Industry representation, structural change and the older male worker in Australia (1985 to 2005).

Abstract

Employment segregation and unfavourable changes to the industrial composition of employment are a possible reason for the long term decline in labour force participation of older males in Australia over recent decades. In this paper, I explore this proposition by analysing employment data for older males over the 1985 to 2005 period. Findings from segregation and shift-share analyses suggest that, although older males appear to be segregated by industry and generally over-represented in stagnant or declining industries, the net aggregate effects of structural change for older males' employment trends are minimal.

Introduction

Labour force participation rates for older males (aged 55-64 years) in Australia have fallen since the mid 1960s. (1) Explanations for these trends are divided between those emphasising labour supply and those emphasising labour demand forces. The former predict that workers have left the labour force early in response to financial incentives, such as greater accumulation of wealth over the career, private pensions and social security eligibility and value (for example, Merrilees 1982, Miller 1983). The latter explanation suggests that employment separation and a dearth of available employment will force workers to leave the labour force as a pool of hidden unemployed; that is, as discouraged workers (for example, Stricker and Sheehan 1981). This latter explanation is supported by more recent studies (for example, O'Brien 2001a and 2001b, Argyrous and Neal 2001)

What remains unresolved, however, is the exact nature of changes in labour market conditions that may have brought about this discouraged worker effect. For example, has it come about as a result of industrial restructuring, changes to hiring and firing practices within industries, or some change in aggregate labour demand conditions? In this paper, two major aspects of older male worker employment changes are explored, specifically, the role of industry representation and changing industrial composition. The first aspect explores whether older workers are segregated by industry. That is, are they disproportionately represented across industries? Second, has there been structural change away from their favoured industries that would be expected to impact on their overall employment levels?

The rest of this paper is structured as follows. First, a brief review of theoretical and empirical literature is presented, followed by a brief discussion of the data used. Next, the segregated nature of labour markets with respect to age and industry is explored. Then the role of changes to industry structure, and age composition within industries, from 1985 to 2005 for older male employment is examined using shift-share analyses. This extends the time period of previous analysis published in the Australian Bulletin of Labour (Moir 1982). The results suggest that a very different industry composition of employment for older males would have existed had past employment structures held. But there is only a small net aggregate effect on older male employment levels. In contrast, a larger role is found for changes to age composition within industries for older male employment. Finally, an extension to this methodology, incorporating the role of female employment trends, is explored.

A Review of Literature

Most empirical literature surrounding older workers' employment and labour force participation was published in the early to mid 1980s, in response to the unusual decline in labour force participation rates for older males in the 1970s and subsequent stabilisation in the 1980s. BLMR (1983) Hughes (1984) and Merrilees (1986) provided summaries of the main debates to be found in this literature. Stricker and Sheehan (1981), Merrilees (1983), Miller (1983) and BLMR (1983) developed quantitative estimates of various influences on older male labour force participation rates.

The male LFPR debate of the 1970s and 1980s specified three causal forces for explaining trends in older male labour force participation rates: first, increased private wealth and the choice of earlier substitution of leisure for work; second, increased availability and financial attractiveness of social security pensions; third, discouraged worker effects resulting from high levels of unemployment. Merrilees summarised the key conclusions that came out of this debate as follows;

   In summary, a consensus is at last emerging which
   emphasises the life cycle private wealth effects as the key
   to the long term trend to early retirement, though not
   necessarily the short run (1973-76) acceleration in such.
   (Merrilees 1986, p. 220)

In short, the conclusion from this early debate was that the primary cause of declining older male labour force participation rates was supply-side in orientation and was not a reflection of labour demand-side forces resulting in high levels of hidden unemployment.

More recent literature on older workers employment and labour force participation has been inspired by the ageing society phenomenon and associated policy to encourage increased labour force participation of older workers (for example, DFaCS 2002). In contrast to the earlier period of analysis, more recent results have emphasised the role of labour demand and labour force discouragement for older male worker employment and participation trends (for example Argyrous and Neale 2001, O'Brien 2001a and 2001b).

Most of the preceding analyses were conducted using micro or aggregate data. Relatively few have analysed or quantified the influence of industrial representation and composition for older older worker employment trends. An exception to this is the study by Moir (1982), published in the Australian Bulletin of Labour. Her findings suggested that older males were represented disproportionately across industries over the period 1966 to 1980, thus supporting a proposition that labour markets are segmented by age. She dismissed, however, the idea of industry structure substantially influencing the number of older men (aged 55+ years) in employment over this period. McCormack (1996), applying Moir's methodology, suggested a large influence for changes in age composition and industry structure for older males' employment between 1991 and 1995: both intra- and inter-industry influences combined to impact negatively on older male employment.

More recent Australian studies have discussed a role of structural changes in employment, with the segmented nature of the labour market, for employment outcomes of older workers in Australia over recent decades. For example, Pickersgill et al (1996) provide an Australian literature review of attitudes of employers to older workers, productivity and human resource practices. They criticise the aggregate nature of much research and instead emphasise the influence of different markets, organisations and employment strategies for older workers in Australia (as did Taylor and Walker (1994) in the UK context). Case studies are used to show the different experiences of older workers in different industries and occupations. There appears, however, to have been no advance in the methodology or quantitative analysis of the role of industry representation and structural changes for older males in Australia since Moir's work.

Finally, structural unemployment was analysed by Hogue and Inder (1991). Their findings emphasise the segmented nature of the labour market with respect to age, sex, industry, occupation and duration of unemployment. They argue, however, that a low unemployment rate for older workers implies low structural unemployment. Case studies of post-retrenchment behaviour elsewhere indicate that older workers are more likely to display a low attachment to the labour force post-retrenchment after controlling for other factors (For example, Gordon et al 1989 and 1991). This implies that the unemployment rate may be a misleading indicator of the role of structural change for the older worker employment position.

In summary, recent literature indicates that industry structural changes may have influenced older male employment trends in Australia, but little formal quantitative analysis exists since Moir (1982). Thus Moir's methods are used as a starting point, with an extension subsequently suggested.

Data and Definitions

Australian Bureau of Statistics (ABS) Labour Force Survey (LFS) employment data disaggregated by industry, age and sex, from 1985 to 2005 are utilised for the following analyses (ABS 2005). Unemployment rates have not been used to analyse structural changes (as in Hogue and Inder) because of the prevalence of discouraged worker effects and hidden unemployment, rendering this measure misleading, particularly for older males.

The 1985-2005 timeframe has been split into the two sub-periods of 1985-1995 and 1995-2005 for the shift-share analysis in an attempt to avoid masking broad fluctuations in employment trends over the twenty year period. These sub-periods are of particular interest in the context of an ageing society. With reference to Figure 1, the sub-period 1995-2005 saw a large growth in the older male population associated with the baby boomer generation and the ageing society phenomenon.

[FIGURE 1 OMITTED]

The definition of the older worker adopted by the International Labour Organization (ILO 1995, p. 39) at the 66th Session of the International Labour Conference in 1980 was '... all those who are liable to encounter difficulties in employment and occupation because of advancement in age.' This definition leaves the actual age group covering older workers open to interpretation. VandenHeuval (1999) has established that males over the age of 55 years are disadvantaged in the labour market in Australia, while O'Brien (1999 and 2000) suggests that this definition may even extend to those in the 45-54 year-old age group. It is evident elsewhere that there has not been a formal definition established or adopted by research literature in Australia or overseas for the age group(s) actually occupied by older workers. It is defined in various Australian sources as aged greater than 50 years (Encel 1996), 55 (Moir 1982), 60 (Miller 1983) or even 65 (Merrilees 1983). Likewise, older workers are defined in various overseas studies as aged over 45 years (OECD 1998), 50 (Costa 1996), 55 (OECD 1995), 60 (Johnson 1989) or 65 (Johnson 1994). Older workers are defined as the age groups 55-59 and 60-64 years for the purposes of this study, while the prime-aged group is defined as 25-44 years and is used as a comparison or control group.

Industrial Representation of Older Males (Segregation Analysis)

Age segregation implies that certain workers are disproportionately represented across industries. We start with the most rudimentary measure of the distribution of older males across industries, the percentage of employed males of each age group within each industry. These percentages will necessarily sum to one hundred percent over all industries for each age group. That is:

Employment Distribution = [E.sub.ikt] * 100/[E.sub.kt]

where

[E.sub.ikt] = the number of male workers in age group k and industry i at time t.

[E.sub.kt] = the number of male workers in age group k at time t.

Moir (1982) also reports an industry representation score as a measure of older males' relative distribution within an industry. This is a more accurate measure of segregation. That is,

Industry representation coefficient = [E.sub.ikt] * 100/[E.sub.it] / [E.sub.kt]/[E.sub.t]

where

[E.sub.it] = the number of male workers in industry i at time t.

[E.sub.t] = total male employment at time t.

An industry representation coefficient above 100 indicates that workers in an age group are over- represented in that industry; a coefficient below 100 signifies under-representation. Moir (1982) presents this representation measure for the 55+ year-old age group as a whole, with no disaggregation within older male age groups, no comparison to other age groups, and no summary measure of overall age segregation. A number of enhancements are therefore made to Moir's methodology in this paper. First, findings for both the 55-59 and 60-64 year-old age groups are presented, along with comparisons with the prime-age group, to indicate relative segregation across male age groups. Second, a summary measure of segregation for each age group is reported. A relatively uneven pattern of representation across industries, implying industrial segregation of an age group, will be indicated by a relatively high coefficient of variation of the industry representation coefficients.

Furthermore, we investigate some indices of employment segregation by gender. Watts (1992) reviews the use of two such indices for the measurement of occupational sex segregation: the Index of Dissimilarity (ID) (Cortese et al 1976) and the IP index (Karmel and Maclachlan 1988). The ID represents the share of an age group that must be removed (without replacement) to achieve zero segregation, where zero segregation implies that each industry contains the same proportion of the age group in question equal in turn to the that age group's total age-share of employment ([[bar.a].sub.k] = [E.sub.kt]/[E.sub.t]).

For industries over-represented by an age group k, [R.sub.ikt] older males must be removed to eliminate segregation and achieve the total age share [a.sub.k].

That is,

([M.sub.ikt] - [R.sub.ikt])/([M.sub.it] - [R.sub.ikt]) = [[bar.a].sub.k]

where [M.sub.ikt] = the number of male workers of age group k in an industry i that is over-represented with age group k at time t.

[R.sub.ikt] = the number of males in age group k in an industry i that is over-represented with age group k at time t that must be removed in order to achieve zero segregation [bar.a].sub.k].

[M.sub.it] = the total number of male workers in an industry i that is over-represented with age group k at time t.

The ID for an age group is then calculated by aggregating [R.sub.ikt] over all m industries over-represented by age group k and expressed as a ratio of that age group's total employment.

I[D.sub.k] = [m.summation over (i=1)] [R.sub.ikt] / [E.sub.kt]

In comparison, the IP index measures the fraction of total employment which must relocate, with replacement, to achieve zero segregation of [[bar.a].sub.k] [E.sub.it] in all industries. That is,

I[P.sub.k] = [I.summation over (i-1)] [absolute value of [E.sub.ikt] - [[bar.a].sub.k] [E.sub.it]]/ [E.sub.t]

Because the ID is expressed as a fraction of the age group's employment rather than total employment, it would appear to offer a better relative comparison of segregation than the IP when applied to a number of age groups of differing sizes. This is especially the case with comparisons of prime-aged males covering a twenty year age span, versus the five year age span of each of the two older male worker groups. This issue would not be expected to be as much a problem when applied to only two sexes whose total employment would be more similar in magnitude. Therefore, only the ID is reported in subsequent analysis.

The employment distribution and industry representation coefficients of prime-aged and older male workers across industries from 1985 to 2005 are presented in Table 1. In terms of employment distribution, most employed older males were found in Manufacturing, Agriculture, and Construction industries. In contrast, relatively few were found in Finance and Insurance, Accommodation, Cafes and Restaurants, and Services industries generally. A relatively high percentage of 60-64 year-old age group workers was found in Agriculture. Over time, the percentage of older males in Manufacturing has been falling, but rising in Property and Business Services. Similar to older males, prime-aged males were common in Manufacturing and Construction industries, however, a much lower percentage were working in Agriculture.

In terms of their relative distribution across industries, older males generally tended to be over-represented compared to other age groups in industries such as Agriculture, Transport and Storage, Government Administration and Defence, Education, and Health and Community Services but under-represented in Mining, Retail Trade, Accommodation, Cafes and Restaurants, Communication Services and Finance and Insurance. These findings are in broad agreement with those of Moir (1982) for the period 1966 to 1980, though her representation scores were for the one aggregated group aged 55 years or more. Therefore, on the surface, older males seem to be over-represented in low growth or declining industries such as Agriculture and Government Administration, and under-represented in higher growth industries such as Retail and other Services industries.

In terms of summary measures of segregation, a consistent finding in all periods from the coefficient of variation of representation coefficients is that older males are concentrated more disproportionately across industries as age increases. This measure of relative dispersion was generally around 10 per cent for prime-aged males, over 20 per cent for males aged 55-59 years and over 30 per cent for males aged 60-64 years. Similarly, older males displayed the greatest segregation with reference to the index of dissimilarity. The ID coefficients suggest that approximately 5 per cent of prime aged males, 7-8 per cent of males aged 55-59 years and over 10 per cent of males aged 60-64 needed to be removed (without replacement) from that respective age group to achieve zero segregation. In summary, older males appear to be disproportionately represented across industries with reference to all measures of segregation presented. Furthermore, they appear to be over-represented in stagnant and decline industries.

A methodology that takes into account an age group's industry representation and the effects of structural changes to industry employment, as well as changes to age composition within industries, is shift-share analysis. Thus, shift-share analysis is able to quantify the potential effects of both inter-as well as infra-industry changes in employment for a defined group of workers. Inter-industry changes are captured as employment shifts across industries. These would be expected to impact on older male employment in segmented labour markets, being disadvantageous if employment shifts away from industries to which older workers are over-represented. Intra-industry employment changes are captured as a changing share of older males within each industry. Intra-industry changes would adversely impact on older workers in the case of changes to employer practices via the substitution for older workers of other age groups within industries

Effects of Industry Structure and Age Composition Changes, 1985 to 2005 (shift-share analysis)

Again, using Moir's (1982) methodology, we can decompose the number of male workers in an age group employed in the year 1999 and express the results as:

[I.summation over (i=1)] [E.sub.ikt] = [I.summation over (i=1)] [a.sub.ikt] x [p.sub.it][E.sub.t]

where [E.sub.ikt] = the number of male workers in age group k and industry i in year t.

[a.sub.ikt] = the proportion of male workers in industry i aged k in year t = [E.sub.ikt/[E.sub.it].

[P.sub.it] = the proportion of employed men in industry i in year t = [E.sub.it/[E.sub.t].

[E.sub.t] = total male employment in year t.

The industry structure effect, expressed as the hypothetical level of a male age group's employment in time period t, assuming past (t-10 years) industry structure, can be calculated as:

[I.summation over (i=1)] [E.sup.*.sub.ikt] = [I.summation over (i=1)] [a.sub.ikt] x [p.sub.it][E.sub.t]

where [E.sup.*.sub.ikt] = the hypothetical number of male workers in age group k assuming the industrial structure from 10 years ago.

[P.sub.it-10] = the proportion of employed men in industry i from 10 years ago = [E.sub.it-10]/[E.sub.t-10].

The age composition effect, expressed as the hypothetical employment of a male age group in 1999, assuming past (1984) age composition, can be calculated as:

[I.summation over (i=1)] [E'.sub.ikt] = [I.summation over i=1] [a.sub.ikt-10] x [p.sub.it] x [E.sub.t]

where [E'.sub.ikt] = the hypothetical number of male workers in age group k if we assume the age composition of 10 years ago.

[a.sub.ikt-10] = the proportion of male workers in industry i aged k 10 years ago

= [E.sub.ikt-10/[E.sub.it-10]

These hypothetical employment numbers for each industry are reported, as well as being expressed as a percentage of actual employment in time ~period t. A percentage over 100 would imply that employment would have been greater if past industry structure or age composition had held. As with the segregation measures, no formal testing procedures exist such as tests of statistical significance. Therefore, analysis consists of three components: first, a comparison of the relative size of the net aggregate shift versus share effects; second, where possible, estimates are compared with those from Moir's (1982) analysis covering the period 1966 to 1980; (2) third, analyses of the disaggregated industry estimates of shift and share effects are conducted.

Similar forms of (non-age specific) shift-share analyses have been conducted in Australia for the analysis of casual employment (for example, Dawkins and Norris 1990, De Ruyter 1997) and part-time employment growth (for example, Sadler and Aungles 1990). These studies, however, instead of showing the shift and share effects in terms of a hypothetical current employment number, have estimated the shift and share effects as a decomposition of the actual change in employment over the time period of interest. As these other methods presented do not appear to be a major step away from those of Moir (1982), Moir's method of shift-share analysis has been chosen to offer a comparison with the findings from 1966 to 1980.

On the surface, shift-share analysis appears to offer a simple analysis of structural change and segmented labour market theory propositions; for example, that a displaced coal miner is not likely to pick up a supermarket check-out job. This aspect is captured by the effect of shifts in industry structure. Furthermore, the analysis also deals with the possible substitution for older males of other workers within industries (the share effect). This form of analysis, however, requires the imposition of a number of restrictive assumptions and fails to address other important aspects of segmented labour market theory. Therefore, it would be misleading to equate the shift-share analysis proposed with a test of segmented labour market theory in general.

First, the use of this methodology assumes that existing industry classifications adequately represent labour market segments. Industry is used as a convenient grouping of employees rather than reflecting an in-depth analysis of key features of employment segmentation such as job mobility, wages, conditions and benefits.

Second, the analysis of the influence of industry shifts assumes that workers of all ages are likely to be hired (fired) in growing (declining) industries according to prior employment age-share concentration only. This ignores an important aspect of segmented labour market theory that assumes different treatment of age groups in different segments. For example, even though older males may be over-represented in a certain industry, employment loss associated with a structural employment shift away from that industry may fall more disproportionately upon relatively younger workers if internal labour markets and seniority rules are prevalent in that industry.

Third, the analysis of the changing age composition of employment within industries ignores key aspects of demographic change. For example, changing age composition of employment within industries over time may reflect the changing proportions of various age groups within the general population, rather than explicitly showing changes to employer preferences and employment practices.

Fourth, the findings may be influenced by the choice of the start and finish dates relative to the phase of the business cycle.

Finally, the analysis largely ignores the role of human capital considerations. That is, that changing age composition of employment within industries, or the effects on older male employment of changing industry structure, may reflect the changing skill requirements of modern employment which older males may lack, rather than being the result of labour market segmentation. Nevertheless, shift-share analysis still offers an interesting, albeit simplistic, counter-factual analysis of potential explanations for older male employment patterns over time.

Tables 2 to 4 report the hypothetical employment levels in year t that would have eventuated had past (t-10 years) industry structure, or age composition within industries, remained unchanged. The hypothetical employment levels are also expressed as percentages of actual time t employment in brackets. 3 These are compared with Moir's (1982) findings from 1966 to 1980 for the older male age groups. To start with the industry structure effects, the results in Table 2 indicate that employment levels for prime aged males would have been 1 per cent higher in 1995 had the 1985 industry structure held, while employment would have been unchanged in 2005 after applying the 1995 structure. With reference to Table 3, employment levels for males aged 55-59 would have been 1 per cent higher in both 1995 and 2005 had either the 1985 or the 1995 industry structure applied.

Finally, Table 4 shows that employment levels for males aged 60-64 would have been unchanged in 1995 had the 1985 industry structure remained, but would have been 3per cent higher in 2005 if the 1995 industry structure had persisted. Therefore, as in Moir's (1982) findings, the net aggregate effect of structural industry change for older males has generally been to decrease total employment for older males, but the aggregate net magnitude suggested is relatively small. Furthermore, the hypothetical effects from changing industry structure do not appear significantly larger for older workers compared to prime aged males.

An interesting picture, similar to Moir's (1982), emerges at the industry level, suggesting large underlying gross shifts of job creation and destruction in industry employment underlying the small net aggregate change. The analysis suggests that employment in Agriculture, Government Administration, Mining, and Manufacturing industries would have been much higher had the past industry structure remained. This effect was offset, however, by favourable industry shifts in employment, primarily toward Property and Business Services. The outsourcing of many business services from the former industries over recent times to Property and Business Services (for example, Hodge 1998) makes it unclear whether the structural change estimated represents true job creation and destruction or simply represents similar employment renamed under a different industry classification. For example, a cleaner employed by a manufacturing firm will be classified within Manufacturing industry employment. If the cleaning contract is outsourced to a firm specialising in cleaning services, the individual will now be reclassified under Property and Business Services, even though the job description and place of employment may not have altered.

In contrast to the industry structure effects, much larger changes were suggested for the age composition effects. Results in Table 2 suggest that 1995 prime-aged employment would have been 1 per cent higher had the 1985 age composition within each industry continued. Employment would have been 10 per cent higher in 2005, however, had the 1995 age composition held. Employment for males aged 55-59 years would have been 11 per cent higher in 1995 had the age composition from 1985 held. But employment would have been 30 per cent lower in 2005 with the 1995 composition (Table 3). Employment for males aged 60-64 years would have been 3 per cent and 28 per cent lower in 1995 and 2005 for the 1985 and 1995 age compositions respectively (Table 4). Especially for the 1995-2005 period, however, these apparently dramatic results appear to reflect changing demographics and an increase in the number and proportion of older people amongst the population associated with the ageing society, rather than changes to employer hiring and firing practices and the replacement of older by younger workers.

Extension to Moir's Method--Incorporating the Female Labour Market

One possible reason for the small aggregate net estimates for industry structure when we use Moir's (1982) method of shift-share analysis is the exclusion of the female labour market employment in calculations. This may be a significant omission in the light of Gregory's (1990) suggestion that employment had grown in industries dominated by female employment and diminished in areas previously dominated by male employment. Thus, if one were to include industry and age composition effects as a proportion of total (male and female) employment, rather than just male employment, a richer insight into possible employment shifts and substitutions may be available.

The total number of employed males in an age group in year t may now be expressed as:

[I.summation over (i=1)] [E.sub.mikt] = [I.summation over (i=1)] [a.sub.mikt] x [p.sub.mfit][E.sub.mft]

where [E.sub.mikt] = the number of male workers in age group k and industry i in year t.

[a.sub.mikt] = the proportion of male workers in (male + female) industry i aged k in year t

= [E.sub.mikt]/[E.sub.mft].

[p.sub.mfit] = the proportion of employed persons in industry i in year t = [E.sub.mfit]/[E.sub.mft]

[E.sub.mft] = total persons employment in year t.

The hypothetical total number of employed males in an age group in year t with past (t-10) industry structure can be calculated as:

[I.summation over (i=1)] [E.sup.*.sub.mikt] = [I.summation over (i=1)] [a.sub.mikt] x [p.sub.mfit-10] x [E.sub.t]

where [E.sup.*.sub.mikt] = the hypothetical number of male workers in age group k in year t had the industrial structure from 10 years ago.

[p.sub.it-10] = the proportion of persons employed in industry 10 years ago

[E.sub.mfit-10]/ [E.sub.mft-10].

The hypothetical total number of employed males in an age group in year t with past (t-10) age composition can be calculated as:

[I.summation over (i=1)] [E'.sub.mikt] = [a.sub.mikt-10] x [p.sub.mfit] x [E.sub.mft]

where [E'.sub.mikt] = the hypothetical number of male workers in age group k for the age composition of 10 years ago.

[a.sub.mikt] = the proportion of male workers in (male + female) industry i aged k 10 years ago

= [a.sub.mikt-10]/[E.sub.mfit].

Indeed, estimates presented in Tables 5 to 7 confirm the suspicion that female employment growth had potentially displaced older male employment through structural changes to industry employment. Employment for all age groups could have hypothetically been 3-4 per cent higher had past industry structure held. For example, results in Table 6 suggest that employment for males aged 55-59 would have been 4 per cent higher in 1995 and 3 per cent higher in 2005 if the 1985 and 1995 industry structures, respectively, had persisted. Comparing the results with those from the previous section implies that females may have displaced up to 3 per cent of males over the periods analysed. For example, employment was estimated in Table 3to be only 1 per cent higher for males aged 55-59 years on the assumption of past (male) industry structure. This finding is consistent across the three age groups presented rather than being specific to older males only.

Summary of Findings

The analyses showed that older males were more segregated by industry than prime-aged males. It was also found that they were generally overrepresented in stagnant or decline employment industries. Thus, it appeared as if older workers were particularly vulnerable to the effects of structural changes to industry employment The findings from the shift-share analyses, however, suggested that there were minimal effects for older male employment levels from changes to industry structure

The introduction of female employment into the shift-share calculations hinted that larger industry structure and age composition effects may have been significant for explaining declining older male employment. Again, however, the magnitude of this effect was quite small. The findings suggested that females may have displaced up to 3 per cent of males from employment during the periods of analysis. Furthermore, this was a common finding across the age groups presented rather than an effect specific to older males only.

In comparison to the industry structure effects, there were much larger changes to employment levels suggested by changes to age composition (the share effect). This, however, would appear largely to reflect demographic change over time. Over the 1995-2005 period, there were large increases in the proportions of older workers within most industries, reflecting a large increase in the older male population over this period. This is a particularly encouraging sign for older male employment in the context of an ageing society.

References

ABS, Labour Force, Australia, Detailed - Electronic Delivery, Quarterly, May 2005, cat. 6291.0.55.001.

Argyrous, G. and Neal, M. (2001), 'Labor Market Disability: Implications for the Unemployment Rate, Economic and Labour Relations Review, vol. 12, pp. 263-284.

BLMR (1983), Retired, Unemployed or at Risk: Changes in the Australian Labour Market for Older Workers, Bureau of Labour Market Research Report No. 4, Australian Government Publishing Service, Canberra.

Cortese, C.F., Falk, R.F. and Cohen, J.K. (1976), 'Further Considerations on the Methodological Analysis of Segregation Indices', American Sociological Review, vol. 41, pp. 630-737.

Costa, D.L. (1996), 'Health and Labor Force Participation of Older Men, 1900-1991', The Journal of Economic History, vol. 56, pp. 62-89.

Dawkins, P. and Norris, K. (1990), 'Casual Employment in Australia', Australian Bulletin of Labour, vol. 16, pp. 156-173.

De Ruyter, A. (1997), 'Workforce Casualisation in Australia: A Shift-share Analysis', Australian Journal of Labour Economics, vol. 1, pp. 143-159.

DFaCS (2002), What Australians Working Together Means for Mature People, Department of Family and Community Services, Canberra, available at http://www.together.gov.au/Publications/default.asp (last accessed 2002)

Encel, S. (1996), 'Involuntary Early Retirement and Labour Force Re-entry for Older Workers', in DSS (ed), Early Retirement Seminar, Australian Government Publishing Service, Canberra, pp. 113-120.

Gordon, M., Gordon, B. and Clarke, P. (1989), 'Education, Age and Redundancy: Labour Market Experience After BHP, 1982-84' Industry Economics Discussion Paper No. 39, Department of Economics, University of Newcastle.

Gordon, M., Gordon, B. and Clarke, P. (1991), 'Labour Market Status After Redundancy: An Econometric Approach', mimeo, University of Newcastle.

Gregory, R. G. (1990), 'Jobs and Gender: A Lego Approach to the Australian Labour Market, Centre for Economic Policy Research Discussion Paper No. 244, The Australian National University, Canberra.

Hodge, G. (1998), 'Contracting Public Sector Services: A Meta-analytic Perspective of the International Evidence', Australian Journal of Public Administration, vol. 57, pp. 98-111.

Hoque, A., Inder, B. A. (1991), 'Structural Unemployment in Australia', Applied Economics vol. 23, pp. 723-730

Hughes, B. (1984), 'Labour Force Participation: What are the Issues?', in Kaspura, A. (ed.), Labour Force Participation in Australia: The Proceedings of a Conference, BLMR Monograph Series No 1, Australian Government Publishing Service, Canberra, pp. 1-11.

ILO (1995), World Labour Report 1995, ILO, Geneva.

Karmel, T. and MacLachlan, M. (1988), 'Occupational Sex Segregation--Increasing or Decreasing?', Economic Record, vol. 64, pp. 187-195.

Johnson, P. (1989), 'The Labour Force Participation of Older Men in Britain, 1951-81', Work, Employment and Society, vol. 3, pp. 351-368.

Johnson, P. (1994), 'The Employment and Retirement of Older Men in England and Wales, 1881-1981, Economy History Review, vol. 47, pp. 106-128.

McCormack, J. (1996), 'Labour Market Exit and Re-entry of Male Workers 55+', in DSS (ed), Early Retirement Seminar, AGPS Canberra, pp. 121-160.

Merrilees, W. (1982), 'The Mass Exodus of Older Males From The Labour Force: An Exploratory Analysis', Australian Bulletin of Labour, vol. 8, pp. 81-94.

Merrilees, W. (1986), 'Economic Determinants of Retirement', in Mendelsohn, R. (ed.), Finance of Old Age, Centre for Research on Federal Financial Relations, Australian National University, Canberra.

Miller, P. (1983), 'On Explanations of Declining Labour Force Participation Among Older Males', Centre for Economic Policy Research Discussion Paper No. 74, Australian National University, Canberra.

Moir, H. (1982), 'Age Structure of Industries and the Position of Older Men in the Labour Market', Australian Bulletin of Labour, vol. 8, pp. 156-75.

O'Brien, M.J. (1999), 'Old Enough for the Scrapheap? Employment Issues for the Baby Boomers', in Lee, J., Probert, B. and Watts, R. (eds), Work in the New Economy: Policies, Programs, Populations, Centre for Applied Social Research, RMIT University, pp.207-223.

O'Brien, M.J. (2000), 'Labour Force Participation of Older Males and Females in Australia, paper presented to the 7th Interdisciplinary Gender Studies Conference, University of Newcastle, June 30.

O'Brien, M.J. (2001a), 'Older Male Labour Force Participation: the Role of Social Security and Hidden Unemployment', Australian Journal of Labour Economics, vol. 4, pp. 206-223.

O'Brien, M.J. (2001b), 'Older Males and Social Security Reliance', in Mitchell, W.F. and Carlson, E. (eds), The Tip of the Iceburg, CAER, Sydney, pp. 135-156.

OECD (1995), The Transition From Work to Retirement, OECD Social Policy Studies No. 16, OECD, Paris.

OECD (1998), 'Workforce Ageing in OECD Countries', Employment Outlook, June, pp. 123-151.

Pickersgill, R., Briggs, C., Kitay, J., O'Keeffe, S., and Gillezeau, A., (1996), Productivity of Mature and Older Workers: Employer's Attitudes and Experiences, ACIRRT Monograph No. 13, University of Sydney, Australia.

Sadler, C. and Aungles, P. (1990), 'Part-time Employment Growth in Australia, 1978-1989', Australian Bulletin of Labour, vol. 16, pp. 286-296.

Stricker, P. and P. Sheehan (1981), Hidden Unemployment: The Australian Experience, Institute of Applied Economic and Social Research, University of Melbourne, Melbourne.

Taylor, P.E. and A. Walker (1994), The Ageing Workforce: Employers: Attitudes Towards Older People' Work, Employment and Society, vol. 18, pp. 569-591.

VandenHeuvel, A. (1999), 'Mature Age Workers: Are They a Disadvantaged Group?', Australian Bulletin of Labour, vol. 25, pp. 11-22.

Watts, M. (1992), 'How Should Occupational Sex Segregation be Measured?', Work, Employment and Society, vol. 6, pp. 475-487.

Endnotes

(1) The decrease in labour force participation rates from 1966 to 1999 is equal to six6, eighteen 18 and thirty one31 percentage points for males prime-aged, 55-59, and 60-64 years, respectively (O'Brien 2001).

(2) Different industry classifications are evident between the two periods of analysis, with Moir (1982) using the previous ASIC classifications. However, ; but broad comparisons with ANZSIC estimates are still available.

(3) A percentage of 10x% would imply that employment for older workers would have been x% per cent higher had past industry structure or age composition held.

Martin O'Brien, Economics and Information Systems, University of Wollongong 

Table 1: Employment Distribution and Industry
Representation Coefficients (brackets)--1985,
1995 and 2005

                                 Prime aged

INDUSTRY                1985       1995       2005

                        6.1        4.5        3.6
Agriculture             (88)       (79)       (79)

                        2.5        2.0        2.0
Mining                 (118)      (118)      (108)

                        18.4       18.1       15.4
Manufacturing           (90)      (104)      (108)

Electricity,            3.2        1.8        1.1
Gas & Water            (108)      (116)       (99)

                        10.9       11.7       14.2
Construction           (105)      (105)      (105)

                        7.6        7.9        5.8
Wholesale Trade        (106)      (104)      (108)

                        8.9        9.9        10.8
Retail Trade            (82)       (79)       (81)

Accommodation,
Cafes &                 2.1        3.3        3.7
Restaurants             (88)       (87)       (90)

Transport and           7.8        7.0        6.4
Storage                (106)      (108)      (101)

Communication           3.3        2.9        2.7
Services               (120)      (127)      (114)

Finance &               3.7        3.3        3.8
Insurance              (108)      (114)      (125)

Property &              6.7        9.3        12.4
Business Services      (107)      (101)      (107)

                        5.6        5.1        4.0
Govt Admin             (112)      (111)       (94)

                        5.2        4.0        3.7
Education              (121)       (95)       (91)

Health &
Community               3.6        4.0        4.0
Services               (102)      (105)      (103)

Cultural &
Recreational            1.7        1.9        2.6
Services               (106)       (89)      (100)

                        2.7        3.3        3.8
Personal Services      (106)      (108)      (106)

                       100.0      100.0      100.0
Total                  (100)      (100)      (100)

Coefficient of
Variation (%)           10.8       13.0       11.4

Index Of
Dissimilarity (ID)      .051       .048       .046

                                55-59 years

INDUSTRY                1985       1995       2005

                        9.0        8.5        5.3
Agriculture            (128)      (148)      (l15)

                        1.4        1.3        2.4
Mining                  (67)       (79)      (126)

                        20.9       17.0       13.6
Manufacturing          (102)       (98)       (95)

Electricity,            3.7        1.4        1.6
Gas & Water            (123)       (94)      (142)

                        8.9        10.3       11.6
Construction            (86)       (93)       (87)

                        6.6        8.3        5.5
Wholesale Trade         (93)      (109)      (102)

                        8.3        10.2       10.4
Retail Trade            (76)       (81)       (78)

Accommodation,
Cafes &                 2.2        2.9        2.9
Restaurants             (92)       (76)       (70)

Transport and           8.9        7.2        7.5
Storage                (121)      (110)      (119)

Communication           2.8        1.2        3.4
Services               (102)       (54)      (142)

Finance &               2.1        2.1        2.4
Insurance               (63)       (73)       (79)

Property &              5.7        9.8        11.1
Business Services       (91)      (107)       (95)

                        7.1        4.3        6.2
Govt Admin             (142)       (94)      (148)

                        4.1        6.6        5.4
Education               (95)      (158)      (133)

Health &
Community               4.8        4.0        4.4
Services               (136)      (104)      (112)

Cultural &
Recreational            1.4        1.9        2.1
Services                (86)       (92)       (83)

                        2.3        3.0        4.1
Personal Services       (90)       (96)      (112)

                       100.0      100.0      100.0
Total                  (100)      (100)      (100)

Coefficient of
Variation (%)           23.4       26.0       22.7

Index Of
Dissimilarity (ID)      .081       .073       .082

                                60-64 years

INDUSTRY                1985       1995       2005

                        12.6       14.0       8.3
Agriculture            (180)      (244)      (181)

                        1.0        0.3        1.5
Mining                  (50)       (19)       (77)

                        21.4       16.3       14.9
Manufacturing          (105)       (94)      (105)

Electricity,            4.2        0.5        1.3
Gas & Water            (142)       (34)      (109)

                        7.7        8.6        11.1
Construction            (73)       (78)       (83)

                        6.2        7.4        6.7
Wholesale Trade         (86)       (98)      (125)

                        8.7        11.1       9.3
Retail Trade            (80)       (89)       (70)

Accommodation,
Cafes &                 2.3        3.1        2.9
Restaurants             (97)       (82)       (69)

Transport and           8.3        5.8        7.4
Storage                (113)       (89)      (117)

Communication           3.2        0.4        2.1
Services               (115)       (17)       (90)

Finance &               1.5        1.4        2.1
Insurance               (43)       (47)       (71)

Property &              6.8        8.5        12.3
Business Services      (109)       (93)      (106)

                        5.8        5.4        4.6
Govt Admin             (116)      (119)      (109)

                        2.9        5.2        5.5
Education               (68)      (125)      (136)

Health &
Community               4.2        5.8        4.4
Services               (119)      (151)      (113)

Cultural &
Recreational            1.7        2.5        1.9
Services               (104)      (117)       (72)

                        1.4        3.6        3.5
Personal Services       (56)      (117)       (97)

                       100.0      100.0      100.0
Total                  (100)      (100)      (100)

Coefficient of
Variation (%)           35.5       56.7       28.6

Index Of
Dissimilarity (ID)      .114       .130       .100

Source: ABS (2005), all years use May figures, Author's calculations

Table 2: Industry Structure and Age Composition
Effects: Prime Aged Males 1985-1995 and 1995-2005

                                 Industry                Age
                            Structure Effects     Composition Effects

INDUSTRY                     1985-95    1995-05    1985-95    1995-05

Agriculture                    130        116        120        101
                              (122)      (125)      (112)      (109)

Mining                          59         46         47         62
                              (125)       (89)      (101)      (121)

Manufacturing                  506        475        374        416
                              (118)      (122)       (87)      (107)

Electricity, Gas & Water        82         39         39         37
                              (196)      (133)       (94)      (128)

Construction                   260        296        276        396
                               (94)       (82)      (100)      (110)

Wholesale Trade                176        210        193        157
                               (94)      (141)      (103)      (106)

Retail Trade                   203        258        248        295
                               (86)       (94)      (105)      (108)

Accommodation, Cafes &          49         86         80        100
Restaurants                    (63)       (91)      (103)      (106)

Transport & Storage            188        167        164        190
                              (113)      (102)       (99)      (117)

Communication Services          83         67         67         84
                              (119)       (98)       (95)      (122)

Finance & Insurance             93         92         74         96
                              (119)       (95)       (95)      (100)

Property & Business            150        249        236        327
Services                       (68)       (79)      (107)      (104)

Govt Admin                     131        110        122        131
                              (109)      (108)      (102)      (129)

Education                       96         98        122        107
                              (101)      (104)      (128)      (115)

Health & Community              88        100         92        115
Services                       (93)       (97)       (97)      (112)

Cultural & Recreational         35         53         54         64
Services                       (78)       (82)      (120)       (98)

Personal Services               65         83         78        110
                               (81)       (85)       (99)      (112)

                              2393       2543       2388       2789
All industries                (101)      (100)      (101)      (110)

Source: ABS (2005), all years use May figures, Author's calculations

Table 3: Industry Structure and Age Composition Effects:
Males Aged 55-59 Years 1966-1980, 1985-1995 and 1995-2005

                                  Industry Structure Effects

INDUSTRY                         1966-80    1985-95    1995-05

Agriculture                         34         28         30
                                  (138)      (122)      (125)

Mining                               5          4         10
                                   (85)      (125)       (89)

Manufacturing                       94         55         75
                                  (117)      (118)      (122)

Electricity, Cas & Water                        8         10
                                             (196)      (133)

Construction                        29         26         43
                                  (106)       (94)       (82)

Wholesale Trade                                21         35
                                              (94)      (141)

Retail Trade                        41         24         45
                                  (100)       (86)       (94)

Accommodation, Cafes &                          5         12
Restaurants                                   (63)       (91)

Transport & Storage                 24         22         35
                                   (99)      (113)      (102)

Communication Services                          4         15
                                             (119)       (98)

Finance & Insurance                             7         10
                                             (119)       (95)

Property & Business Services        15         18         40
                                   (71)       (68)       (79)

Govt Admin                                     13         31
                                             (109)      (108)

Education                                      18         25
                                             (101)      (104)

Health & Community                  16         10         19
Services                           (64)       (93)       (97)

Cultural & Recreational                         4          8
Services                                      (78)       (82)

Personal Services                    9          7         16
                                   (85)       (81)       (85)

Other industries                    32
                                   (87)

All industries                     299        274        460
                                  (101)      (101)      (101)

                                    Age Composition Effects

INDUSTRY                         1966-80    1985-95    1995-05

Agriculture                         24         23         22
                                   (98)       (98)       (91)

Mining                               8          3          5
                                  (144)       (96)       (44)

Manufacturing                       63         55         45
                                   (78)      (118)       (72)

Electricity, Cas & Water                        6          3
                                             (148)       (47)

Construction                        31         29         40
                                  (113)      (104)       (76)

Wholesale Trade                                22         19
                                              (96)       (75)

Retail Trade                        47         29         35
                                  (113)      (106)       (73)

Accommodation, Cafes &                         11         10
Restaurants                                  (137)       (76)

Transport & Storage                 25         24         22
                                  (103)      (124)       (66)

Communication Services                          7          4
                                             (214)       (27)

Finance & Insurance                             6          7
                                              (97)       (65)

Property & Business Services        20         26         40
                                   (99)       (96)       (79)

Govt Admin                                     20         13
                                             (171)       (45)

Education                                      12         20
                                              (68)       (84)

Health & Community                  28         16         13
Services                          (112)      (148)       (65)

Cultural & Recreational                         6          8
Services                                     (105)       (79)

Personal Services                   15          9         11
                                  (138)      (105)       (60)

Other industries                    36
                                   (98)

All industries                     297        303        317
                                  (100)      (111)       (70)

Source: ABS (2005), (Moir 1982), all years
use May figures, Author's calculations

Table 4: Industry Structure and Age Composition Effects:
Males Aged 60-64 Years--1966 to 1980, 1985-95 and 1995-2005

                                  Industry Structure Effects

INDUSTRY                          1966-      1985-      1995-
                                    80         95          5

Agriculture                         20         26         26
                                  (138)      (122)      (125)

Mining                               0          1          3
                                   (85)      (125)       (89)

Manufacturing                       49         30         45
                                  (117)      (118)      (122)

Electricity, Gas & Water                        2          4
                                             (196)      (133)

Construction                        10         13         23
                                  (106)       (94)       (82)

Wholesale Trade                                11         24
                                              (94)      (141)

Retail Trade                        19         15         22
                                  (100)       (86)       (94)

Accommodation, Cafes &                          3          6
Restaurants                                   (63)       (91)

Transport & Storage                 10         10         19
                                   (99)      (113)      (102)

Communication Services                          1          5
                                             (119)       (98)

Finance & Insurance                             3          5
                                             (119)       (95)

Property & Business                  7          9         24
Services                           (71)       (68)       (79)

Govt Admin                                      9         12
                                             (109)      (108)

Education                                       8         14
                                             (101)      (104)

Health & Community                   9          8         11
Services                           (64)       (93)       (97)

Cultural & Recreational                         3          4
Services                                      (78)       (82)

Personal Services                    4          5          7
                                   (85)       (81)       (85)

Other industries                    11
                                   (87)

All industries                     139        155        254
                                  (102)      (100)      (103)

                                    Age Composition Effects

INDUSTRY                          1966-      1985-      1995-
                                    80         95          5

Agriculture                         19         16         20
                                   (98)       (73)       (99)

Mining                               0          1          1
                                    (0)      (255)       (18)

Manufacturing                       44         28         24
                                  (104)      (111)       (66)

Electricity, Gas & Water                        3          1
                                             (415)       (23)

Construction                        21         12         19
                                  (223)       (94)       (69)

Wholesale Trade                                10         10
                                              (88)       (57)

Retail Trade                        32         15         22
                                  (171)       (90)       (93)

Accommodation, Cafes &                          6          6
Restaurants                                  (117)       (87)

Transport & Storage                 17         11         10
                                  (174)      (126)       (56)

Communication Services                          4          1
                                             (685)       (14)

Finance & Insurance                             2          3
                                              (91)       (49)

Property & Business                 13         15         20
Services                          (140)      (116)       (65)

Govt Admin                                      8          9
                                              (97)       (80)

Education                                       4          9
                                              (54)       (67)

Health & Community                  18          7         11
Services                          (137)       (78)       (98)

Cultural & Recreational                         3          5
Services                                      (89)      (119)

Personal Services                   14          3          8
                                  (273)       (47)       (89)

Other industries                    22
                                  (170)

All industries                     203        150        178
                                  (149)       (97)       (72)

Source: ABS (2005), (Moir 1982), all years
use May figures, Author's calculations

Table 5: Industry Structure and Age Composition Effects
Prime Aged Males 1985-1995 and 1995-2005 (Using Male and
Female Industry Employment)

                            Industry Structure      Age Composition
                                 Effects                Effects

INDUSTRY                     1985-95    1995-05    1985-95    1995-05

Agriculture                    133        121        128        100
                              (124)      (131)      (119)      (107)

Mining                          61         48         50         61
                              (129)       (94)      (106)      (118)

Manufacturing                  542        488        379        417
                              (126)      (125)       (88)      (107)

Electricity, Gas & Water        83         37         42         40
                              (198)      (128)      (101)      (137)

Construction                   277        308        283        391
                              (100)       (86)      (102)      (109)

Wholesale Trade                186        219        199        155
                               (99)      (147)      (106)      (104)

Retail Trade                   220        260        249        301
                               (93)       (95)      (106)      (110)

Accommodation, Cafes &          55         87         76        102
Restaurants                    (71)       (92)       (98)      (108)

Transport & Storage            190        164        177        198
                              (115)      (101)      (107)      (122)

Communication Services          87         67         69         87
                              (125)       (97)       (99)       (27)

Finance & Insurance             83        102         89         89
                              (107)      (106)      (115)       (92)

Property & Business            149        261        256        322
Services                       (68)       (83)      (117)      (102)

Govt Admin                     130         99        133        150
                              (108)       (97)      (111)      (147)

Education                       91         95        140        113
                               (96)      (102)      (148)      (121)

Health and Community            83         93        107        127
Services                       (88)       (90)      (112)      (124)

Cultural & Recreational         35         55         58         64
Services                       (78)       (84)      (130)       (98)

Personal Services               72         90         76        105
                               (91)       (92)       (96)      (107)

All industries                2477       2593       2511       2822
                              (104)      (103)      (106)      (111)

Source: ABS (2005), all years use May figures, Author's calculations

Table 6: Industry Structure and Age Composition Effects
Males Aged 55-59 Years 1985-1995 and 1995-2005 (Using
Male and Female Industry Employment)

                               Industry Structure     Age Composition
                                     Effects              Effects

INDUSTRY                       1985-95    1995-05    1985-95    1995-05

Agriculture                       29         31         24         21
                                (124)      (131)      (104)       (89)

Mining                             5         10          4          5
                                (129)       (94)      (101)       (43)

Manufacturing                     58         77         55         45
                                (126)      (125)      (119)       (73)

Electricity, Gas & Water           8          9          6          4
                                (198)      (128)      (159)       (50)

Construction                      28         45         30         40
                                (100)       (86)      (106)       (75)

Wholesale Trade                   22         37         22         19
                                 (99)      (147)       (99)       (74)

Retail Trade                      26         45         30         35
                                 (93)       (95)      (107)       (75)

Accommodation, Cafes &             6         12         10         10
Restaurants                      (71)       (92)      (131)       (77)

Transport & Storage               22         35         26         23
                                (115)      (101)      (133)       (68)

Communication Services             4         15          8          4
                                (125)       (97)      (222)       (28)

Finance & Insurance                6         11          7          7
                                (107)      (106)      (117)       (60)

Property & Business               18         42         28         39
Services                         (68)       (83)      (105)       (78)

Govt Admin                        13         28         22         15
                                (108)       (97)      (186)       (51)

Education                         17         25         14         22
                                 (96)      (102)       (78)       (88)

Health & Community                 9         18         18         14
Services                         (88)       (90)      (171)       (72)

Cultural & Recreational            4          8          6          8
Services                         (78)       (84)      (114)       (78)

Personal Services                  7         17          8         11
                                 (91)       (92)      (102)       (58)

All industries                   283        466        318        321
                                (104)      (103)      (117)       (71)

Source: ABS (2005), all years use
May figures, Author's calculations

Table 7: Industry Structure and Age Composition Effects Males
Aged 60-64 Years 1985-1995 and 1995-2005 (Using Male and Female
Industry Employment)

                            Industry Structure     Age Composition
                                  Effects               Effects

INDUSTRY                     1985-95    1995-05    1985-95    1995-05

Agriculture                     27         27         17         20
                              (124)      (131)       (78)       (97)

Mining                           1          3          1          1
                              (129)       (94)      (268)       (18)

Manufacturing                   32         46         28         24
                              (126)      (125)      (112)       (66)

Electricity, Gas & Water         2          4          4          1
                              (198)      (128)      (446)       (24)

Construction                    13         24         13         19
                              (100)       (86)       (96)       (68)

Wholesale Trade                 11         25         10          9
                               (99)      (147)       (90)       (57)

Retail Trade                    16         22         16         22
                               (93)       (95)       (90)       (95)

Accommodation, Cafes &           3          7          5          6
Restaurants                    (71)       (92)      (112)       (89)

Transport & Storage             10         19         12         11
                              (110)      (101)     (136)        (58)

Communication Services           1          5          4          1
                              (125)       (97)     (71l)        (14)

Finance & Insurance              2          6          2          2
                              (107)      (106)      (110)       (45)

Property & Business              9         25         17         19
Services                       (68)       (83)      (127)       (64)

Govt Admin                       9         11          9         10
                              (108)       (97)      (106)       (92)

Education                        8         14          5         10
                               (96)      (102)       (62)       (71)

Health & Community               8         10          8         12
Services                       (88)       (90)       (90)      (109)

Cultural & Recreational          3          4          4          5
Services                       (78)       (84)       (96)      (119)

Personal Services                5          8          3          7
                               (91)       (92)       (46)       (85)

All industries                 160        258        158        181
                              (104)      (104)      (102)       (73)

Source: ABS (2005), all years use
May figures, Author's  calculations