others this emphasis is not mirrored by immigrants. For example, in LSIA2 only 38.2% of immigrants gave at least one economic reason for coming to Australia (and interestingly given the policy changes between LSIA1 and 2 this was not remarkably different to the 31.2% for LSIA1). These proportions partially explain why the relationship between satisfaction and labour market status is found not to be particularly strong. The stated aims of Australia s immigration policy may be based on economic outcomes, and selection policies can be aimed at this outcome, but 50% of immigrants still enter Australia on non-economic visas and for these immigrants especially (but by no means exclusively) labour market outcome may be an inappropriate measure of successful settlement. Third, although at a point-in-time in aggregate the proportion of immigrants reporting they are at least satisfied with Australia is very high (for example, 92.8% in LSIA22) as is the proportion indicating their decision to immigrate to Australia was correct considering the gross change in satisfaction levels across waves, as opposed to aggregate net responses, suggests that there is more information available from this question, and more avenues for analysis, than appears when considered as a net response at a point-in-time. Note that such a high survey response to questions relating to satisfaction is not unusual. For example, in answer to a similar question in the first wave of the HILDA (scaled from 0 totally unsatisfied, to 10 totally satisfied) 97.7% reported that were in the 5-10 range (Shields & Wooden 2002). Fourth, other measures in the LSIA that are related to satisfaction with Australia may provide some guidance or insights as to the reliability of the question regarding satisfaction. For example, in LSIA21, although there is a statistical relationship between the level of satisfaction and the intention to apply for Australian citizenship, the correlation coefficient is only (t-statistic 8.083, p-value 0.000). This is 7

8 driven partial by the fact that 62.8% of those who were not satisfied with Australia intended applying for citizenship, and partly by the fact that 58.7% of reasons given for not applying for citizenship (multiple answers accepted) were that the immigrant wanted to maintain citizenship of their former country. Only 10.3% of responses were that the immigrant had not yet decided to stay in Australia. Similar results are found when considering immigrants intentions regarding encouraging others to immigrate to Australia (Section ), and intentions relating to leave Australia permanently (Section ) although there is a statistical relationship between these measures and satisfaction the correlation is not particularly strong, suggesting that the reasons given are not driven by dissatisfaction with Australia but by other considerations. Fifth, the changes to immigration policy (see for example Richardson & Lester 2004a) between LSIA1 and LSIA2 are associated with changed levels of satisfaction for those visa categories affected by the changes, but not the others. The policy change also altered the proportions of those reporting various levels of satisfaction, but changes were not particularly large. This result is not surprising; the change in policy would have further influenced immigrant self-selection, with those who previously may have made a marginal decision in favour of emigration to Australia being more likely to decide not to move. An important sixth conclusion is that if there is congruence between those factors that lead to labour market success for those immigrants who enter the labour force, and the factors that contribute to immigrant satisfaction (and potentially success from their perspective), then immigration policy is perhaps by a happy coincidence doubly appropriate. If there are contradictions, this requires further examination and consideration of future policy objectives and outcomes. In summary, this Paper discusses factors that contribute to immigrants satisfaction with Australia. It finds that labour market outcomes matter, but they matter to some immigrants more than others, and according to LSIA data, they do no matter to some immigrants at all. If only labour market outcomes are considered as a measure of successful settlement government policy design has been emphasised but applied as a test of success to immigrants to whom such a test is of little concern, or simply of no interest. Immigrants satisfaction with Australia, and their decision to immigrate to Australia, is for such immigrants an appropriate indicator of success and this appears to be, at best, given only superficial consideration by government and in academia. 8

9 This Paper uses exploratory data analysis techniques to ascertain whether there is a statistical relationship between satisfaction and other immigrant attributes that theory or previous empirical evidence have suggested are associated with satisfaction; statistical significance of correlation is tested for where appropriate. To keep the Paper manageable, disaggregation of variables is generally kept to a minimum, and often only one cohort or wave is discussed. It is expected that if there is a relationship in the aggregate data, or in a single wave, it will also exist for other waves and at some level of disaggregation What Does Successful Settlement Mean? Of primary importance to policy makers (as representatives of the Australia population), and to immigrants is whether immigrants have settled (or will settle) successfully into their new country. It is not necessarily so that successful settlement will be defined by each party in the same way and so success must be viewed from two perspectives. Immigration is known to be associated with numerous sources of stress: uncertainty about the future, financial stress (particularly if employment is to be sought), and emotional stress due to initial exclusion from general society (Lev-Wiesel & Kaufman 2004). To address the issue of successful settlement it is necessary first to define what constitutes success and how it is to be measured. In essence, successful settlement is complex and must include multiple factors. Moreover, the LSIA data cover only a short period (for example, LSIA2 includes information relating to just 18 months). Thus, there are dimensions to immigrant s satisfaction that are not, and could not be, captured in the LSIA data (or elsewhere) and consequently the data will provide only part of the explanation for individual satisfaction and success. This is a constraint that cannot be overcome, but one that must be accepted. Similarly, it is important to note that capturing the many contributors to satisfaction and success is not possible. For example, happiness can be influence by both internal and external factors. An individual who suffers from endogenous clinical depression is not likely to be happy due to internal factors which may be genetic; or an individual who has been injured by being struck by lightening is not likely to be happy. Notwithstanding best efforts, many influences on an individual s happiness may not (or cannot) be quantified and hence modelled, and this is also true for successful settlement. 9

10 As discussed below (perhaps for want of a better measure) a common criterion used by government to assess successful settlement is labour market success, but as noted previously, labour market success, from the perspective of individual immigrants may by only one matter of interest and it may be little (or possibly no) interest to some immigrants. Jupp, McRobbie et al. (1991), in a seminal text on settlement and emerging ethnic communities, outlined requirements for measuring successful settlement based on a minimalist (achieved within two or three years) or maximalist approach (which may take a lifetime to achieve). The minimalist approach requires that an immigrant: secures accommodation and employment; does not become a charge on the public purse; does not have a physical and psychological condition that inhibit employment; are not alienated from society to the extent of becoming a social menace; and can socialise with workmates and neighbours. This approach assumes major importance to labour market success. The maximalist approach suggests an immigrant is settled when they are employed at the same level of qualifications and experience as in their former country of residence (FCR); when their lifestyle is better than in their FCR; they have full command of English; they have full access to public services available to all; they have taken out citizenship and intend to remain in Australia; and they do not feel discriminated against once again, the labour market is an important feature of successful settlement. According to Jupp, McRobbie et al. (1991) the maximalist approach is the ideal way to measure successful settlement, as this approach is premised on the principles underlying multiculturalism: access, participation, and equity that is, the focus is on inclusiveness not upon assimilation. Nonetheless, due to shortness of time covered by the LSIA it is not possible to reasonably assess settlement success by the maximalise approach and assessment by the minimalist approach does not appear to focus sufficiently on non-labour market factors. Cox (1996) suggests that settlement is the period during which immigrants need to find accommodation, organise a source of income, develop communications with current residents, and organise a suitable social life once again defined too broadly to measure completely. Others define successful settlement more broadly Adelman, Borowski et al. (1994) suggests successful settlement is related to long-term equality of access to resources and power, and to self-reliance; Wooden, Holton et al. (1994) 10

11 equate successful settlement to immigrants sense of security and optimism. Such indicators of successful settlement inform the debate but do not allow successful settlement to be measured: first, they are too broad to fully measure, and second, there is not a benchmark group against which to measure the relative progress of immigrants. Khoo & McDonald 2001) address the factors associated with positive settlement outcomes, with reference to the first cohort of the LSIA. Their approach is based on the requirement of DIMA that settlement indicators be compared against an Australian-born benchmark population (or where relevant the total overseas-born population with more than 10-years residence or the total Australian population) and by assumption that settlement success is achieved when migrants reach parity with the reference population. Based on assessment of previous studies Khoo & McDonald (2001) suggest that there are four dimensions if immigrant settlement: social participation (measured by English proficiency, and citizenship status); economic participation (employment status); economic well-being (home ownership and income); and physical well-being (mental health status, and self assessments of general health). Analysis shows that the indicators of successful settlement are correlated with immigrants characteristics (which themselves may also be an indicator of successful settlement) such as: age, sex, educational qualifications, health status, employment status, English language proficiency, and visa category. Khoo & McDonald (2001) conclude that settlement success and the roles of social and economic participation and wellbeing are complex issues requiring further exploration, but that the four dimensions of immigrant settlement outcomes are highly correlated forming an interlinked system which presents significant problems in differentiate between their relative importance. It is worth exploring whether one way to summarise the success of immigrants is to equate success with satisfied with their life in Australia, with their decision to immigrate to Australia, and their intention to remain permanently in Australia. Adopting a single criterion however may not be the best, or only, indicator as it fails to considerer other matters of interest to immigrants (and fails to take advantage of the complete set of indicators provided in the LSIA). Nonetheless, for the purpose of 11

12 traditional econometric analysis (to be reported on in a forthcoming Paper) it is necessary to either choose a single measure of success or to construct an index of success to be the dependent (endogenous) variable in regression analysis. Therefore, one of the focuses of this Paper is to investigate immigrants satisfaction, and to consider its appropriateness as the single representative measure of success. In addition, consideration is also given to the potential for construction of an index of success (that is, a more encompassing measure which allows several simultaneous indictors of successful settlement to be combined), based on several indictors Difficulties in Defining and Measuring Success Governments aim to influence population characteristics through immigration policies where intervention is to regulate arrival (and departures) of foreigners, according to their nationality, the purpose of their visit (for example, settlement, temporary work, or holiday) and the duration of their stay (Doomernik 1998). There is little agreement on how best to measure settlement success, or integration and satisfactory outcomes for immigrants. Reasons for this include the difficulty in defining settlement and integration; a paucity of data; and the lack of knowledge regarding the interconnection between longer-term integration and shorter-term indicators of settlement outcome. For example, integration is a widely used term, but unless the sense in which it is used is specified there is little likelihood that comparisons are legitimate. Similarly, Doomernik (1998) in work for the ILO Migration Programme, states that it is important to spell out what is meant by the term integration, suggesting that in its broadest sense it indicates the situation where immigrant s position in society are similar to those of non-immigrants. According to Doomernik (1998) any method aimed at providing tools that will assess effectiveness of immigration polices should make clear the link between immigration and integration into society. Such tools should rely on a set of agreed indicators that are reliably available but as yet no such methodology has been developed it is suggested that it is necessary to rely on a collection of indicators that are be considered suitable, but imperfect. Although there has been considerable research on measuring and comparing international immigrant settlement outcomes it has predominantly related to assessing 12

13 the outcomes for refugees. Fletcher (1999) suggests that the comments on that work however could also relate to other groups of immigrants. Indicators of success should also be benchmarked; they are subjective measurements in comparison with some pre-defined group. For immigrants, this would generally be native or non-immigrant residents with similar attributes, but within the LSIA these comparative data are not available. Comparison with non-lsia data is potentially possible but there are numerous difficulties in constructing a synthetic group from, for example, Census data. For example, approximately 20% of the Australia population has a disability resulting in lower labour force participation and wages, and reduced occupational choice and hours worked (Wilkins 2004). In theory, recent immigrants should not be disabled, but 7.5% of immigrants over age 65 in LSIA21 reported long-term health conditions. Generally, the Census data would have to be adjusted for age, gender, education, language, country of birth, health, marital status, number of children, and household membership. In addition, Census data are not available for the date at each wave of interviews. This leads to the conclusion that the LSIA does not provide a comparison sample of native-born individuals. As the lack of benchmark data in the LSIA does not allow comparisons between immigrants and similar non-immigrant individuals assessment of whether settlement is successful must be by other means. This can be by an assessment of the immigrants position. As guidance, there appears to be some consensus about those factors that are important indicators of successful settlement, but once again many of these factors are not provided in the LSIA. At a high level factors indicating success can be as broad as achieving physical and emotional security, and a sense of self-esteem; achieving aims and ambitions; and acceptance by the wider community (ACRMA 1981). Such measures are, however, rarely available for analysis and are not measured in the LSIA. At a medium level factors can be classified as economic, social, cultural and political (CCR 1998), and some of these can be measured and collected in surveys. In the LSIA there are few such measures (excluding economic measures), and where they are measured they may not cover more than one wave so that progress can not be measured. For example, in wave 2 of LSIA2 (Section I), some questions are asked about participation in the community but similar questions are not asked in wave 1. 13

14 The responses cannot provide information regarding changes in immigrants; participation. Thus, in wave 2, 14.5% of PAs did not talk to people who lived in their street or near them, but over 70% spoke to between 1 and 10 such people but there is no way of knowing if this indicates an improvement in meaningful participation since their previous interview nor does it allow comparison to other people in their street (or elsewhere). At the lower level factors or indicators are classified as labour market, financial, language, citizenship, accommodation, and sponsorship (Fletcher 1999), and it is these that are generally available for analysis. It is these lower level classifications that lend themselves to analysis of the LSIA data. Work by Khoo & McDonald (2001) has recently provided a useful summary and guide to some factors that can be used to assess successful settlement of immigrants to Australia Social Integration One general measure of successful immigration is social integration, which is of significant importance to the level of satisfaction recorded by immigrants (as with all groups who are minorities). This does not necessarily imply that society is integrated to the extent that it is homogeneous. Integration may be in the form of attachment to a particular ethnic or religious sub-group that has formed by choice. There is a tendency for immigrants to seek to live in areas in which there is a resident national or cultural presence. Since this common to all immigrants (OECD 2003) it can be concluded that there are benefits for immigrants. Benefits can include both practical and psychological matters for example, cultural support and improved job prospects. Fennelly and Palasz (2003) suggest limited English ability is a major barrier to integration. Thus, access to services is impeded, and workers are relegated to the secondary labour market. Empirical research shows that these effects persist after controlling for education, age at immigration, employment, and other factors and there is some evidence that gender differences may mask some differences in English proficiency. There are two indictors that social integration may not be a particular problem for immigrants in the LSIA. First, in response to the question asking about things disliked 14

15 about Australia, 5 relatively few immigrants indicted a problem (for example, only 14% in LSIA22 responded and the largest category was the weather (6.3%) followed by isolated and remote (3.5%) and taxation (3.3%)). When considering immigrants perceptions of life in Australia the responses to questions 6 suggests that, for major social indicator, immigrants view Australia as superior to their former country of residence (Table 1-3) (see also Richardson, Healy et al. 2004b). Second, a lack of social integration could be expected to reduce the level of satisfaction with Australia and given the high level of satisfaction recorded by immigrants it appears that social integration is not a significant problem to immigrants in the LSIA. Table 1-3 Perceptions held by Migrants about Various Aspects of Life in Australia (%) Proportion of Each cohort/wave cohort 2 cohort 2 cohort 1 cohort 1 (Question from LSIA) wave 1 wave 2 wave 2 wave 1 I.3 Religious Tolerance Lot Some Little I.5 Contact between Cultures/Nations Lot Some Little I.6 Race/Culture/Nationality Tolerance Lot Some Little I.8 Racial Discrimination Lot n.a. 4 6 n.a. Some n.a n.a. Little n.a n.a. I.9 Religious Discrimination Lot n.a. 2 2 n.a. Some n.a n.a. Little n.a n.a A Policy Perspective on Immigrant Success As described in Paper [3], Australia s visaed immigration policy has three main aims. First, to attract economic immigrants: that is, immigrants who are expected to contribute to human capital growth, who enter Australia in the Skilled Australian Sponsored, 7 Independent, and Business skills/ens visa categories. Second, to accept 5 For example, question V.4 (LSIA22). 6 For example, Section 1 (LSIA22). 7 Previously called Skilled Australian Linked, which replace the Concessional Family category between the arrival of LSIA1 and LSIA2. 15

16 certain non-economic immigrants including Family stream immigrants (who enter in the Concessional family, Preferential family/family stream); and third, Humanitarian entrants. As with several other countries, Australia seeks to influence the socioeconomic characteristics of its immigrants by a point scoring system which ranks potential immigrants. On the other hand, immigrants chose Australia as a destination due to, amongst other things, what Australia offers in comparison with their current place of residence and other potential destinations (Bell 1997). Notwithstanding that a significant proportion of immigrants to Australia are non-economic the view is perpetuated that employment is the appropriate measure of success (see for example, Marston 2004). There were a number of important changes to Australia s immigration policy between the LSIA1 and the LSIA2 that were designed, amongst other things, to reinforce Australia s emphasis on economic immigrants. Policy changes increased the continuing economic focus of Australian governments, with an increased emphasis on the labour market outcomes of immigrants (Hawthorne 2004), by strongly favouring younger, English language proficiency, post-secondary educated applicants with experience in particular occupations (considered to be in short supply) or with skills in high demand. Essentially, immigration in the noneconomic group were not affected by changes in policy other than the important reduction in the proportion of these immigrants whereas economic immigrants faced significant changes Government reports acknowledged the increased emphasis on an economic approach to immigration, for example: Since 1995/96 there have been significant changes to migration policy to increase its economic focus. The level of Family migration has been cut, and Skill Stream migration and the general points tests have been refocused on employability attributes (Econtech 2001, p.i). Reports also recognise the ongoing adjustments in favour of successful labour market outcomes:...the strengthening of the points test requirements relating to skills, age, and English ability, introduced from July 1999, makes clear the 16

17 increased emphasis on skills adopted by the Australian government (Richardson & Lester 2004a, p.15). This increased emphasis on an economic view continued to be publicised. According to the (then) Minister for Immigration and Multicultural and Indigenous Affairs in commenting on Australia s immigration program: The emphasis must continue to be a program which provides tangible economic, demographic and social benefits (DIMIA 2004b, p.iii). Birrell (2003) confirms the direction of Australia s policy emphasis when he concludes (tentatively) in his assessment of immigration and the labour market that the evidence he examined supported the hypothesis that Australia s immigration policy was tied to immigrants labour market success. Chiswick & Miller (2004) also support this view when discussing changes to the selection criteria. Moreover, government have used the successful labour market outcomes of immigration as an indicator that immigration policy has been successful (Hawthorne 2004). Between LSIA1 and LSIA2 the change in policy emphasis appears to have resulted in a significant change in the proportion of economic and non-economic immigrants. This is most noticeable in the increase in economic immigrants from 35% of the total in LSIA1 to 50% in LSIA2, but the fall in the proportion of non-points tested immigration is also an important result of changing policy. Table 1-4 Immigrant Visa Categories LSIA1 and LSIA2 (%) LSIA 1 LSIA 2 Economic Family Humanitarian 16 9 It should also be noted however, that, notwithstanding the Government s stated emphasis on skilled immigration, as a proportion of permanent immigration visas issued, Family stream immigration fell to 36% in (from 41% in ) but has fallen little since. It averaged 35% in the four years following the LSIA2, and increased its share of the total by about one percentage point in the 12 months to

18 Thus, from a policy perspective, success is indicated for immigrants, other than dependent immigrants, by their rapid movement into employment (either as wage or salary earners or in their own business): Finding a job is a crucial step in successful settlement into Australian life for migrants who are not dependent family members (Richardson, Miller-Lewis et al. 2002b, p.5). This view of immigrant success mirrors that at the international level. For example, in an overview or the successful integration of refugees the UNHCR indicated that lack of employment was the foremost constraint (Marston 2004) An Immigrant s Perspective of Success If the immigrant s reasons for immigration to Australia were solely or predominantly economic, from their perspective success may also be defined principally by labour market success. For immigrants who have only, or predominantly, non-economic reasons for immigration success defined only by labour market status is not appropriate. This is particularly the case for immigrants in non-economic visa categories, and such immigrants make up a majority of the LSIA. For example, in LSIA22, overall only about 38% of immigrants gave an economic reason for immigrating to Australia and in the Family stream 79% of immigrants gave only non-economic reasons for immigration (see Table 1-8 below). There are also a significant proportion of immigrants in both LSIA1 and LSIA2, in all visa categories, who were not in the labour forces. For example, at wave 2, 39% of LSIA1 and 33% of LSIA2 were not in the labour force (see Table 1-5 below). For such immigrants, employment cannot be used as a measure of successful settlement, and their responses to the questions regarding their level of satisfaction and other questions regarding their circumstances are more appropriate as indicators of their successful settlement. To emphasise this point, at any one time about one-third of the total visaed offshore immigrants in the LSIA are nonparticipants, or not in the labour force (NLF). For example, as shown in Table 1-5 based on the LSIA data, 13,389 persons (or 33.4%) of the total of 40,045 offshore Program visaed immigrants who arrived in Australia during the period September 1999 to August 2000 and were interview at wave 2, reported that they were NLF. 18

19 Table 1-5 Current Main Activity in Australia LSIA22 (%) Current Main Activity All Immigrant Off-shore Visaed Immigrants PA Only Wage or Salary Earner , Conducting Own Business 7.1 2, Other Employed Unemployed 6.5 2, NLF , Student 7.6 3, Home Duties , Retired, Pensioner 3.9 1, Other Total , Whilst it is not possible from the LSIA to know if a meaningful proportion of these immigrants were involuntarily out of the labour force (but noting that unemployed immigrants are recorded as such) considering the transition table for main activity (including labour market status) from wave 1 to wave 2 is instructive. According to Table 1-6, 18.6% of immigration remained as Home duties or Retired, Pensioner, Other ; 4.4% remained as Students ; and 5.6% moved between these three categories. That is, 28.6% of immigrants were NLF in both waves of LSIA2. Based on this data, over 11,450 immigrants who arrived in a 12 months period did not seek employment within up to 18 months of arrival. Table 1-6 Transition in Current Main Activity in Australia LSIA2, Wave 1 & 2 (%) Labour Market Wage, Conduct Other Unemp. Student Home Retired, Total Status Salary Own Employ Duties Pension, wave 1 wave 1/ wave 2 Earner Business Other Wage or Salary Earner Conducting Own Business Other Employed Unemployed Student Home Duties Retired, Pensioner, Other Total wave Notes: (1) Entries enclosed in a box are those immigrants who were NLF in both waves of LSIA2. (2) Represents cell contents too small (or zero) for analysis. 19

20 As an approximation, this proportion can be extrapolated to later immigrant intakes. In May 2002, the Government announced that the number of annual places (for offshore visaed immigrants would be set at about 117,000 (Ruddock 2002; Econtech 2004). The government also confirmed in March 2003 that until barring exceptional circumstances the size of the annual immigrant intake would remain at that level (Ruddock 2003a; Ruddock 2003b). On this basis Econtech (2004) estimated that there were 135,700 permanent (off-shore and on-shore) arrivals in , and this would continue until Of the total, about 118,100 places are permanent arrivals in the Migration Program (and the remaining 17,600 are non-program arrivals). If, as for LSIA2, about 28% of these immigrants are NLF then it is possible that over 30,000 permanent immigrants per year from Australia s Migration Program will remain out of the labour force for between 12 and 18 months at least after arrival. This suggests, on a cumulative basis, a significant number of visaed immigrants to Australia, whether economic immigrants or not, are not in the labour force for extended periods if not indefinitely. Consequently, labour market outcomes are irrelevant for testing successful settlement notwithstanding the government s emphasis on immigrants adding to Australia s stock of (useable and employed) human capital. Thus, Ho and Alcorso (2004, p.238) criticise current research as it is: dominated by studies presenting a success story narrative of recent, mostly highly skilled, migrants achieving increasingly positive outcomes in the Australia labour market. For the reasons outlined above, although labour market outcomes can be considered as an indicator of successful settlement for some immigrants, this is not so for significant numbers of immigrants. To inform policy makers, and other interested parties, other measures are necessary to assess successful immigration from the immigrants perspective. One such avenue is immigrants satisfaction Why Does Immigrants Satisfaction Matter? Considering whether satisfaction matters starts from the premise that: Most societies see happiness as a desirable state of affairs for their people [even if] Increasing individualism changes the focus from society to the individual (Marks & Fleming (2004, p. 301). 20

21 If as is generally accepted, individuals seek to reach the highest level of satisfaction that they can (that is, maximize utility) then: understanding and analyzing the determinants of satisfaction over a population seems a necessary condition to understand human behavior (van Praag, Frijters et al. 2000, p.4). If the behaviour of immigrants is understood, it may be possible to increase their level of satisfaction, and contribute to higher levels of satisfaction for immigrants to come, and so contribute to their remaining in Australia; adding to Australia economically and socially; and help encourage the inflow of further high quality immigrants either through chain migration or new chains by enhancing Australia s reputation as an immigrant destination. At a practical level, Australia s immigration policy aims, amongst other things, to increase Australia s population but: The phenomena of falling fertility rates and aging populations (resulting in amongst other things increasing dependency ratios) is not confined to Australia and the number of countries attempting to solve this growing problem by seeking to attract young skilled migrants will continue to grow (Richardson & Lester 2004a, p.40). Moreover, Productivity Commission (2004) research estimates that Australia s population distribution will (continue to) change radically in the coming 40 years. In 1925, only about four% of the population was aged 65 years or above, in 2000 the proportion had risen to about 12%, and the Productivity Commission estimates are that it will have reached 25% by 2045 (see Figure 1-1). 21

22 Figure 1-1 The Changing Age Structure of the Australian Population, Age male females males females males females share of population (%) Source: Productivity Commission (2004, p.xx). Although immigration may not be the solution to the problems caused by an aging population, (including importantly the reduction in the labour force participation rate and increased Government budget pressure) it can contribute in some small way if successful, satisfied, immigrants remain in Australia; and if those immigrants in some way encourage further immigration. Moreover, others express the view that external sources of labour may be important for aging populations. When discussing the issues that emerged during the World Economic Forum in 2005 Gottliebsen (2005, p.38) wrote, India will emerge as a provider of labour to solve shortages caused by ageing populations. Although the Productivity Commission (2004) recognises the population issues, it suggests that immigration will have little impact: Population policies are not likely to reduce the effects of ageing. Only very large (and increasing) migration flows could make a difference. Even existing migration targets may be frustrated, with other ageing countries competing with Australia for skilled migrants (Productivity Commission 2004, p.xliii). Nonetheless, according to Birrell (2003) the main objective of immigration policy, during the period between the Second World War and the late 1960s, was population growth for national defence (see also, Teicher, Shah et al. 2000). Between the 1960s and 1970s the rationale for immigration became population growth for economic reasons such as self-sufficient industries, and an expanding domestic market. Further, 22

23 in the 40 years to 1983, immigration was responsible for 40% of Australia s population growth (DIEA 1985). In , 52% of Australia s growth in population was from net overseas immigration, and as at 30 June 2002, approximately 23% of Australia s resident population were overseas-born (ABS 2004). That is, although the Productivity Commission (2004) suggests that immigration may have only a minor impact on Australia s future population size and age structure, it has often been an important part of Australia s long-term population and economic policy and it is one component of labour market growth which policy makers control directly (Cobb-Clark & Connelly 1997, p.670). Moreover, in a comprehensive review of the literature on the impact of immigration on Australia s population McDonald & Kippen (1999) conclude that given current trends in fertility and mortality the first 50,000 to 100,000 migrants may have a worthwhile impact on reducing the ageing of the Australian population. Immigrant satisfaction matters because the more satisfied the immigrants with their life in Australia the more likely they will remain in Australia, become active participants in Australian society and, for those in the labour force, productive participants. In addition: It is also important, of course, for the wellbeing of the migrants themselves [and] word-of-mouth of satisfied migrants may also be a powerful tool in attracting further migrants (Richardson, Healy et al. 2004a, p.75). There are also practical reasons for being concerned about levels of immigrant satisfaction. There is evidence that unhappy individuals do not participate fully in the labour market, reducing productiveness and cause financial distress (and increase reliance on Government assistance where available) (see Section 1.17). There is also evidence that illness may result causing further financial burden on families or Government (Borjas 1999b; Econtech 2004). As is well recorded, there has been an ongoing debate about the benefits to the host country from immigration; for example, Borjas (2000) refers to work by Sir John Hicks in 1932; and debate continues (Gartner 1996; Econtech 2001; Garnaut, Ganguly et al. 2003; Withers 2003; Econtech 2004) and see Borjas (1999a) for a comprehensive survey. For example, do immigrants take jobs from current residents 23