THE RESULTS SECTION - QUANTITATIVE ANALYSIS
7.1 Introduction
In this section the reader will find a detailed discussion of the analysis of the 1991 ABS one percent sample. While the previous section provided a literature review and some preliminary findings, this section will proceed with a detailed statistical analysis of the data. Ordinary least squares and logistic regression models will be developed to explore the relationship between unemployment, hourly income and a range of predictor variables. An analysis of the 1986 census total AI working population is also provided.
7.2 Participation Rates
Overall, it was believed that AIs would have lower participation rates compared to the UKDs and ADs. This view was not supported with AIs on average having higher participation rates. It appears that those researchers that have suggested similar or higher participation rates for immigrants to the Australian born (BLMR, 1986; Junankar and Pope, 1990; Brooks and Volker, 1983; Wooden and Robertson, 1989) are correct in regard to the AIs.
Table 7.1 Participation rates for AIs, UKDs and ADs compared. 1991 One Percent Sample.
AI UKD AD Rate N Rate N Rate N
ANGLICAN 68.48 92 56.96 5518 62.39 21389 AT16 74.17 120 79.34 3141 32.49 3201 BORNASIA 71.46 410 50.67 75 63.41 82 BORNAUST 68.63 51 52.85 2704 64.75 69571 CATHOLIC 72.75 345 63.96 1795 66.32 17216 DEGREE 87.50 40 83.43 688 86.13 3851 DIPLOMA 72.00 25 71.16 801 78.80 3873 FEMALES 62.60 254 47.94 6510 54.08 35960 FOREIGNQ 83.58 67 83.33 1080 77.78 27 HIDEGREE 75.00 16 86.85 251 88.72 1312 MARRIED 75.00 308 61.24 8067 66.12 38479 NSW 80.41 97 58.97 3624 63.67 24084 PERTH 66.67 96 59.95 1895 65.10 3327 TRADEQ 84.62 52 76.71 2181 83.11 9845 VIC 70.71 198 61.46 2662 65.14 16794 TOTAL 1991 71.27 479 59.72 12888 64.69 70153 TOTAL 1981 68.22 264 57.89 7171 62.38 38331
Participation rates have been calculated for all respondents 15+ years of age. However, there are four categories where AIs had lower participation rates compared to UKDs and ADs. These were, HIDEGREE, DEGREE, DIPLOMA and FOREIGNQ. These findings are consistent with the work of Stromback et al. (1992) who found that the more highly educated the immigrant the lower their participation rates. The findings, with regard to overseas qualifications are consistent with research that employers place less value on qualifications from NESB countries.
The AIs with the highest participation rates were those with a trade qualification, while for UKDs and ADs, it was for those with higher degrees. While no one group had consistently higher participation rates in all categories than the others, on average the AIs had the highest participation rates followed by UKDs and ADs. The most interesting finding was that participation rates had increased for AIs between 1981, when it was (68.2 percent), and 1991, when it increased to (71.2 percent).
Work conducted by Young (1995) indicated that once labour force participation rates were standardised by age, immigrants had substantially lower participation rates. When this age standardisation was performed with the AIs, their participation rates remained high, going against previous findings (Ackland and Williams, 1992).
Table 7.2 Participation Rates for AIs, UKDs and ADs Compared by Age Group. 1991 Census One Percent Sample AI UKD AD Age Rate N Rate N Rate N
15-19 37.1 35 54.9 541 53.8 7750 20-24 73.2 41 83.3 855 85.4 7120 25-34 90.2 184 82.4 2392 81.2 14646 35-44 93.1 194 84.4 2416 83.0 14189 45-54 84.9 136 81.5 2054 78.5 9532 55-59 71.4 44 60.3 912 58.3 3428 60+ 20.7 112 11.7 3718 14.7 13488 TOTAL 71.2 479 59.7 12888 64.7 70153
From Table 7.2, it can be seen that participation rates for AIs between the ages of 15 and 24 years, are lower than for the UKDs and ADs. After the age of 25, the AIs have much higher participation rates relative to the other two groups. Finally, after the age of 60, the AI participation rates remain substantially higher than that for the other two groups. The higher participation rate for the oldest AI age group can be explained by the need for financial security. Those AIs who immigrated to Australia late in life would need to keep on working for as long as possible so as to pay off mortgages, one of the main markers of having made it in the new country (Lobo, 1988).
7.3 Unemployment
Unemployment rose rapidly between 1975 and 1983, doubling from (4.6 percent) to (9.9 percent) of the labour force (Jones, 1992b: 131). At the time of the 1991 Census, the general rate of unemployment had gone even higher, to around (12 percent). The figure observed among the ADs in the ABS one percent sample used in this study was about (10.5 percent). Aggregate unemployment was somewhat higher among UKDs at (11.3 percent). The AIs had the highest aggregate unemployment of the three groups at (11.7 percent).
The AIs appear to have substantially higher unemployment levels in most of the categories explored. But it should be emphasised that sample sizes are small and as such may affect the extension of these findings to the general population.
Those AIs with higher degrees and diplomas appear to be of most concern, with substantially higher unemployment rates than UKDs and ADs. AIs with post-graduate qualifications have an unemployment rate of (8.3 percent) that is 5.5 times that of ADs at (1.5 percent). Further, those AIs with degrees have an unemployment level of (5.7 percent) which is 1.7 times greater than ADs at (3.4 percent). UKDs have unemployment levels that are intermediate to the AIs and AD unemployment rates.
Table 7.3 Unemployment Rates for AI, UKDs and ADs By Variable. 1991 One Percent Sample.
AI UKD AD Unemp. N Unemp. N Unemp. N Rate Rate Rate
ANGLICAN 15.9 63 10.5 3143 10.7 13345 AT16 9.0 89 10.9 2492 18.3 1040 BORNASIA 10.2 293 7.9 38 3.8 52 BORNAUST 25.7 35 14.8 1429 10.4 45050 CATHOLIC 10.8 251 11.2 1148 10.0 11418 DEGREE 5.7 35 5.4 574 3.4 3317 DIPLOMA 16.7 18 3.7 570 3.7 3052 FEMALE 9.4 159 9.2 3121 9.1 19448 FOREIGNQ 17.9 56 9.2 900 4.8 21 HIDEGREE 8.3 12 3.2 218 1.5 1164 MARRIED 7.4 231 7.5 4940 5.8 25441 NSW 10.3 78 9.8 2137 10.5 15335 PERTH 15.6 64 12.5 1136 11.9 2166 TRADEQ 11.4 44 10.5 1673 7.8 8182 VIC 10.7 140 10.5 1636 10.1 10939 TOTAL 11.7 341 11.3 7697 10.5 45376
Another worrying feature of the unemployment statistics for the AIs was for those with overseas qualifications. While AIs with a qualification, whether a post-graduate qualification or a certificate had lower than average unemployment levels, those AIs with an overseas qualification had unemployment levels that were substantially higher than average. AIs with an overseas qualification had unemployment rates of (17.9 percent), about 3.7 times higher than for ADs (4.8 percent) with overseas qualifications.
7.3.1 Modelling Unemployment
To get an idea of the type of effect that different variables have on unemployment rates, the probability of being unemployed was calculated at the mean level for each of the three groups. This was done so as to provide a reference point or "control" for group comparisons (Flatau and Hemmings, 1991: 22). The mean values used in the Logistic Regression equations can be seen in the tables that appear as an appendix.
The results reported refer to "predictions" derived from the modelling exercise. These "predicted" unemployment figures do not refute the actual data but allow the researcher to examine the data for the possibility of discrimination.
Table 7.4, below give results from logistic regressions predicting the probability of unemployment from a range of background variables. A more complete listing of the logistic regression may be found in the Appendix at the end of the thesis. Because employment is scored zero and unemployment is scored one, variables with negative signs reduce the chances of unemployment; those with positive signs increase them.
Table 7.4, shows the parameter estimates for different variables used to predict unemployment rates among the AIs, UKDs and ADs. Table 7.5, shows the effects of the parameter estimates on average unemployment for the three groups. Those variables that have large negative parameter estimates substantially reduce average unemployment levels, conversely those variables with large positive parameter estimates substantially increase unemployment levels.
Table 7.4 Measures of Effects from a Logistic Regression of Unemployment among AIs, UKDs and ADs, in descending Size of AI Effects. 1991 One Percent sample.
Variable AIs UKDs ADs
HIDEGREE -1.676 0.021 -1.478 DEGREE -1.529 -0.070 -0.901 MARRIED -1.269 -0.974 -0.977 FEMALE -0.691 -0.504 -0.411 TRADEQ -0.578 -0.338 -0.512 DIPLOMA -0.549 -0.440 -0.766 NSW -0.162 -0.234 0.045 VIC -0.137 -0.143 0.026 AUSLFE -0.045 -0.026 -0.017 FORGNLFE 0.016 0.019 -0.032 CATHOLIC 0.048 -0.159 -0.124 AUSTED 0.108 -0.206 -0.105 FORGNED 0.277 -0.193 -0.059 PERTH 0.435 -0.024 0.132 FOREIGNQ 0.600 -0.101 -1.382 ANGLICAN 0.674 -0.197 -0.007 AT16 0.767 0.009 -0.496 BORNASIA 0.892 0.057 -0.840 BORNAUST 3.480 0.377 -0.216 Constant -5.011 1.368 0.292
-2 * ln LR (a) 299.256 5933.351 32617.586 -2 * ln LR (b) 275.470 5593.167 30653.339 Model Chi-Sqr 23.787 340.184 1964.247 DFs 12 12 12 No. of Resps. 341 7697 45379
* The positive coefficients indicate an increased probability of unemployment while the negative coefficients indicate a decreased probability of unemployment.
Table 7.5 Predicted Unemployment Levels for AIs, UKDs and ADs. 1991 One Percent sample. AI UKD AD
Unemp. % Unemp. % Unemp. % Variable Rate Change Rate Change Rate Change
ANGLICAN 20.68 8.95 9.49 -1.84 10.39 -0.07 AT16 22.25 10.52 11.42 0.09 6.64 -3.82 AUSLFE 11.27 -0.46 11.07 -0.26 10.31 -0.15 AUSTED 12.89 1.16 9.42 -1.91 9.52 -0.94 BORNASIA 24.48 12.75 11.91 0.58 4.80 -5.66 BORNAUST 81.18 69.45 15.71 4.38 8.61 -1.85 CATHOLIC 12.24 0.51 9.83 -1.50 9.35 -1.11 DEGREE 2.80 -8.93 10.65 -0.68 4.53 -5.93 DIPLOMA 7.13 -4.60 7.61 -3.72 5.15 -5.31 FEMALE 6.24 -5.49 7.17 -4.16 7.19 -3.27 FOREIGNQ 19.49 7.76 10.36 -0.97 2.85 -7.61 FORGNED 14.91 3.18 9.53 -1.80 9.92 -0.54 FORGNLFE 11.89 0.16 11.53 0.20 10.17 -0.29 HIDEGREE 2.43 -9.30 11.54 0.21 2.59 -7.87 MARRIED 3.60 -8.13 4.60 -6.73 4.21 -6.25 NSW 10.15 -1.58 9.18 -2.15 10.89 0.43 PERTH 17.03 5.30 11.09 -0.24 11.76 1.30 TRADEQ 6.94 -4.79 8.35 -2.98 6.54 -3.92 VIC 10.39 -1.34 9.97 -1.36 10.71 0.25 OBSERVED UNEMPLOYMENT 11.73 11.33 10.46
See Jones (1992b: 132) for an example of how to obtain an estimate of unemployment
With regard to the AIs and ADs, it is having a qualification, being female and being married that reduces predicted unemployment the most. For the UKDs, the results are mixed, with Higher Degrees and Degrees not reducing predicted unemployment to the same extent as with the other two groups.
7.3.2 Proxy Variable for Being Light Skinned - Anglican
Hypothesis H1: that Anglicans would have lower unemployment rates than Catholics was not supported. Being a self described Anglican AI was expected to indicate the increased possibility of physical similarity to ADs, thus leading to lower unemployment rates. Those AIs who were Anglican, (15.9 percent), had much higher observed unemployment levels than average. UKDs, (10.5 percent), and ADs, (10.7 percent), had unemployment levels near average.
The predicted unemployment levels for Anglican AIs at (20.68 percent), was almost twice the observed rate. Anglican UKDs had predicted unemployment levels of (9.49 percent), and ADs (10.39 percent). Their predicted unemployment rates were close to the observed rates.
7.3.3 Proxy Variable for Being Dark Skinned - Catholic
Being a Catholic AI was used as a proxy for looking similar to ethnic Indians. Catholic AIs, (10.8 percent), UKDs, (11.2 percent) and ADs, (10.0 percent), had slightly lower than average unemployment rates. The predicted unemployment rates was about that of the observed levels, AIs, (12.24 percent), UKDs, (9.83 percent) and ADs (9.35 percent).
7.3.4 The Effect of Overseas Work Experience on Unemployment
For AIs with one year of overseas work experience there was an unemployment level of (7.50 percent) that increased slightly to (9.16 percent) after 15 years. For UKDs predicted unemployment increased from (8.63 percent) to (11.03 percent) over 15 years. For ADs predicted unemployment decreased from (8.13 percent) to (5.38 percent) over 15 years. The hypothesis H2, that AIs with overseas work experience would have higher unemployment rates than UKDs and ADs with overseas work experience, was not supported. Please see Figure 7.1.
For AIs with AD human capital, predicted unemployment with one years overseas experience was (20.23 percent), and after 15 years had increased to (23.96 percent). The hypothesis H3, that AIs with overseas work experience would have higher unemployment rates than AIs with overseas work experience and AD human capital, was not supported.
In the case of UKDs with AD human capital predicted unemployment increased from (10.73 percent) after one year, to (13.62) percent after 15 years. ADs with AI human capital, had the lowest predicted unemployment of (4.48 percent) after 1 year that dropped down to (2.92 percent) after 15 years.
7.3.5 The Effect of Overseas Schooling on Unemployment
Increasing years of overseas education appeared to lead to an increase in predicted levels of unemployment for AIs. An AI with average 1 year of overseas education had a predicted unemployment rate of just over (.88 percent), with 15 years of overseas education the predicted unemployment figure rises to (29.94 percent). For UKDs with 1 year of overseas education predicted unemployment was about (25.23 percent), after 15 years of overseas education this figure drops to (2.22 percent). After 1 year of overseas education ADs had an unemployment rate of about (7.95 percent), after 15 years of overseas education this figure had decreased to (3.64 percent). The hypothesis H4, that AIs with overseas education would have higher unemployment rates than UKDs and ADs with overseas education, was supported.
When examining the effects of transferring the human capital from one group to another, it was found that AIs with AD human capital, would actually increase their unemployment rates after 15 years by (63.62 percent) to a very high (93.56 percent). The hypothesis H5, that AIs with overseas education would have higher unemployment rates than AIs with overseas education and AD human capital, was not supported.
UKDs with AD human capital had a level of unemployment, (8.92 percent), that dropped to (0.65 percent) after 15 years. ADs with AI human capital were also well performed, with low initial levels of unemployment, (5.98 percent), that tapered to (2.70 percent), after 15 years.
7.3.6 The Effect of Australian Work Experience on Unemployment
The level of predicted unemployment for AIs with just one year of Australian labour force experience was (12.21 percent), with 15 years of experience unemployment dropped to (6.91 percent). For UKDs unemployment was at (12.49 percent) after one year and decreased to (9.05 percent) after 15 years. For ADs unemployment began at (10.5 percent) and decreased to (8.52 percent) after 15 years. The hypothesis H6, that AIs with Australian work experience would have higher unemployment rates than UKDs and ADs with Australian work experience, was not supported.
For AIs with AD capital and just one year of Australian labour force experience predicted unemployment was at a high level (34.5 percent) that slowly decreased to (21.93 percent) after 15 years. The hypothesis H7, that AIs with Australian work experience would have higher unemployment rates than AIs with Australian work experience and AD human capital, was not supported. In the case of UKDs with AD capital predicted unemployment began at (14.81 percent) and decreased to (10.82) percent after 15 years. ADs with AI human capital had the lowest predicted unemployment rates after one year of Australian labour force experience, (4.38 percent), that dropped down to (3.51 percent) after 15 years.
7.3.7 The Effect of Australian schooling on Unemployment
Increasing years of Australian education appeared to lead to an increase in predicted levels of unemployment. An AI with 1 year of Australian education had a predicted unemployment rate of just over (6.54 percent), with 15 years of Australian education the predicted unemployment figure rises to (24.08 percent). Part of the reason for this unusual prediction is young Australian born AIs. These AIs have the highest number of years of Australian education and also very high levels of observed unemployment (25.7 percent).
For UKDs with 1 year of Australian education predicted unemployment was (17.30 percent), after 15 years of Australian education this figure drops to (1.15 percent). After 1 year of Australian education ADs had an unemployment rate of about (21.63 percent), after 15 years of Australian education this figure had decreased to (6.01 percent). The hypothesis H8, that AIs with Australian education would have higher unemployment rates than UKDs and ADs with Australian education, was supported overall.
When examining the effects of transferring the human capital from one group to another, it was found that AIs with AD human capital, or means, would actually increase their unemployment rates after 15 years by (2.0 percent) to (26.00 percent). The hypothesis H9, that AIs with Australian education would have higher unemployment rates than AIs with Australian education and AD human capital, was not supported.
UKDs with AD human capital, had a very high initial level of unemployment that dropped to (5.50 percent) after 15 years. ADs with AI human capital were the best performed of the six groups, with low initial levels of unemployment, (4.67 percent), that tapered to (1.12 percent) after 15 years.
7.3.8 Unemployment Summary
The effects of post-school qualifications for AIs are extremely large. For AIs, having a qualification of some sort substantially reduced their unemployment rate. In contrast ADs with qualifications experienced much smaller effects with UKDs falling somewhere in-between. For UKDs being married had the largest effect on reducing unemployment, while for ADs it was having a higher degree. There were sex differences, with women being less likely to be unemployed then men. Overall about half of the unemployment hypotheses were supported.
Having examined the issue of unemployment, the variables that effect it, and compared the unemployment rates for AIs, UKDs and ADs, the issue of hourly income will now be examined. Firstly, using the 1986 census the issue of hourly income is examined with regard to industry employed in, occupation and educational qualifications. Secondly, we test a series of hypotheses using the 1991 census.
Fig 7.1: The Effects of Overseas Labour Force Experience on Predicted Unemployment Levels
Fig 7.2: The Effects of Overseas Education on Predicted Unemployment Levels
Fig 7.3: The Effects of Australian Labour Force Experience on Predicted Unemployment Levels
Fig 7.4: The Effects of Australian Education on Predicted Unemployment Levels
7.4 Analysing the 1986 Census for Hourly Income
The analysis of the total 1986 data set was based on information obtained by the ABS. This information provided hourly incomes for AIs by Occupation, Industry and Qualification. The information for AIs was then compared to that for UKDs and ADs. Overall the ABS provided the average hourly income for 7692 AIs ($10hr). This figure was compared to that for UKDs $10hr and ADs ($9hr), using the 1 percent 1986 ABS sample tape.
The average hourly income figures indicate that AIs are doing reasonably well in Australia. This finding was unexpected given that the literature almost exclusively paints AIs as low achievers, both in terms of education and jobs. The present study differs from other studies dealing with AIs (Lobo, 1988; 1994) in that it relies on a data set that includes all AIs working full-time. For the first time, the researcher has a large and unbiased sample of AIs to work with.
It would have been interesting to have continued the 1986 analysis by building regression models for the three groups and comparing them, as has been done in later in the chapter for the 1991 One Percent Sample. High monetary costs prohibited developing regression models for the 1986 census data, since the three tables that appear below alone cost $700.
7.4.1 Hourly Income by Industry for the 1986 Census
The hourly income figures indicated that in most categories AIs earned at similar or higher levels than ADs. When the AIs and UKDs were compared, the AIs earned more than the UKDs in the areas of "Construction" and in the "Wholesale/Retail Trade". The UKDs earned more than the AIs in the areas of "Agriculture" and "Electricity", "Transport and Communication".
Table 7.6 Comparisons of Hourly Earnings of AIs, UKDs and ADs by Industry. 1986 Census, Total Data.
AI UKD AD INDUSTRY N % $ N % $ N % $
AGRICULTURE 42 0.6 5 112 2.2 6 2051 7.6 5 MINING 99 1.3 14 88 1.8 14 481 1.8 14 MANUFACTURING 1518 19.7 10 963 18.8 10 3923 14.5 9 ELECTRICITY 201 2.6 11 138 2.7 12 689 2.5 12 CONSTRUCTION 228 3.0 11 392 7.7 10 1885 7.0 9 WHOLESALE-RETAIL 1220 15.9 9 953 18.6 8 4932 18.2 8 TRANSPORT 622 8.1 10 255 5.0 11 1713 6.3 10 COMMUNICATION 297 3.9 10 108 2.1 11 642 2.4 11 FINANCE 1030 13.4 11 612 12.0 11 2845 10.5 11 PUBLIC ADMIN 677 8.8 12 364 7.1 12 2135 7.9 11 COMMUNITY SERVICES 1370 17.8 11 793 15.5 11 4076 15.0 11 RECREATION 254 3.3 8 226 4.4 8 1231 4.5 8 NON- CLASSIFIABLE 72 0.9 10 55 1.1 9 211 0.8 8 NOT STATED 60 0.8 7 59 1.2 8 307 1.1 7 TOTAL 7690 10 5118 10 27121 9
The AIs were well represented in the category "Transport" (8.09 percent), compared to UKDs (4.98 percent) and ADs (6.32 percent). This was expected, given their history of working on the railways in India (Anthony, 1969; Gaikwad, 1967; Gist and Wright, 1973). Further, the large proportion of AIs working in "Community Services" (17.82 percent), compared to UKDs (15.49 percent) and ADs (15.03 percent), was to be expected given their long history of working for the Government in India.
The unexpected finding was in the area of "Manufacturing". More AIs were working in "Manufacturing" (19.74 percent), than UKDs (18.82 percent) and ADs (14.46 percent). In India AIs would quite possibly have refused to work in a manufacturing environment where many of the jobs were "unskilled" (Gist and Wright, 1973: 63). But with immigration came a change in the social environment and a willingness to accept less skilled jobs.
The reported unwillingness of AIs to "dirty their hands" and risk their "pride" by accepting unskilled employment in India may have been exaggerated. Gaikwad (1967: 101) produces figures that suggests that about half the AIs in his study were employed in Government and Private factories. Similarly, Gist and Wright (1973: 60) report that skilled and semi-skilled AIs were employed as "mechanics, electricians, engineers, welders [and] factory operatives..." .
7.4.2 Qualifications and Hourly Income - 1986 Census
The AIs who earned the highest hourly incomes were the managers and professionals ($13hr). In the case of the UKDs ($14hr) and ADs ($13hr) it was the professionals who earned the highest hourly income. There is substantial gap in the proportion of AIs (8.55 percent) who have managerial positions compared to UKDs (11.22 percent) and ADs (14.94 percent).
The UKD professionals had the highest hourly earnings $14hr of all. They earned, on average, $1 an hour more than AI and AD professionals. Part of the reason for this, almost certainly, lies with the devaluation of Asian qualifications and labour force experience that occurs in Australia (Stromback, 1984, Tran-Nam and Neville, 1988; Jones, 1992a; 1992b).
Table 7.7 Comparison of Average Hourly Income, for AIs, UKDs and ADs by Occupation. 1986 Census Total Data.
AI UKD AD N % $ N % $ N % $
MANAGERS 658 8.6 13 574 11.2 11 4051 14.9 9 PROFESSIONALS 986 12.8 13 602 11.8 14 3064 11.3 13 PARA- PROFESSIONALS 616 8.0 12 409 8.0 12 1871 6.9 11 TRADESPERSONS 954 12.4 9 970 19.0 9 4613 17.0 9 CLERKS 2359 30.7 9 922 18.0 9 4628 17.1 9 SALESPERSONS 722 9.4 9 559 10.9 9 2712 10.0 8 PLANT&MACHINE OPERATORS 476 6.2 10 386 7.5 9 2458 9.1 9 LABOURERS 787 10.2 8 575 11.2 8 3176 11.7 8 INADEQUATELY DESCRIBED 92 1.2 9 89 1.7 11 372 1.4 9 NOT STATED 42 0.6 9 32 0.6 9 176 0.7 8 TOTAL 7692 10 5118 10 27121 9
While the AIs were under-represented in managerial positions, they were over-represented in professional 12.8 percent and para-professional 8.0 percent positions. UKDs and ADs were about a percentage point behind in these important categories. The other category that stood out for the AIs was the clerical category. About twice as many AIs 30.7 percent, as UKDs 18.0 percent and ADs 17.1 percent worked as clerks. This is consistent with the AI preference for white collar work in India. In general, the AIs earned about the same as the UKDs and more than ADs in many of the categories.
7.4.3 Hourly Income and Educational Qualifications - 1986 Census
The most interesting result, with reference to educational qualifications, is that the AIs appear to be relatively well qualified. Most of the literature dealing with the AIs, in India and England, indicate that, as a rule, AIs are poorly educationally qualified (Gist and Wright, 1973; Gist, 1975). The ABS data below indicates quite clearly, that on average, AIs are better qualified than the UKDs and ADs. There is a higher proportion of AIs with bachelor or higher degrees, 9.7 percent, than there are UKDs, 7.6 percent and ADs 7.9 percent.
Table 7.8 Comparison of Average Hourly Income, for AIs, UKDs and ADs by Qualifications. 1986 Census Total Data.
AI UKD AD N % $ N % $ N % $
HIGHER DEGREE 76 0.9 16 77 1.5 17 213 0.8 17 GRADUATE DIPLOMA 88 1.1 14 45 0.9 14 259 1.0 14 DEGREE 580 7.5 14 265 5.2 15 1679 6.2 14 DIPLOMA 409 5.3 13 214 4.2 14 1214 4.5 13 TRADE CERTIFICATE 948 12.3 10 913 17.8 10 4288 15.8 10 OTHER CERTIFICATE 1222 15.9 11 600 11.7 11 2554 9.4 10 NOT CLASSIFIABLE 38 0.5 10 14 0.3 12 88 0.3 10 INADEQUATELY DESCRIBED 516 6.7 9 215 4.2 9 1085 4.0 8 NO QUALIFI-CATIONS 3285 42.7 9 2519 49.2 9 14315 52.8 8 NOT SHOWN 528 6.9 9 256 5.0 9 1426 5.3 9 TOTAL 7690 10 5118 10 27121 9
While the AIs are better qualified than the other two groups in the study, they are often on lower hourly rates of pay. AIs with a higher degree $16hr, earn less than UKDs and ADs with higher degree's $17hr. Both AIs and ADs with bachelor degrees, $14hr, earn less than similarly qualified UKDs $15hr.
7.4.4 Summary of Findings for the 1986 Census
In general, the AIs in Australia appear to be doing reasonably well. They are a better educated and more professional group compared to the UKDs and ADs. Still, there are some areas of concern, such as the under-representation of AIs in management positions and the lower hourly earnings of AIs with higher degrees. Having now established, in general terms, the socio-economic attainment of AIs, the rest of the chapter will be focused on a more detailed and statistically sophisticated analysis of the AI situation in Australia using more recent 1991 census data.
7.5 The Model of Hourly Income Using 1991 Census Data
The model of hourly income was based on three main areas, educational qualifications and experience, job sector and experience, and region. A perusal of the means of hourly income in Table 7.9, indicates that it quite literally pays to have a qualification of some sort. The higher the level of the qualification the better the hourly wage. The highest hourly earnings were for UKD and AD respondents with postgraduate qualifications. They earned substantially more than those AIs who had higher degrees, albeit this difference was not statistically significant. On average, AIs earned more than UKDs who in turn earned more than ADs.
7.5.1 Average Hourly Income
With regard to the OLS regressions, the dependent variable, hourly income, is usually in log units (Flatau and Hemmings, 1991: 25; Jones, 1992a: 85; Stromback, 1984: 3), with the regression coefficients being "interpreted in percentage terms (Norusis, 1988: 38)." The present study rather than using log units as the dependent variable, uses hourly income in dollars. This approach is more typical of the approach sociologists rather than the economist take (Jones, 1992a: 85). Further, it is less complex to interpret the regression coefficients.
A comparison of hourly incomes appears in Table 7.9. On average, AIs earn $14.41hr, with the UKDs $13.73hr and ADs $12.84hr. The AIs and UKDs earned significantly more than the ADs on average. Further, the AIs earned more on average than the UKDs. The above hourly income figures and those that appear below are based on actual incomes from the 1991 one percent sample.
7.5.2 Examining Significant Differences in Hourly Income
In most categories AIs earn more than UKDs and ADs, although in many cases this difference is not statistically significant. Examining only significant differences: AIs who are employed in one of the secondary industries, $15.12hr, earn significantly more than ADs $12.86hr. Similarly, Married AIs, $14.97hr, earn significantly more than ADs $20.00hr. Those AIs who are Catholic, $14.31hr, earn substantially more than AD Catholics' $12.80hr.
7.5.3 Skin Colour and Hourly Income
For AIs there is a difference of $0.46hr, between the possibly lighter skinned Anglicans and darker Catholics. For UKDs this gap drops to $0.15hr. While for ADs, there is almost no difference at all, $0.05hr.
There seems to be some evidence, that in the case of AIs, a respondent's religion could be used as a proxy for differential economic performance as a consequence of skin colour. In general, the Anglican AIs were not as educated as the Catholic AIs and they held fewer managerial and professional positions. But, in spite of this they still managed to have higher hourly incomes.
TABLE 7.9 Average Hourly Earnings in Dollars and Cents for AIs, UKDs and ADs Compared in Order of Highest to Lowest AI earnings. 1991 One Percent Sample.
1 2 3 AI UKD ADs Groups Variable N Mean N Mean N Mean Different
DEGREE 47 19.53 391 18.47 2322 18.47 n.s DIPLOMA 22 19.24 386 16.77 1869 15.49 2,3 HIDEGREE 10 18.93 147 21.20 840 20.00 n.s MANAGER 30 18.86 625 16.05 3190 16.21 n.s PROF 58 17.89 1059 17.45 5622 16.60 2,3 FEDGOVT 23 16.62 355 15.59 2007 15.00 n.s STATEGVT 53 15.54 744 15.27 4334 15.28 n.s NSW 56 15.22 1389 14.57 8999 13.27 2-3 SECONDRY 51 15.12 1261 13.40 5805 12.86 1,3-2,3 MARRIED 155 14.97 3065 14.36 15851 13.59 1,3-2,3 FINANCE 32 14.82 576 15.73 3217 14.79 2,3 BORNAUST 15 14.80 811 12.30 26330 12.84 2,3 ANGLICAN 40 14.77 1901 13.55 7884 12.85 2-3 VIC 82 14.69 1006 13.84 6430 13.08 2-3 TRADEQ 31 14.68 1156 13.28 5665 12.87 2,3 FOREIGNQ 40 14.60 609 15.83 17 13.25 n.s BORNASIA 189 14.50 16 14.58 30 13.37 n.s AT16 54 14.33 1510 13.45 99 13.59 n.s CATHOLIC 159 14.31 708 13.40 6628 12.80 1,3-2,3 HRS49 23 14.25 888 12.69 5484 12.09 2,3 SERVICE 42 12.70 1242 12.01 7341 11.21 2,3 FEMALE 89 12.57 1474 12.05 8781 11.37 2,3 CLERKSAL 80 12.48 1264 12.21 7018 11.55 2,3 PERTH 38 11.90 655 13.32 1202 13.33 n.s TRADES 49 11.54 1538 11.79 8773 11.13 2,3 SELF_EMP 7 11.06 462 10.70 2671 9.00 2,3 AVEHRINC 213 14.41 4654 13.73 26493 12.84 1,3-2,3
* These are significantly different groups at the .05 level using Chi-Squares for binary data, with One Way ANOVA's and post hoc Tukey test's for Continuos data
7.5.4 The Effects of Age on Income Most comparative studies of the overseas-born and Australian born do not take into account the differential effects that age can have on the socio-economic attainment of different groups (Young, 1995). In other words, the reason that immigrants may earn more than the Australian born may be simply be due to their being, on average, older.
Table 7.10 Hourly Income for AIs, UKDs and ADs Compared by Age Group. 1991 Census One Percent Sample. AI UKD AD Age Group $ N $ N $ N
15-19 9.49 2 6.81 94 6.85 1085 20-24 11.22 16 10.73 404 10.24 3573 25-34 13.80 55 13.69 1264 13.14 7480 35-44 14.71 63 15.06 1296 14.09 7351 45-54 14.63 51 14.17 1045 14.10 4796 55-59 16.77 16 14.00 336 12.85 1215 60-64 17.14 10 11.95 215 11.25 993 TOTAL 14.41 213 13.73 4654 12.84 26493
The hourly income figures in Table 7.10, indicate that AIs don't earn more than UKDs and ADs simply because of differences in the age structure of the three populations. At every age grouping AIs earn more than ADs. Further, with the exception of the 35 to 44 year old grouping, the AIs also earn more than the UKDs. It appears then that relatively high AI earnings are more than an age related "artefact (Young, 1995: 36)" and represent the effects of greater human capital.
7.5.5 Predicting Hourly Income
Once the average income of AIs had been established in a number of different sub-categories and then compared to UKDs and ADs an attempt was made to model hourly wages. Towards this end a series of regression models were set up. The research now progresses from examining average hourly earnings, a fairly static procedure, towards a more dynamic approach. In general it is regression models that are used to observe the effects that varying certain variables have on predicted hourly income. Below are the means, see Table 7.11, and regression coefficients, see Table 7.12, that are to be used in the analyses.
The table below lists descriptive labour force statistics for AIs, UKDs and ADs. Further all persons have full-time jobs, which involves working for 35 or more hours per week, and they all speak only English at home.
TABLE 7.11 Descriptive Labour Force Statistics. Sample Means of Variables for AIs, UKDs and ADs. 1991 One Percent Sample.
1 2 3 Variable AI UKD AD Pr. Different
ANGLICAN 0.188 0.408 0.298 <.0001 Chi-square AT16 0.254 0.324 0.004 <.0001 Chi-square AUSLFE 11.604 14.736 17.586 <.0001 Anova-1,2,3 AUSTED 3.023 4.418 11.814 <.0001 Anova-1,2,3 BORNASIA 0.887 0.003 0.001 <.0001 Chi-square BORNAUST 0.070 0.174 0.994 <.0001 Chi-square CATHOLIC 0.746 0.152 0.250 <.0001 Chi-square CLERSAL 0.376 0.272 0.265 .0010 Chi-square DEGREE 0.122 0.084 0.088 .1423 Chi-square DIPLOMA 0.047 0.083 0.071 .0039 Chi-square FEDGOVT 0.108 0.076 0.076 .2094 Chi-square FEMALE 0.418 0.317 0.331 .0034 Chi-square FINANCE 0.150 0.124 0.121 .4063 Chi-square FOREIGNQ 0.188 0.131 0.001 <.0001 Chi-square FORGNED 9.732 7.408 0.037 <.0001 Anova:1,2,3 FORGNLFE 7.289 5.197 0.012 <.0001 Anova:1,2,3 HIDEGREE 0.047 0.032 0.032 .4488 Chi-square HRS49 0.108 0.191 0.207 <.0001 Chi-square MANAGER 0.108 0.134 0.120 .0233 Chi-square MARRIED 0.728 0.659 0.598 <.0001 Chi-square NSW 0.263 0.298 0.340 <.0001 Chi-square PERTH 0.178 0.141 0.045 <.0001 Chi-square PROF 0.272 0.228 0.212 .0078 Chi-square SECONDRY 0.239 0.271 0.219 <.0001 Chi-square SELF_EMP 0.033 0.099 0.101 .0044 Chi-square SERVICE 0.197 0.267 0.277 .0136 Chi-square STATEGVT 0.249 0.160 0.164 .0028 Chi-square TRADEQ 0.146 0.248 0.214 <.0001 Chi-square TRADES 0.230 0.330 0.331 .0076 Chi-square VIC 0.385 0.216 0.243 <.0001 Chi-square HRINCOM 14.410 13.730 12.840 <.0001 Anova:1,3-2,3
Number of, Respondents 213 4654 26493
Chi-Squares were used to analyse binary data (0,1). One Way ANOVA's and post hoc Scheffe test's were used to analyse continuos data.
Table 7.12 Regression Coefficients from an Ordinary Least Squares (OLS) Regression of Hourly Earnings among AIs, UKDs and ADs (Fulltime workers 1991 Census of Australia One Percent Sample. Dependent Variable Hourly Wages)
AI UKD AD Variable B Sig. B Sig. B Sig.
ANGLICAN 2.008 0.218 -0.039 0.819 0.193 0.011 AT16 2.353 0.212 0.269 0.412 -0.705 0.683 AUSLFE 0.175 0.011 0.075 0.000 0.102 0.000 AUSTED 0.446 0.245 0.694 0.000 0.568 0.000 BORNASIA 0.578 0.753 -0.627 0.637 -0.066 0.952 BORNAUST 4.355 0.194 -0.349 0.447 -0.447 0.812 CATHOLIC 1.015 0.499 -0.250 0.282 0.030 0.709 CLERKSAL 4.607 0.130 1.129 0.013 2.313 0.000 DEGREE 4.060 0.029 0.838 0.091 2.565 0.000 DIPLOMA 3.675 0.081 -0.301 0.500 0.636 0.001 FEDGOVT 2.585 0.038 1.061 0.001 1.413 0.000 FEMALE -1.175 0.235 -2.503 0.000 -1.968 0.000 FINANCE 1.869 0.119 1.732 0.000 1.549 0.000 FOREIGNQ -0.758 0.476 1.031 0.001 -0.836 0.596 FORGNED 0.529 0.129 0.779 0.000 0.519 0.000 FORGNLFE 0.115 0.085 0.026 0.125 0.068 0.533 HIDEGREE 1.873 0.503 1.074 0.145 2.744 0.000 HRS49 -2.193 0.076 -2.041 0.000 -1.367 0.000 MANAGER 7.575 0.017 3.517 0.000 5.173 0.000 MARRIED -0.517 0.542 1.135 0.000 1.008 0.000 NSW 1.142 0.320 1.328 0.000 0.673 0.000 PERTH -2.349 0.059 0.107 0.665 0.569 0.000 PROF 6.691 0.031 3.436 0.000 4.043 0.000 SECONDRY 2.637 0.013 0.338 0.172 0.579 0.000 SELF_EMP -1.937 0.347 -2.880 0.000 -3.179 0.000 SERVICE 0.408 0.707 -0.862 0.000 -0.732 0.000 STATEGVT 0.825 0.375 0.089 0.722 0.796 0.000 TRADEQ 2.863 0.016 0.572 0.005 1.048 0.000 TRADES 1.367 0.648 0.005 0.991 1.110 0.000 VIC 0.164 0.877 0.422 0.049 0.358 0.000 (Constant) -2.889 2.347 2.145
R Squared 0.442 0.254 0.295 Adj R Sqr. 0.383 0.251 0.295 N 213 4,654 26,943
7.5.6 Examining the "Significant" Parameter Estimates
Having examined average earnings in Table 7.9, we can now examine Table 7.11 and Table 7.12, before making predictions about the effects of certain key variables on hourly earnings. The earnings of AIs are dependent on a number of factors, many of which have been explored in the above regressions analyses.
Examining the significant regression coefficient's first, see Table 7.12. Being employed as a manager increases an AIs average hourly wages by ($7.58hr), being a professional by ($6.69hr), a degree by ($4.06hr), having a trade qualification by ($2.86hr) and working for the federal government by ($2.59hr). Each additional year of Australian labour force experience resulted in an increase of ($0.18hr) while working in a secondary industry increased earnings by ($2.64hr).
When comparing AIs, UKDs and ADs, in most cases predicted hourly earnings for AIs are higher than those for UKDs and ADs. Compared to AI managers, UKD ($3.52hr) and AD ($5.17hr) managers earn less all other things being equal. For professionals, UKDs ($3.44hr) and ADs ($4.04hr), achieve lower returns than AIs. A degree increased hourly wages by ($0.84hr) for UKDs and ($2.57hr) for ADs. Working for the Federal Government resulted in a ($1.06hr) increase for UKDs and a ($1.41hr) increase for ADs. Those UKD respondents who worked in secondary industries earned ($0.34hr), compared to ADs ($0.58hr).
7.5.7 Examining the Parameter Estimates as a whole
The brief discussion dealing with those regression coefficients that reached significance at the .05 level is now followed by a detailed discussion of all regression coefficients. A convention in statistical research is that if a variables makes a contribution to the level of variance at the .05 level it is considered significant and is included in the discussion. The present study did not overly concern itself with issues of statistical significance and sampling error when evaluating group differences, because of the small number of AIs in the analysis. The emphasis is rather on examining trends for the three groups included in the study.
7.5.8 Proxy Variable for Being Light Skinned - Anglican
This apparently "religious" variable was inserted into the analysis for the following reason. To try and explore the issue of labour force differences among the AIs based on skin colour. Significantly fewer AIs, (18.8 percent), were Anglican compared to UKDs, (40.8 percent), and ADs, (29.8 percent).
The regression coefficients indicated that being Anglican increased hourly wages by ($2.01hr) for AIs and ($0.19hr) for ADs. UKDs experienced a slight decrease of ($0.04hr). On average, AIs who were Anglican, earned ($14.77hr) compared to UKDs, ($13.55hr) and ADs ($12.85hr).
7.5.9 Proxy Variable for Being Dark Skinned - Catholic
This variable was inserted into the analysis for the same reasons as the "ANGLICAN" variable. About (74.6 percent) of AIs were Catholic, compared to (15.2 percent) of UKDs and (25.0 percent) of ADs.
The regression coefficients indicated that for AIs being Catholic resulted in an increase of ($1.02hr), for ADs an increase of ($0.03hr) and for UKDs there was a decrease in average return of ($0.25hr). AI Catholic's ($14.31hr) had higher hourly incomes than UKDs ($13.40hr) and ADs ($12.80hr) who were Catholics.
7.5.10 Skin Colour and Hourly Income
Anglican AIs ($14.77hr) earned substantially more ($0.46hr) than Catholic AIs ($14.3hr). Hypothesis H9, that AIs who are Anglican will have higher hourly earnings than AIs who are Catholic, was supported. Further, the difference in hourly income between the Anglican and Catholic UKDs was ($0.15hr) and for the Anglican and Catholic ADs it was ($0.05hr). The hypothesis H10, that UKDs and ADs who are Anglican will have hourly earnings that are similar to UKDs and ADs who are Catholic, was supported. It appears that religion is quite possibly a marker of differential treatment by employers in the case of AIs.
7.5.11 Years of Australian Education
The AIs had the fewest average years of education in Australia, 3.0 years compared to the UKDs 4.4 years and ADs 11.8 years. The number of years of education completed in Australia was positively related to hourly earnings for all three groups.
A perusal of Figure 7.5, shows that AIs with just one year of Australian education had substantially higher predicted hourly earnings, ($14.09hr), than UKDs, ($11.91hr) and ADs ($7.22hr). With the completion of 15 years of Australian education the UKDs now earn more, ($21.62hr), than the AIs, ($20.33hr) and ADs ($15.17hr). The hypothesis H12, that AIs with Australian education would have lower hourly income than UKDs and ADs with Australian education, was in general not supported.
There are six regression lines in Figure 7.5, the first three use the regression coefficients and means for each group. The latter 3 attempt to make group comparisons by the method of mean substitution (Jones, 1992b: 146), see section 5.13.
AIs with the human capital, or means, of ADs would do substantially worse than other AIs by about ($2.74hr). The hypothesis H13, that AIs with Australian education would have lower hourly income than AIs with Australian education and AD human capital, was not supported. UKDs with AD means, would do worse with their earnings decreasing about ($6.44hr). In contrast, ADs with AI human capital increased their hourly wage by ($5.66hr) and after about 11 years of Australian schooling started to do better than the AIs.
7.5.12 Years of Overseas Education
The AIs being the group that has migrated most recently has the highest mean years of overseas education, 9.7 years, followed by the UKDs 7.4 years. The number of years of education completed overseas was positively related to earnings for all three groups. The average hourly rate of return for AIs was ($0.53hr), for UKDs ($0.78hr) and ADs ($0.52hr).
Using the method of mean substitution (Jones, 1992b: 146) we can compare the hourly income of the three groups while varying years of Overseas education. After just one years' overseas education both AIs, ($10.37hr), and UKDs, ($9.28hr), were earning less than the ADs ($13.86hr). After 15 years the position of AIs, ($17.78hr), had not improved relative to the ADs, ($21.12hr), and had slipped compared to the UKDs, ($20.19hr). The hypothesis H14, that AIs with overseas education would have lower hourly income than UKDs and ADs with overseas education, was supported. This is another example of the devaluation of overseas education with the exception of that gained in the U.K and perhaps the U.S. and Canada, see Figure 7.6.
Replacing the means, a measure of human capital, for the AI and UKD regression equations with that of the ADs, resulted in a substantial increase in average hourly income. AIs with AD human capital increased their hourly income by ($6.31hr), UKDs with AD hourly income increased their predicted hourly income by ($4.44hr). ADs with AI human capital suffered a drop in hourly income of ($4.19hr). The hypothesis H15, that AIs with overseas education would have lower hourly income than AIs with overseas education and AD human capital, was supported.
7.5.13 Australian Labour Force Experience
Expectably, the ADs had the highest mean number of years of Australian Labour Force Experience, 17.6 years, followed by the UKDs, 14.7 years and the AIs, 11.6 years. The number of years of Australian Labour Force Experience, while being positively related to hourly earnings, explained relatively small amounts of the variance in the model. AIs with one years' labour force experience had an average return of ($0.18hr), UKDs ($0.08hr) and ADs ($0.10hr).
AIs with one year of Australian labour force experience had a predicted hourly rate substantially higher, ($13.14hr), than ADs ($11.67hr), but lower than the UKDs, ($13.24hr). After 15 years of Australian labour force experience AIs had predicted hourly incomes of ($15.58hr), this compared to ($14.30hr), for UKDs and ($13.09hr), for ADs. The hypothesis H16, that AIs with Australian work experience would have lower hourly incomes than UKDs and ADs with Australian work experience, was not supported.
Replacing AI human capital with AD human capital resulted in a marginal increase in hourly income ($0.14hr), while replacing UKD means with AD human capital resulted in a drop in income of about ($1.51hr). Replacing AD human capital with AI human capital resulted in an increase in hourly income of about $1.52hr. The hypothesis H17, that AIs with Australian work experience would have lower hourly income than AIs with Australian work experience and AD human capital, was supported.
7.5.14 Overseas Labour Force Experience
The AIs had significantly more overseas labour force experience, 7.3 years, compared to the UKDs, 5.2 years. Having worked overseas was positively related to hourly income. For AIs the average return for a year of overseas labour experience is ($0.12hr), for UKDs ($0.03hr) and for ADs ($0.07hr).
With just one years overseas labour force experience, AIs predicted hourly income ($14.27hr), is more than that for UKDs, ($14.17hr) and ADs, ($13.42hr). With increasing overseas labour force experience the AIs improved their position even further, after 15 years AIs earned ($15.87hr), compared to UKDs ($14.54hr) and ADs ($14.38hr). The hypothesis H18, that AIs with overseas work experience would have lower hourly income than UKDs and ADs with overseas work experience, was not supported.
Substituting AI human capital with AD human capital, resulted in an average hourly increase of about ($2.02hr), substituting UKD human capital with AD human capital resulted in a drop in income of ($1.16hr). When AD human capital was substituted with AI human capital there was an increase in predicted hourly earnings of ($0.40hr). The hypothesis H19, that AIs with overseas work experience would have lower hourly income than AIs with overseas work experience and AD human capital, was supported. There seems to be support for the view that Asian immigrants such as the AIs have their overseas work experience devalued by Australian employers.
Fig 7.5: The Effects of Australian Education on Predicted Hourly Income
Fig 7.6: The Effects of Overseas Education on Predicted Hourly Income
Fig 7.7: The Effects of Australian Labour Force Experience on Predicted Hourly Income
Fig 7.8: The Effects of Overseas Labour Force Experience on Predicted Hourly Income
7.6 Chapter Summary and Key findings
AIs with overseas labour force experience had lower predicted unemployment levels than UKDs but higher levels of predicted unemployment than ADs. With regard to overseas education, AIs had substantially higher levels of predicted unemployment levels compared to the UKDs and ADs. Similar findings were made regarding Australian labour force experience and Australian education.
For AIs with AD human capital, unemployment increased dramatically. Whether we examine the effects of overseas work experience, see Figure 7.1, and overseas education, see Figure 7.2, or Australian work experience, see Figure 7.3, and Australian education, see Figure 7.4, of AIs with AD human capital, the effects are the same. AIs with AD human capital have substantially higher levels of predicted unemployment.
The AIs like many other ethnic groups in Australia are well educated and quite affluent. Hopefully, as AIs become better established in Australia they will begin to occupy more managerial positions and their unemployment rates will come down to that of the Australian born. But, they already earn more than UKDs and ADs who are similarly qualified or are employed in a similar job. Further younger AIs, those who arrived in Australia at the age of 16 or younger and those who were born in Australia are doing better, earnings wise on average, than UKDs and ADs.
What the findings of the present study confirm is that overseas education and work experience are heavily discounted by employers. An AI with AD human capital and one year of overseas schooling would have earned ($6.31hr) more than a similarly qualified AI, see Figure 7.6. Similarly, those AIs with AD human capital and one year of overseas work experience would have earned ($2.02hr) more than an average AI, see Figure 7.8.
The discounting of overseas education and work experience would strongly suggest the presence of discrimination if a similar process operated with Australian education and work experience. It does not. An AI with AD human capital and one year of Australian schooling would have earned ($2.74hr) less than an average AI, see Figure 7.5. With regard to the effects of Australian work experience. An AI with AD human capital and one year of Australian work experience earns ($0.14hr) more than an average AI, see Figure 7.7.
Stromback et al (1992: 61) in their study of the Vietnamese, Maltese and Lebanese concluded that "...the evidence [for] discrimination [in Australia] is, while not unambiguous, still fairly strong". The AIs appear not to be experiencing any discrimination to the extent that their hourly income is above that of UKDs and ADs. In the case of unemployment and participation rates there appears to be some evidence that AI position has worsened over the last decade and it is quite possible that new immigrants are being discriminated against by some employers. But, on the whole, it appears difficult to suggest that consistent and structurally based discrimination is effecting the careers of AIs in Australia.
