Living Arrangements: Changes across Australian generations
By the time the group of people born in 1965-69 had reached 30-34 years, nearly 24% remained unmarried, almost twice as many as those born 15 years earlier, when 13% remained unmarried at the same age.
Changes in the economic and social circumstances of successive Australian generations have implications for society, and for its institutions and policies. Generational changes in incomes, living standards, family size and living arrangements can affect the economy, the communities where people live and the provision and funding of services ranging from schools and hospitals, through to pensions and other income support. Understanding generational changes and trends will assist in assessing their impact and in developing appropriate responses.
Most studies of the economic circumstances of Australians use cross-sectional data generated by surveys conducted at a point in time. This provides a basis for comparing how people vary in relation to their age, household type, income level, country of birth, and so on. What these comparisons do not fully reveal is how the circumstances of individuals are changing as they move through the life cycle.
This article uses a cohort analysis to examine selected living arrangements and income characteristics of successive generations of Australians. Changes in these characteristics are examined for a series of five-year age groups (or cohorts) over a 15 year period from 1984 to 1998-99, a period that has been characterised by rapid economic and social change.
The article illustrates the contribution of cohort comparisons to understanding changes in household structures, incomes and standards of living over this period. These comparisons provide information about what is happening to different age groups at a point in time and how the circumstances and fortunes of different age groups are changing over time.
| Household Expenditure Surveys This article uses data from the four most recent ABS Household Expenditure Surveys. These surveys were conducted in 1984, 1988-89, 1993-94 and 1998-99. While there have been changes in methodology from survey to survey, the data used in this article are broadly comparable across all four surveys. While the analysis in this article is based on people in particular age groups (or cohorts), two of the characteristics that are examined, household income and children within the household, relate to the households to which they belong. For example, each person within a particular household may be in a different cohort but they will each have the same household income as all of the others in that household. |
Constructing synthetic cohorts
ABS Household Expenditure Surveys were conducted approximately five years apart from 1984 to 1998-99, and each collected information on the age of household members that can be expressed in five-year ranges. This makes it possible to construct a synthetic cohort by gathering together information on people who were of a given age in 1984 and a comparable group who were 5 years older in 1988-89, 10 years older in 1993-94 and 15 years older in 1998-99. This information can then be viewed not as a series of separate cross-sectional snapshots, but from the perspective of a cohort that has aged by 15 years over the course of the four surveys.
This method has been used to construct 10 separate synthetic cohorts from the four surveys. The first or youngest cohort was aged 15-19 years in 1984 and ended up aged 30-34 years in 1998-99, the second cohort was aged 20-24 years in 1984 and ended up aged 35-39 years in 1998-99, and so on. It is then possible to examine how the circumstances of people in each cohort change over time and to compare different cohorts at a given age.
Currently it is only possible to track changes over 15 years using ABS Household Expenditure Survey data. As further surveys are conducted, it will become possible to extend the scope of this kind of analysis in ways that may make it possible to identify key trends with more certainty.
SYNTHETIC COHORTS |
Family and household characteristics
The basic structure of households change as people marry, possibly get divorced and remarry, have children, and see those children attain adulthood. At the same time social and economic conditions, and attitudes also change. Comparison across different cohorts allows the extent of these generational changes to be examined. This issue is explored by examining for people in each age group: the proportion who had never entered a registered marriage; the proportion who were in households that contained children (0-14 years); and the average number of children that were in their households, where there were some present.
PROPORTION OF PERSONS IN SELECTED BIRTH YEAR GROUPS WHO HAVE NEVER BEEN IN A REGISTERED MARRIAGE(a)
(a) Includes those people who are in, or have been in, a de facto relationship but have never been in a registered marriage.
Source: ABS, 1984, 1988-89, 1993-94 and 1998-99 Household Expenditure Surveys.
The proportion of people who had never been in a registered marriage within each cohort declined over time as more people met and were married. For example, 24% of those aged 25-29 years in 1984 had never been married; 5 years later in 1988-89 only 15% of this age group (who were then
aged 30-34 years) had never been married.
A change in attitudes to registered marriage among people from different cohorts is likely to have resulted in increasingly larger proportions of people in younger cohorts remaining unmarried (at least up until their early 40s). By the time the youngest cohort (those born 1965-69) had reached 30-34 years, nearly 24% were unmarried, almost 6 percentage points higher than that of the previous cohort at the same age, and almost double that of the cohort born 15 years earlier (13%). This is consistent with an increase in de facto relationships, which delay or replace registered marriage. (A person in, or who has been in, a de facto relationship is included as never married, unless the person had previously been in a registered marriage.)
| Cohort analysis Cross-sectional surveys, that are representative of a population of interest, provide a snapshot of what is happening at a particular point in time. While cross-sectional surveys are often repeated, they rarely include the same households and people and thus cannot be used to examine how the structure and fortunes of specific people and households have changed (i.e. they can not be used for longitudinal analysis). Comparing people across different ages using cross-sectional results to draw conclusions about changes over time can be misleading when there are generational effects operating. For example, it cannot be assumed that the circumstances and experience of future generations will be similar to those of current generations. Questions about changes that occur across generations can usually only be answered with the help of longitudinal data that track how the circumstances of the same people vary over time. However, in the absence of longitudinal data, it is possible to link data from a sequence of cross-sectional surveys to form what is referred to as ‘synthetic cohorts’. These can then be used to estimate how the circumstances of cohorts change as they move through a statistically constructed life cycle. In the cohort analysis conducted for this article people who migrated to Australia since the first survey in 1984 were excluded to ensure that there were consistent populations in the age cohorts. However, the populations in each age group for each of the four surveys does vary slightly because of differences in the timing of the first survey compared with later surveys and because the precise age recorded in each survey depends upon when a person’s birthday falls relative to when they are interviewed. The trends within individual cohorts can be affected slightly by sampling variation and the effect of mortality and outward migration rates differing in relation to the characteristic being examined. For example, such factors would explain small increases in the proportion of people never married in some cohorts. |
Within each cohort, the probability of a person living in a household containing a child (0-14 years) increased with the age of the householder and then declined from approximately age 40-44 years, as the children attained adulthood. However, there are differences in these life-cycle patterns across cohorts. In keeping with the trends towards some families having no children (see Australian Social Trends 2002, Trends in childlessness) and increasing divorce (see Australian Social Trends 2002, Family: national summary table) the proportion of people of a given age with children present in their household tended to fall below that of the previous cohort. Around 70% of the second youngest cohort were living with a child by the time they had reached 35-39 years of age in 1998-99. This compares with around 80% of the people in the cohorts born 10 and 15 years earlier when they were in their late 30s. Similarly, of the cohort aged 30-34 years in 1984, 30% lived with a child when they had reached the age 45-49 years, compared with over 40% for the cohort born 15 years ahead of them.
While people in each cohort were less likely than those in previous cohorts at the same age to be living in a household with a child, for those that were living with a child, the number present remained relatively stable over successive cohorts. There was a consistent age pattern, with the average number of children present tending to peak at around 30-39 years of age and fall thereafter. However, changing patterns in the number of children present from survey to survey are likely to be related to the timing of births, which can be influenced by changing attitudes to family formation and prevailing economic circumstances.
SELECTED FAMILY AND HOUSEHOLD CHARACTERISTICS OF PERSONS
 | Age (years) in 1984, Year of birth | |||||||||
| ||||||||||
| Age at time of survey (years) | 15-19 1965-69 | 20-24 1960-64 | 25-29 1955-59 | 30-34 1950-54 | 35-39 1945-49 | 40-44 1940-44 | 45-49 1935-39 | 50-54 1930-34 | 55-59 1925-29 | 60-64 1920-24 |
| Persons never married(a) | |||||||||
| % | % | % | % | % | % | % | % | % | % |
| 15-19 | 97.2 | - | - | - | - | - | - | - | - | - |
| 20-24 | 70.8 | 60.6 | - | - | - | - | - | - | - | - |
| 25-29 | 37.3 | 30.4 | 24.3 | - | - | - | - | - | - | - |
| 30-34 | 23.7 | 18.0 | 15.3 | 12.5 | - | - | - | - | - | - |
| 35-39 | - | 14.1 | 11.9 | 8.2 | 7.3 | - | - | - | - | - |
| 40-44 | - | - | 10.4 | 8.3 | 4.8 | 5.3 | - | - | - | - |
| 45-49 | - | - | - | 6.7 | 6.0 | 5.1 | 4.9 | - | - | - |
| 50-54 | - | - | - | - | 4.1 | 4.1 | 4.3 | 5.8 | - | - |
| 55-59 | - | - | - | - | - | 5.1 | 3.0 | 5.0 | 4.0 | - |
| 60-64 | - | - | - | - | - | - | 3.7 | 5.6 | 4.9 | 5.2 |
| 65-69 | - | - | - | - | - | - | - | 2.8 | 4.2 | 5.3 |
| 70-74 | - | - | - | - | - | - | - | - | 4.3 | 4.3 |
| 75 and over | - | - | - | - | - | - | - | - | - | 4.6 |
| Persons living in households where children aged 0-14 years are also present | |||||||||
| % | % | % | % | % | % | % | % | % | % |
| 15-19 | 52.5 | - | - | - | - | - | - | - | - | - |
| 20-24 | 22.1 | 30.4 | - | - | - | - | - | - | - | - |
| 25-29 | 35.1 | 46.9 | 54.4 | - | - | - | - | - | - | - |
| 30-34 | 59.5 | 62.4 | 72.1 | 75.6 | - | - | - | - | - | - |
| 35-39 | - | 69.7 | 68.7 | 79.3 | 80.1 | - | - | - | - | - |
| 40-44 | - | - | 57.0 | 52.2 | 61.5 | 69.2 | - | - | - | - |
| 45-49 | - | - | - | 30.0 | 24.8 | 35.3 | 41.7 | - | - | - |
| 50-54 | - | - | - | - | 11.3 | 9.7 | 17.5 | 17.8 | - | - |
| 55-59 | - | - | - | - | - | 6.7 | 5.9 | 4.7 | 8.5 | - |
| 60-64 | - | - | - | - | - | - | 4.6 | 2.1 | 3.2 | 4.6 |
| 65-69 | - | - | - | - | - | - | - | 1.7 | 1.9 | 2.7 |
| 70-74 | - | - | - | - | - | - | - | - | 2.5 | 1.3 |
| 75 and over | - | - | - | - | - | - | - | - | - | 1.8 |
| Average number of children aged 0-14 years or less in the household (where such children are present) | |||||||||
| no. | no. | no. | no. | no. | no. | no. | no. | no. | no. |
| 15-19 | 1.49 | - | - | - | - | - | - | - | - | - |
| 20-24 | 1.40 | 1.44 | - | - | - | - | - | - | - | - |
| 25-29 | 1.91 | 1.77 | 1.86 | - | - | - | - | - | - | - |
| 30-34 | 2.07 | 2.39 | 2.17 | 2.17 | - | - | - | - | - | - |
| 35-39 | - | 2.07 | 2.25 | 2.15 | 2.18 | - | - | - | - | - |
| 40-44 | - | - | 1.75 | 1.87 | 1.77 | 1.77 | - | - | - | - |
| 45-49 | - | - | - | 1.51 | 1.61 | 1.54 | 1.49 | - | - | - |
| 50-54 | - | - | - | - | 1.44 | 1.30 | 1.36 | 1.41 | - | - |
| 55-59 | - | - | - | - | - | 1.64 | 1.83 | 1.28 | 1.44 | - |
| 60-64 | - | - | - | - | - | - | 1.42 | 1.76 | 1.61 | 1.85 |
| 65-69 | - | - | - | - | - | - | - | 1.14 | 1.36 | 1.45 |
| 70-74 | - | - | - | - | - | - | - | - | 1.71 | 1.42 |
| 75 and over | - | - | - | - | - | - | - | - | - | 1.98 |
Source: ABS 1984, 1988-89, 1993-94 and 1998-99 Household Expenditure Surveys.
Income and living standards
The income and living standards of households also change as people move through life-cycle stages. At various times people within a household may enter paid employment, leave the labour force for different lengths of time when children are born, eventually see these children enter the labour force and earn money of their own (which increases the household income), and see these children leave to form new households.
In this section three measures are used to compare the incomes and living standards of different age cohorts. The first measure is average weekly household disposable income (adjusted to 2000-01 dollars in line with movements in the Consumer Price Index). The second measure further adjusts these income data using an equivalence scale to account for the needs of households of different size and composition, resulting in a measure referred to as household equivalised income. The third measure examines the percentage of persons whose average weekly household equivalised income places them in the bottom 20 per cent of all persons, when ranked by household equivalised incomes.
Over a person’s life cycle, there is a well established tendency for average weekly household disposable income to increase up until middle-age (i.e. age 45-54 year olds), before declining slightly and then more rapidly after retirement (see Australian Social Trends 1998, Income distribution and life cycle). In broad terms, the cohort analysis confirms this tendency, with the exception of the two youngest cohorts (those born between 1960-69). The average weekly household disposable income for these two cohorts was at least $170 a week higher, at ages 15-19 years and 20-24 years, than when they were in the 25-29 years age group. This is likely to be associated with many young people still living with their parents, where many of these young people study up until age 20-24 years (see Australian Social Trends 2000, Young adults in the parental home). It remains to be seen if the life-cycle tendency of declining household disposable income experienced by people aged 45 years and over in 1984 will prevail among those people in the 6 youngest cohorts, most of whom were still younger than 45 years at the time of the last survey.
| Household income Household disposable income is defined as the combined income received from all sources by all members of the household, after personal income tax and other periodic levies (such as the Medicare Levy) are taken out. Household equivalised income is household disposable income adjusted on the basis of the household’s size and composition. This allows the standard of living of different households to be compared. For example, an adjustment is made to account for the difference that would exist in the standard of living between a couple with children and a couple without children who both receive the same combined household disposable income. The ‘modified’ OECD equivalence scale has been used in this article to estimate household equivalised income (for further information see Income Distribution, Australia 2000-2001, ABS cat. no. 6523.0). When comparing household equivalised income over time or between cohorts the relative magnitudes of the figures is most relevant, rather than the absolute levels. The household income data from the 1998-99 survey used in this article incorporate revisions to initially published results. For more information see Australian Economic Indicators April 2002, Upgrading Household Income Distribution Statistics (ABS cat. no. 1350.0). |
Comparisons across cohorts indicate each cohort had a higher average weekly household disposable income than its predecessor at some ages, but a lower income at other ages. This will, at least partly, reflect differences over time in general economic conditions, differences in household structures, and changes in the likelihood of household members being in paid work.
Most of the patterns observed in average weekly household disposable income remain in household equivalised income, where adjustments have been made for differences in household size and composition using an equivalence scale. However, the differences in average weekly household equivalised income between cohorts and within cohorts were smaller than those in (unadjusted) household disposable income. As household size is one of the main factors used to adjust household disposable income to obtain household equivalised income, the reduced difference in household equivalised incomes indicates that household disposable income tends to increase with household size. This implies that differences in living standards are less pronounced than those observed when simply comparing household disposable incomes.
Another issue of social concern is whether a specific group of people (such as an age cohort or people of a particular age) are more likely to have lower incomes than others. The likelihood of a person being in the bottom 20% of all persons, when ranked by household equivalised incomes, tends to decrease after age 30-34 years and increase from age 50-54 years. This is likely to reflect the higher unemployment rates among older men compared with middle-aged men and the impact of retirement on income (see Australian Social Trends 1999, Economic resources of older Australians). Between age 25 and 39 years, people in the younger cohorts (those born 1960-64 and 1965-69) were generally less likely to have household equivalised income in the bottom 20% than those in older cohorts at the same age. For example, in 1998-99, the proportion of people aged 30-34 years in the bottom 20% of people’s household equivalised income was lower (13%) than for the three previous cohorts (each 17%). This reflects the higher average household incomes of people in these younger cohorts and is consistent with the greater likelihood of women from these younger cohorts to be working, which will have increased household income for people in these cohorts (see Australian Social Trends 2001, Women’s incomes).
SELECTED HOUSEHOLD INCOME MEASURES OF PERSONS(a)
| Age (years) in 1984, Year of birth | |||||||||
| ||||||||||
| Age at time of survey (years) | 15-19 1965-69 | 20-24 1960-64 | 25-29 1955-59 | 30-34 1950-54 | 35-39 1945-49 | 40-44 1940-44 | 45-49 1935-39 | 50-54 1930-34 | 55-59 1925-29 | 60-64 1920-24 |
| Average weekly household disposable income | |||||||||
| $ | $ | $ | $ | $ | $ | $ | $ | $ | $ |
| 15-19 | 1,075 | - | - | - | - | - | - | - | - | - |
| 20-24 | 1,042 | 1,052 | - | - | - | - | - | - | - | - |
| 25-29 | 870 | 869 | 892 | - | - | - | - | - | - | - |
| 30-34 | 928 | 815 | 810 | 855 | | - | - | - | - | - |
| 35-39 | - | 939 | 832 | 849 | 901 | - | - | - | - | - |
| 40-44 | - | - | 964 | 941 | 984 | 973 | - | - | - | - |
| 45-49 | - | - | - | 1,077 | 949 | 1,026 | 1,066 | - | - | - |
| 50-54 | - | - | - | - | 1,033 | 928 | 920 | 953 | - | - |
| 55-59 | - | - | - | - | - | 815 | 694 | 743 | 784 | - |
| 60-64 | - | - | - | - | - | - | 592 | 575 | 532 | 585 |
| 65-69 | - | - | - | - | - | - | - | 518 | 454 | 454 |
| 70-74 | - | - | - | - | - | - | - | - | 475 | 412 |
| 75 and over | - | - | - | - | - | - | - | - | - | 453 |
| Average weekly household equivalised disposable income | |||||||||
| $ | $ | $ | $ | $ | $ | $ | $ | $ | $ |
| 15-19 | 421 | - | - | - | - | - | - | - | - | - |
| 20-24 | 520 | 523 | - | - | - | - | - | - | - | - |
| 25-29 | 496 | 494 | 494 | - | - | - | - | - | - | - |
| 30-34 | 509 | 447 | 427 | 443 | | - | - | - | - | - |
| 35-39 | - | 489 | 428 | 407 | 436 | - | - | - | - | - |
| 40-44 | - | - | 482 | 461 | 457 | 441 | - | - | - | - |
| 45-49 | - | - | - | 527 | 492 | 493 | 489 | - | - | - |
| 50-54 | - | - | - | - | 557 | 496 | 480 | 495 | - | - |
| 55-59 | - | - | - | - | - | 455 | 392 | 425 | 439 | - |
| 60-64 | - | - | - | - | - | - | 362 | 352 | 332 | 354 |
| 65-69 | - | - | - | - | - | - | - | 330 | 293 | 293 |
| 70-74 | - | - | - | - | - | - | - | - | 313 | 280 |
| 75 and over | - | - | - | - | - | - | - | - | - | 309 |
| Persons in the bottom 20% of the income distribution(b) | |||||||||
| % | % | % | % | % | % | % | % | % | % |
| 15-19 | 15.8 | - | - | - | - | - | - | - | - | - |
| 20-24 | 11.5 | 12.4 | - | - | - | - | - | - | - | - |
| 25-29 | 12.2 | 13.5 | 15.0 | - | - | - | - | - | - | - |
| 30-34 | 13.4 | 16.8 | 16.8 | 16.5 | - | - | - | - | - | - |
| 35-39 | - | 14.2 | 16.4 | 17.9 | 16.3 | - | - | - | - | - |
| 40-44 | - | - | 15.3 | 14.8 | 13.9 | 16.1 | - | - | - | - |
| 45-49 | - | - | - | 13.7 | 13.0 | 14.0 | 14.9 | - | - | - |
| 50-54 | - | - | - | - | 15.0 | 14.6 | 14.7 | 14.6 | - | - |
| 55-59 | - | - | - | - | - | 25.0 | 24.5 | 23.6 | 20.2 | - |
| 60-64 | - | - | - | - | - | - | 36.1 | 30.7 | 33.4 | 30.0 |
| 65-69 | - | - | - | - | - | - | - | 38.5 | 36.4 | 39.2 |
| 70-74 | - | - | - | - | - | - | - | - | 38.3 | 39.7 |
| 75 and over | - | - | - | - | - | - | - | - | - | 43.5 |
(b) Distribution of persons by the average weekly household equivalised disposable income of their household.
Source: ABS 1984, 1988-89, 1993-94 and 1998-99 Household Expenditure Surveys.
| Previous Page | Next Page |