Abstracts

Is the labour demand curve downward sloping for Australia?

Melissa Bond, Noel Gaston

Abstract
This paper estimates the effects of immigration on labour market outcomes of native workers in Australia. Following Borjas (2003) we aggregate workers from a national labour market into education-experience groups, assuming that similarly educated workers with differing experience levels are not perfect substitutes. We also examine the role of migrants of different ethnicities on the labour market outcomes of native workers. Using the first five waves from the Household, Income and Labour Dynamics in Australia Survey, we find significant positive effects of immigration on native earnings at an aggregate level. Specifically, a one per cent increase in the labour supply of a particular skill group increases the weekly earnings of natives in that group by about 0.4 per cent. The analysis indicates, however, that not all education groups are positively affected. In fact, those native workers who at the highest hold a Certificate or Diploma appear to be adversely affected by immigration. These findings are robust across differing specifications and migrant definitions. Finally, we find that immigrants originating from primarily non-English speaking countries have greater positive effects on native earnings than immigrants from primarily English-speaking countries.

Household asset portfolio diversification: Evidence from the Household, Income and Labour Dynamics in Australia (HILDA) Survey

Andrew Worthington

Abstract
This paper examines the impact of demographic, socioeconomic and risk aversion factors on diversification in Australian household asset portfolios using Wave 6 of the HILDA Survey. Household assets are categorised as home and other property, superannuation, equity and cash investment, business assets, bank accounts, life insurance, trust funds and collectibles. The characteristics examined include family structure and composition, the source and level of income, age, gender and attitudes towards financial risk taking. The diversification measures comprise a naïve index, a Hirschman–Herfindahl concentration index, a Shannon entropy index, absolute and relative benchmark indexes and a market asset share index. Tobit models are used to identify the source and magnitude of the factors associated with diversification. The results indicate that Australian household portfolios have very low levels of asset diversification and that the factors analysed exert a major impact. Importantly, the behaviour observed in household portfolios appears to bear little relation to the central predictions of classic portfolio theory.

Is homogamy good for marriage? A longitudinal dyadic analysis of factors associated with marital dissolution in early twenty-first-century Australia

Rebecca Kippen, Peng Yu and Bruce Chapman

Abstract
Using data from waves 1–7 of HILDA, 2,520 married couples—where both partners are respondents in the first wave—are traced over six years to identify factors associated with marital dissolution. The data are analysed dyadically; that is, the characteristics of both partners in each couple are considered in tandem. This allows assessment of whether marriages between partners with similar characteristics (homogamy) are more likely to last than are marriages between dissimilar partners, or whether particular characteristics of wives or husbands—independent of their partners’—are more strongly associated with marital stability. Factors considered include country of birth, education, income, labour force status, religiosity, family background, premarital cohabitation, age, length of marriage, number of marriages and presence/absence of children. Pooled longitudinal samples allow us to use fixed-effects modelling to control for unobserved characteristics of the individuals sampled and to test whether relationships between variables changed over the seven years of observation. Through the use of the dyadical approach, the availability of the longitudinal aspects of the panel, and the wealth of variables accessible from HILDA, the modelling allows new insights into the determinants of marital stability and instability.

Work-life tension and its impact on the work hours of Australian mothers in paid work between 2000 and 2005

Ibolya (Ibi) Losoncz,

Abstract
Australian mothers in paid work often try to reconcile two equally important aspirations: participation in the labour market, and commitment to caring for their family and children. Earlier research by Losoncz (2008), using cluster analysis on the HILDA survey data, identified six major homogenous groups of working mothers who had distinctive profiles in terms of their work-family life experienc e.

Descriptive analysis of the clusters found that mothers in the two clusters – Aspiring and struggling, and Indifferent and struggling – experienced strong tension in managing their work and family responsibilities. These mothers were characterised by long working hours, high work overload, lack of support from others, lower outcomes on health measures, and low satisfaction with family life and parenthood regardless of their level of aspiration to be in paid work.

This research will use longitudinal analysis of the first six waves of HILDA to assess if mothers in these two clusters are more likely than mothers in other clusters to reduce their working hours or exit paid work. Furthermore, the research will investigate the impact this reduction in working hours, or exit from paid work, have on reports of work-life balance, physical and mental health, and satisfaction with family life and parenthood by these mothers.

Social participation of youth growing up with disability: a study with the first seven waves of HILDA

Peng Yu

Abstract
This research investigates the impact of disability on social participation of youth in Australia and aims to contribute to the evidence base for the Social Inclusion Agenda of the Australian Government using the first seven waves of the Household, Income and Labour Dynamics in Australia (HILDA) Survey. Among a large stock of literature on the impact of disability, there are relatively few studies which focus on youth with disability, and even fewer published papers focus on their social participation. Social participation may be as important as economic participation, because it not only helps people with disability to accumulate social capital, but also increases their social interaction within the community, influences public attitudes towards disability, and promotes public awareness and acceptance of people with disability as an integrated part of the community.

HILDA is a large, nationally representative Australian panel survey. Along with many individual, family background and contextual variables, HILDA has rich information on social participation. This research chooses active membership of clubs and associations, social interaction with friends and community participation as key indicators of social participation. Community participation is examined using the 12 items collected in wave six of the survey, from which an underlying construct of social involvement is derived through factor analysis.

In this study, disability is defined as having any long-term health condition, impairment or disability that restricts one’s everyday activities, and has lasted or is likely to last, for 6 months or mor e. The effects of the incidence, onset, duration and severity of disability on the selected indicators of social participation are explored using one or multiple waves of HILDA where appropriate; both bivariate and multivariate analytical techniques are applied. The research also examines how disabilities and disadvantages in other aspects (for example, low family income) intersect to impact on social participation of young peopl e. The results are useful for informing appropriate interventions, not only regarding current youth but also for future generations.

Wages, occupational choice and personality

Michelle Tan, Deborah Cobb-Clark

Abstract
A large literature has investigated the implications of gender segregation for men's and women's labour market outcomes, in particular wage rates. However, much less is known about the process which leads to ccupational segregation in the frst plac e. Understanding occupational segregation is particularly important in light of its potential effect in driving the gender wage gap. When examining the aggregate gender wage gap, most researchers account for occupation by including a set of occupational dummy variables into the model. Unfortunately,this approach implicitly assumes that occupational distributions are determined exogenously. This paper re-examines occupational segregation in the context of a model that explicitly accounts for gender differences in a number of psychological traits that are often speculated to be at the heart of gender differences in job choic e. Specifcally, we examine whether an individual's personality (as measured by the Big Five), self{effcacy and locus of control (as measured by the Pearlin and Schooler self-effcacy scale) are important drivers of occupational attainment, and if so whether gender differences in these traits can help us understand occupational segregation in Australia. We then go on to incorporate this model of occupational choice in a decomposition of the gender wage gap. Using data from HILDA (2001-2006) the decomposition employed accounts for the contribution of relative wages both within and across occupations to the overall gender wage gap. The methodology allows us to explicitly consider the ways in which gender differences in psychological traits and the returns to those traits affect occupational segregation, as well as the way in which occupational segregation subsequently affects the relative wages of Australian women.

We fnd that for some occupations, certain personality traits have a non trivial effect on the probability of being observed in that occupation. On aggregate, however, the reasons behind the occupational segregation observed in Australia, cannot be explained by differences in tastes and preferences for occupations, as measured by the Big Five traits, self-effcacy and locus of control. To summarise our results, we nd that Australian women earn less on average because they earn less than men when employed in the same occupation, not because they are employed in different occupations to men. Interestingly, virtually all of the inter-occupational component of the aggregate gender wage gap is unexplained. This suggests that occupational segregation occurs because men and women with the same characteristics have a very different propensities to enter specific occupations. Finally, accounting for gender differences in the process that leads men and women to be in different occupations has important implications for the proportion of the wage gap that can be explained.

Work, household dynamics and housing insecurity

Sharon Parkinson,

Abstract
The expansion of the standard employment contract along with the growth in female labour participation has given rise to a range of employment arrangements amongst households in terms of hours worked and mix of employment contracts. Distinctions can now be made between working and jobless households and also according to the presence of permanent and non permanent employment within a household. These changes have occurred alongside many significant shifts within the housing market, including declining affordability and low rental vacancies. Together these trends raise concerns about the capacity to maintain housing, not only amongst households who become unemployed, but also amongst those relying solely on a non permanent or variable incom e.

The paper uses the first three waves of HILDA data to model the connections between more insecure forms of employment, household conditions and housing insecurity, placing particular emphasis on the employment composition of the household. A static logistic regression model with random effects is initially applied to two different measures of housing insecurity for both renters and purchasers. This is followed by dynamic models of the factors associated with transitions into and out of purchased housing. The results point to a strong relationship between non permanent employment and housing insecurity after controlling for income and other household characteristics.

Disentangling overlapping seams: the experience of the HILDA Survey

Nicole Watson

Abstract
An issue unique to longitudinal surveys is seam effects. These occur when there is a tendency for changes in the data to unusually concentrate in adjoining periods from different interviews. One component of the HILDA Survey subject to seam effects is the labour market activity calendar. In this calendar respondents are asked to recall the various jobs they have had over a 14 to 18 month period, the time spent in unemployment and the time spent outside the labour forc e. As the calendar is administered every wave, an overlap of 2 to 6 months results depending on when the respondent is interviewed.

In this paper, we separately model the likelihood that respondents will make three types of errors in the activity calendar: i) reporting a spell in the first version of events and not in the second; ii) reporting a spell in the second version of events and not in the first; and iii) misplacing a spell in the second version of events compared to the first. The characteristics considered in the model include the various causes of errors in dating events, such as spell length, spell type, duration of the overlapping seam, recall ability of the respondent, and characteristics of the interview that may affect the respondent’s recall. The overlapping seam also permits the study of measurement error over time to identify whether the same people continually make the same mistakes.

With a better understanding of the types of errors that respondents make, the HILDA team hopes to construct a consolidated labour market activity spell file that will encourage greater data use of a section of the HILDA Survey that has so far been underutilised.

Determinants of maternity leave duration in Australia: evidence from the HILDA Survey

Aydogan Ulker, Cahit Guven

Abstract
Length of maternity leave taken has been found to influence the well-being of working mothers and their newborn children through factors such as the quality of mother-infant interactions, the incidence and duration of breastfeeding, and immunisations. Maternity leave duration has also direct implications for pre-birth employed mothers’ future labour market activities in terms of both labour force participation and working hour schedules. We use the first five waves of the HILDA survey to examine the determinants of maternity leave taken (both paid and unpaid) by pre-birth employed mothers in Australia. Our goal is to identify the relevant factors for public policy which influence the mothers’ post-birth return to employment decisions. In this regard, we utilize the rich information on the socio-demographic and economic characteristics of mothers. Specifically, we consider the roles of household wealth, childcare accessibility, pre-birth job characteristics such as maternity leave rights, hours flexibility, and wage levels as well as the roles of household structures, education levels, and attitudes towards family and employment situations. The findings of this paper provide strong insights on the current policy debate regarding universally paid maternity leav e.

Does Labour Market Achievement Matter for the Wellbeing of Australian Immigrants? Culture and Gender Differences

Weiping Kostenko

Abstract
This study explores the Australian immigrants' job-life relationship by simultaneously estimating a bi-variate ordered probit random effects panel model. We found that discrepancy between career goal andemployment reality plays a central role. The study also explores the characteristics related to immigrantswho are likely to have stronger job concerns. Non-western male immigrants are among them. Theyare more disadvantaged in the labour market and have lower life satisfaction compared to their Westerncounterparts, while these situation improve with duration in Australia. Also, immigration age is foundcrucial for this adjustment process. For female immigrants, the results suggest that the well-educatedfemale migrants' subjective wellbeing is impeded by struggling over work-family balance.

Effects of work-life harmonization on the birth of a second child in Australia: an event history analysis using HILDA data

Hideki Nakazato

Abstract

In many countries including Australia and Japan, there has been growing concern about the challenges of balancing (or harmonising) paid work and life outside of work, including family life and parenting. It has been suggested that these concerns are often related to low fertility. Although the recent total fertility rate in Australia is much higher than in lower fertility countries such as Japan, avoiding further fertility decline remains an important priority for the political agenda of both countries. Despite the prevalence of discussions linking work–life issues with fertility, there is little statistical evidence of a causal relationship. This seems to be mainly because of the difficulty in inferring causality with cross-sectional research design. The HILDA (Household, Income and Labour Dynamics in Australia) survey, as a panel data set, is a rich source of information for causal analyses.

In this paper I examine how work-life experiences of partners affect fertility in Australia. Specifically, I focus on the effects of work/life experiences of mothers after her first birth on parity progression to her second birth. I use waves 1 to 6 of the HILDA survey for this analysis. Event history analysis is conducted to make the best use of the longitudinal data within a limited observation period.

In preparing for this analysis, I have matched each woman's responses with those of her partner. We can thus obtain the information on her partner’s work and life situation, such as hours in paid work and those in child rearing, as well as her own experiences. I have also matched responses of each woman across different waves to capture a change in the number of children as well as lagged effects of independent variables at one wave, such as work hours and parenting experiences, on parity progression by the following wav e.

The results obtained from discrete-time event history analyses are as follows. Compared to the reference category (working for 1-19 hours per week), both mothers at home and mothers working fulltime are significantly less likely to proceed to a second child by the following wav e. For mothers working full-time, their partner’s hours in child rearing are positively related with the earlier progression to a second child. Further results will be shown in the complete paper.

Social capital and three forms of health in Australia.

Jennifer Welsh, Helen Berry

Abstract
Social capital is associated with improved health but components of social capital and their associations with different types of health are rarely explored together. The aim of this study was to explore the relationships between structural and cognitive components of social capital with three forms of health – general health, mental health and physical functioning.

Data were taken from Wave 6 of the Household, Income and Labour Dynamics of Australia Survey. The structural component of social capital (community participation) was measured using a twelve-item short form of the Australian Community Participation Questionnaire, and the cognitive component (social cohesion) by sense of belonging, tangible support, trust and reciprocity. Multiple hierarchical regression modelling was used to investigate multivariate relationships among these factors.

Higher levels of participation were related to higher levels of social cohesion and to all three forms of health, particularly strongly to mental health. These findings could not be accounted for by sex, age, Indigenous status, education, responsibility for dependents, paid work, living alone or poverty. Controlling for these and physical health, structural and cognitive components of social capital were each related to mental health, with support for a mediated relationship between the structural component and mental health. Women reported greater social capital, yet worse health.

Social capital was related to three forms of health and may play a particularly important role in mental health. Notable gender differences in this relationship were evident, with women reporting greater social capital than men and worse mental health.

Understanding the mechanisms underlying this apparently anomaly needs further exploration. Because community participation is amenable to intervention, our findings may assist in the development of programs which are effective in promoting mental health and also social cohesion.

Income and happiness: an analysis of adaptation and comparison income effects

Satya Paul and Daniel Guilbert

Abstract
This study has examined the effects of adaptation and social comparison on the self reported life satisfaction (happiness) of individuals in Australia based on panel data drawn from the five waves (2001-2005) of HILDA surveys. The empirical results show that over the period 2001-2005, the incomes have grown steadily yet the level of happiness has remained constant. The results provide no evidence of adaptation to incom e. However results offer strong statistical support to the hypothesis that reference/ comparison group income has a negative effect on the self reported well being of individuals. The increase in peer group income hurts the poorer more than the richer on e. Some other variables such as age, sex, health, maritus status, employment status, location, work hours and voluntary work are found to have significant effect on happiness.

Chronic and transitory poverty over the life cycle

Joan Rodgers and John Rodgers

Abstract
This study estimates chronic and transitory rates of poverty in Australia using the concept of permanent income and seven years of longitudinal data. We sketch a picture of chronic and transitory poverty over the life cycle by decomposing the poverty rate for the entire population into poverty rates for age categories that range from childhood, through adulthood, to old age. The measure of household income that underlies our poverty rates includes imputed rent on owner-occupied housing, which has a large influence on measured poverty, particularly among older members of the population. We also document changes in chronic and transitory poverty from 2001-03 to 2005-07 and use a jack-knife procedure to test the statistical significance of the poverty-rate changes we observe.

The gender wage gap within the managerial workforce: an investigation Australian panel data

Ian Watson

Abstract
This paper examines the gender earnings gap within the managerial workforce in Australia, drawing on the HILDA. As is well known, career paths for women are truncated within the managerial workforce, particularly at higher levels. This has given rise to the aptly named `glass ceiling'. An important dimension of this problem is a lower level of earnings among women managers, compared to their male counterparts. Using regression techniques to decompose this earnings gap allows researchers to ask how much of this gender difference might be due to the differing characteristics of the individuals, and how much might be due to other unobservable factors, including discrimination.

Following a long tradition of decomposing the wage gap using Blinder- Oaxaca type approaches, this paper extends the analysis in a number of ways. It uses several waves of data, and fits multi-level models to these data. This not only deals with the practical problems of small sample size, but it confers several advantages. It takes account of correlated error terms (inherent in clustered observations) and it also produces coefficients which are more reliable. The modelling approach also takes account of several `selection effects', including selection into working and selection into the managerial workforce.

Finally, the paper also looks at more recent decomposition strategies which make use of simulation approaches. This framework is particularly well suited to policy research, since it allows one to specify quite precisely what factors contribute most to the gender earnings gap.

Stability of lone mothers’ employment: Using HILDA calendar data to examine work transitions

Jennifer Renda, Jennifer Baxter and Matthew Gray

Abstract
While more mothers have been participating in the paid workforce over recent years, the employment rate of lone mothers remains lower than that of couple mothers. Increasing employment rates of lone mothers is a goal for policy makers as participation in paid work is seen as central to reducing poverty and welfare dependence within lone parent families. However, if the work that lone mothers’ undertake is unstable they may continue to be at risk of poverty and welfare dependenc e. In order to examine whether or not lone mothers are more likely than couple mothers to experience unstable employment, calendar data from all seven waves of the HILDA survey are used to identify and compare the rate at which lone and couple mothers move into and out of employment. The calendar divides each month into thirds, and captures employment status in each of these time periods, so short periods of employment and non–employment can be measured. These data show that in any period lone mothers are less likely to be employed than couple mothers. Of those employed in a period, lone mothers are more likely to transition out of employment than couple mothers and of those not employed, lone mothers are less likely to transition into employment. The analyses also consider whether it is lone parenthood or other characteristics that differentiates lone and couple mothers in their employment transition rates. These analyses show that educational attainment and age of youngest child may influence, in part, the different employment transition rates of lone and couple mothers.

The retirement – consumption puzzle downunder

Garry Barrett and Mat Brzozowki

Abstract
As increasing numbers of workers approach retirement, an issue of growing importance for policy makers is whether households have sufficient savings to maintain their standard of living in retirement. A substantial body of international research (e.g. Hamermesh 1984; Banks, Blundell and Tanner 1998; Bernheim, Skinner and Tanner 2001) has shown that household expenditure systematically decreases at the time of retirement – a finding that which is inconsistent with the standard life-cycle model of income and savings if retirement is an anticipated event. This fall in expenditure has become known as the ‘retirement-consumption puzzle.’ In this paper we analyse HILDA Survey data from waves 1 to 6 to assess the Australianevidence on the retirement-consumption puzzle and attempt to reconcile disparate finding in the literature.

The main contributions of our analysis are:
(i) Consideration of multiple dimensions of household well-being. In particular, we exploit the richness of the HILDA data by considering the effect of retirement on changes in household expenditure, wealth, indicators of financial hardship, and time use. Examining these alternative
dimensions of wellbeing allows us to reconcile various findings in the literature which focus on single dimension of wellbeing usually drawn from disparate data sources.
(ii) Extension of the recent work by Aguiar and Hurst (2005) and Brzozowski and Lu (2009) who develop measures of food ‘consumption,’ which is a combination of market expenditure and home production drawn from time-use information,
(iii) Implementation of econometric techniques which exploits the panel structure of the HILDA data. In contract to most studies in the literature which rely on cross-sectional data, we estimate variety of richer models which include conditional mean estimators with individual random and
fixed effects, random effect dynamic probit and fixed effect logit models, plus instrumental variable models.

The main findings from the analysis include that there is little evidence of a fall in consumption upon retirement. The results for consumption are consistent for our finding regarding wealth changes over time, where there is little evidence of dissaving among retirees. However, we do
find retirement has strong negative effects on individual household’s self-reported ability to‘make ends meet’ as measured by their ability to pay bills or their need to ask for financial help from family, friends or welfare institutions. Our ongoing research attempts to reconcile these
findings. Finally, in regards to methodology, we assess the performance of alternative instrumental variables for retirement in IV models. We show that the choice of subjective retirement expectations or a series of age indicator variables for instrumenting retirement status
is crucial for the estimated impact of retirement on changes in household expenditure and wealth.

Bivariate income distributions for assessing inequality and poverty under dependent samples

Andrea Vinh, William E. Griffiths and Duangkamon Chotikapanich

Abstract
As indicators of social welfare, the incidence of inequality and poverty is of ongoing concern to policy makers and researchers alike. Of particular interest are the changes in inequality and poverty over time, which are typically assessed through the estimation of income distributions. From this, income inequality and poverty measures, along with their differences and standard errors, can be derived and compared. With panel data becoming more frequently used to make such comparisons, traditional methods which treat income distributions from different years independently and estimate them on a univariate basis, fail to capture the dependence inherent in a sample taken from a panel study. Consequently, parameter estimates are likely to be less efficient, and the standard errors for between-year differences in various inequality and poverty measures will be incorrect. This paper addresses the issue of sample dependence by suggesting a number of bivariate distributions, with Singh-Maddala or Dagum marginals, for a partially dependent sample of household income for two years. Specifically, the distributions being considered are the bivariate Singh-Maddala distribution, proposed by Takahasi (1965), and bivariate distributions belonging to the copula class of multivariate distributions, which are an increasingly popular approach to modelling joint distributions. Each bivariate income distribution will be estimated via full information maximum likelihood using data from the Household, Income and Labour Dynamics in Australia (HILDA) Survey for 2001 and 2005. Parameter estimates for each bivariate income distribution will be used to obtain values for mean income and modal income, the Gini inequality coefficient and the headcount ratio poverty measure, along with their differences, enabling the assessment of changes in such measures over time. In addition, the standard errors of each summary measure and their differences, which are of particular interest in this analysis, will be calculated using the delta method.

Interrelated dynamics of health and poverty in Australia

Hielke Buddelmeyer and Lixin Cai

Abstract
Using the Households, Income and Labour Dynamics in Australia (HILDA) survey, this study examines the joint dynamics of health and poverty of Australian families. Taking advantage of panel data, the modelling approach used in this study allows a better estimation of the causal relation between health and poverty. The results indicate that the causality runs both ways and the relationship could be confounded by unobserved heterogeneity. In particular, it is found that families headed by a person in ill-health are more likely to be in poverty compared with families headed by a person with good health. On the other hand, a family head whose family is in poverty in the current year is more likely to be in ill-health in the next year compared with a family head whose family is not in poverty. In addition, there is evidence that health and poverty are affected by correlated unobserved determinants, causing health to be endogenous to poverty. Consequently, treating health as exogenous in a poverty equation would produce a biased estimate for its effect.

Cohabitation Outcomes:The Effect of Fertility Intentions, Relationship Satisfaction and Union Length on Cohabitation Transitions

Sandra Buchler, Janeen Baxter , Michele Haynes and Mark Western

Abstract
Rates of cohabitation have increased significantly over the past few decades, and while a substantial amount of research has been conducted on cohabiting relationships, relatively little is known about the pathways that these relationships follow. In this research we use waves 1 – 7 of HILDA to
investigate the influence of fertility intentions, relationship satisfaction and union length on the likelihood that a cohabiting person will transition to either married or single over the seven waves. In particular, we focus on whether the influence of these factors on relationship transitions is mediated by a cohabitor’s marital history and intention to marry. We track transitions using lagged variables and predict a number of logistic regression models with and without controlling for covariates and other independent variables. We find that the influence of fertility intentions, relationship satisfaction and union length on relationship transitions varies by marital history and intention to marry.

The effect of life histories on repartnering in Australia and the United Kingdom

Edith Gray, Alexandra Skew and Ann Evans

Abstract
In recent years as a result of a rise in divorce rates coupled with an increased prevalence of cohabitation, a growing percentage of the population has or will experience the breakdown of a relationship and also the possibility of forming another new relationship. It has therefore become increasingly important to understand how people repartner after the dissolution of a previous union. Although a large body of literature already exists on the study of remarriage, there is far less research which has investigated repartnering in the form of a cohabiting union. Further, much of this work focuses on those who have been previously married, and less is known about patterns of repartnering after the breakdown of a cohabiting relationship (Wu and Schimmele, 2005) . This paper seeks to address the issue of repartnering, both in terms of forming cohabiting and marital unions, from a comparative perspective. Using a longitudinal approach we compare the nature of repartnering behaviour in Australia and the United Kingdom, countries with similar policy and legislative frameworks. We find that within five years of becoming single, an estimated 49 per cent of the United Kingdom sample and 43 per cent of the Australian sample had entered a new relationship, most commonly cohabitation. Multivariate analysis reveals important similarities as well as differences in the demographic and socio-demographic determinants of forming a new union in the two countries.

The effect of changes in the aggregate employment rate on the composition of employment:An Australian Case Study (2002-2006)

Jeremy Lawson and Crystal Ossolinski

Abstract
A pronounced and extended rise in the employment rate might be associated with a change in the distribution of employment across the population. Here, we explore which personal characteristics are associated with employment and how, if at all, these relationships may have changed over time. Using unit-record data from the HILDA survey in 2002 and 2006, we estimate a binomial logit model to establish the relationship between employment and personal and household characteristics. The same relationship is then estimated for 2002 and 2006 data to establish if this relationship has changed in light of the significant increase in employment over this period. Although there was not a statistically significant change in the relationship, the point estimates suggest a broadening of employment over this period; the relative employment prospects of people with “low employment” characteristics in 2002 had improved by 2006.

To explore whether a change in participation or unemployment prospects had contributed to this result, we then separate individuals outside of employment into the unemployed, marginally attached and not in the labour force (NILF) and estimate a multinomial logit model. We find that unemployment was associated with different characteristics to those associated with NILF, while the marginally attached shared characteristics with both groups. Between 2002 and 2006, we find that the concentration of “low employment” characteristics in the unemployment pool was broadly unchanged. However, for both NILF and marginal attachment, the concentration of “low employment” characteristics decreased, suggesting that increased supply by these groups contributed substantially to employment growth.

Impact of qualifications and work-related training on employment growth in innovating industries.

Cezary Kapuscinski

Abstract

Post-compulsory education has always been an acknowledged factor enabling success in the labour market. Recent socio-economic developments (such as more competitive labour and product markets, increase in the average level of education over time or demographic changes impacting on the labour force) have emphasized the importance of not only the foundation education (i.e. education gained before entering the labour force) but also of various forms of life-long learning – from formal and informal workplace training and skilling to part-time and full-time upgrading of the level of qualification and subsequent job mobility. Such upgrading of the human capital, of course, does not occur in the vacuum – firms upgrade their physical capital, invest in research and development and innovate their business operations and practices in order to gain competitive advantage. In the process, both the firms and employees are rewarded from innovation and higher productivity.

This paper will investigate the impact that upgrading of qualifications and skills of employees has on job mobility utilizing the labour dynamics information from HILDA data set. Specifically, the analysis will focus on the impact of work-related training on subsequent labour mobility. Using additional information from the Business Characteristics Survey relating to innovation performance of sectors/industries, we will look at the importance of job-related training as a determinant of mobility into the innovating sectors of the economy. In addition, the paper will also investigate differences in participation in work-related training by workers in sectors that are leaders in innovation and in sectors that are at the bottom of the innovation activity. Closely linked to mobility is also the issue of job duration and the paper will also look at the impact of upgrading of human capital on duration of employment in innovating and non-innovating industries.

Low-paid employment and unemployment dynamics in Australia

Hielke Buddelmeyer, Wang Shang Lee and Mark Wooden

Abstract
This paper uses longitudinal data from the Household, Income and Labour Dynamics in Australia (or HILDA) Survey to examine the extent to which the relatively high rates of transition from low-paid employment into unemployment are the result of disadvantageous personal characteristics or are instead a function of low-paid work itself. Dynamic random effects probit models of the likelihood of unemployment are estimated. After controlling for unobserved heterogeneity and initial conditions, we find that, relative to high-paid employment, low-paid employment is associated with a higher risk of unemployment, but this effect is only significant among women. We also find only weak evidence that low-wage employment is a conduit for repeat unemployment.

The effect of permanent income on completed schooling in Australia

Jessica Todhunter

Abstract

Child poverty is a concern to any society and the link between childhood poverty and schooling completion is one of the key factors causing cycles of intergenerational poverty in Australia. In particular, the literature suggests that it is sustained, or permanent, low income which has the most serious and long-lasting consequences on children, rather than transitory income shocks. This paper investigates the effect of permanent income during childhood on adolescent's schooling completion using the first six Waves of the HILDA Survey. The results show that whilst permanent income during adolescence is positively associated with the probability of completing Year 12 or a trade qualification, other factors including parental education and neighbourhood characteristics, have a far greater impact on adolescent's schooling preferences.

Lifetime economic consequences to women informal carers in Australia, 2006

Binod Nepal, Laurie Brown, Geetha Ranmuthugala, Richard Percival

Abstract

Home-based care provided by family members is the most common form of caring for people with disabilities in Australia. However, this model of care is generating enormous financial consequences for informal carers. This study examines the impact of taking on a primary carer’s role on financial well-being of women in Australia over the course of their working life, by comparing women primary carers and other women. Estimates are based on information drawn from the Household, Income and Labour Dynamics in Australia (HILDA) Survey Wave 6 and the life tables for Australian women. The economic indictors examined cover labour force participation and income. Income indicators include individual and family income from wages and salaries, and government benefits (public transfer). While there may be a wide range of caring situations, we focus on women aged 30 to 64 years, with two or more children, who are primary carers to their child with a disability. These women are divided into four categories by partnership status (single vs partnered mothers) and educational attainment (secondary school or less vs post-secondary education). Results show that over half of the women primary carers aged 30 to 64 years are not in the paid labour force compared to less than a third of other women in the same age group. Primary carers who do work spend fewer hours in paid employment than do other women. As a consequence of limited participation in paid work, primary carers earn considerably less income from wages and salaries over their working life compared to women with similar demographics. Mothers caring for a child with a disability are likely to earn over their working life, depending on their level of education, between a quarter and a half the income of women sharing the same demographics but who are not primary carers. While carers receive more in government benefits than other women, these payments do not compensate fully for the income they forgo from paid work. In conclusion, women primary carers have a bleaker financial prospect compared to that of other women. Policies for ensuring financial security of women primary carers are warranted.

Dynamics of female labour supply: evidence of differences in preference between single and couple females

Laurence Lester and Darcy Fitzpatrick

Abstract
This Paper presents the results of econometric modelling of female labour force participation and the supply of hours (contingent on being employed) for females in Australia based on the first six waves of the HILDA data. (The Household, Income and Labour Dynamics in Australia (HILDA) survey (funded by the Australian Government Department of Families, Housing, Community Services and Indigenous Affairs) is designed and managed by the Melbourne Institute of Applied Economic and Social Research at the University of Melbourne.)

We use sophisticated panel (longitudinal) data econometric models which are based on an extensive review of the recent theoretical and applied econometric literature addressing labour supply for single and partnered individuals. Applied econometric models of labour force participation and hours of labour supply in this Paper:

a) investigate the set of factors which influence women’s decisions, the relative importance of explanatory factors, and implied semi-elasticities (i.e. the percent change in the dependent variable for a one unit change in an explanatory variable);

b) control for unobserved individual level attributes or characteristics (i.e. unobserved heterogeneity);

c) incorporate dynamics to control for the influence of previous period values and “state dependency” on the current value of the dependent variable;

d) adopt a two-stage selection model to account for potential bias in econometric estimates due to “selection bias” in models of hours of labour supply (i.e. labour supply is contingent on a labour force participant female being employed); and

e) analyses separate models for single females and for females with male partners.

Following the results of econometric modelling, we canvas the implications of the econometric model results for influencing the labour supply of women.

Analysis of the added worker effect and the discouraged worker effect for married women in Australia

Xiaodong Gong

Do changes in the lives of our peers make us unhappy?

Tony Beatton and Paul Frijters

Abstract
In this paper, we seek to explain the changes in aggregate happiness over the lifecycle. The advantage of looking at the aggregate level of happiness is that it solves the problems of missing peer effects and measurement error that plague models of individual level happiness, though the disadvantage is a dramatic loss of degrees of freedom. We use panel data from the Household Income and Labour Dynamics for Australia (HILDA), which allows us to construct an index of the severity of life changes for each age. This single-variable Stress Index is able to explain over 80% of the variation in happiness over time. Unexpectedly, aggregate ‘positive stress’ (such as marriage rates by age or levels of job promotion) have greater negative effects on aggregate life satisfaction than negative stress (such as negative financial events or deaths of spouses), which we interpret as a strong indication that what is deemed a positive event by the person involved is a highly negative event for his or her peers. We find some evidence that extraverted individuals get affected less negatively by stress. The happiness maximising policy is then to reduce changes over the life cycle to the bare minimum needed to sustain a dynamic economy and to sustain procreation.

Precarious Employment and Employees’ Self-rated Mental Health in Australia

Guangyu Zhang and Sue Richardson

Abstract
There is growing evidence that precarious employment is harmful to health. In Australia, such employment comprised one third of the total work force in 2006, much of which involved employment on casual terms. Australia is an interesting case, since it has a century-old system of protections for workers and a distinctive welfare state. Our hypothesis is that these protections reduce the harm to health arising from precarious employment. Employing six waves of longitudinal survey data and logistic regression models, this paper examines the mental health consequences of employment on a casual or fixed-term basis compared with permanent employment. Gender differences are explored. The self-rated mental health outcome was measured by the SF-36 health instrument. Key control variables include demographic and socio-economic characteristics, occupation, negative affectivity, and the level of social support. The paper also examines whether there were any lagged mental health consequences from current precarious employment. We find that most precarious employees did not have systematically worse mental health; however, male casual and female fixed-term employees working full-time were more likely to report mental health problems. No lagged mental health effects were found.

Economic independence or bargaining power? The relationship between women’s earnings and housework time

Janeen Baxter and Belinda Hewitt

Abstract
Studies of gender differences in the domestic division of labour have typically included measures of the relative earnings of husbands and wives to help explain how much time each spends on household tasks. The theoretical underpinning for this approach is based on a range of perspectives including role specialization, relative resources and household bargaining. Recently, Gupta (2006, 2007) has argued that wives absolute earnings are more important than relative earnings in determining how much time wives spend on housework. Using US data, he shows that women’s housework time depends more on the magnitude of their earnings than their share of earnings relative to their husbands, and that income differences amongst women are as important in explaining women’s housework time as the difference between husbands and wives earnings. One explanation Gupta proposes for this finding is that women with higher earnings can buy their way out of domestic labour because they can afford to hire paid domestic help, although he is unable to explicitly test this.

The current paper builds on Gupta’s earlier work in two main ways. First we use recent Australian data to examine whether women’s housework time in Australia is associated with relative or absolute earnings. Second we examine whether the use of paid domestic help reduces women’s time on domestic labour. We investigate these issues using data on 1,374 married or cohabiting couples from Wave 5 of HILDA where the female partner reports working full or part time. We are limited to using Wave 5 since information about time use in households, as well as use of paid domestic help was not collected in other waves of the survey.

Our results differ from those of Gupta’s in the US. We find more evidence for the relationship between relative earnings and women’s housework time than absolute earnings and housework time. Even though absolute earnings are negatively associated with women’s housework time in baseline models, this association becomes non-significant when relative earnings are included in the models. Moreover, relative earnings remain statistically significant even when other controls are included. Using paid help does not significantly reduce women’s time on housework once other factors are controlled.

Overall our findings, like those of Bittman et al (2003), support an exchange bargaining perspective in Australia. This further suggests that there are real national differences between Australia and the US that are related to the presence of a strong male breadwinner institutional framework in Australia.

Testing Sen’s capability approach to explain objective and subjective well-being using German and Australian Panel Data?

Ruud Muffels and Bruce Headey

Abstract
There is renewed interest in Sen’s capability approach but still a dominant lack of empirical research in measuring and testing his theoretical model especially in a dynamic context. We elaborate a ‘stock-flow-outcome’ approach to explain objective (OWB) and subjective well-being (SWB) indicated by income and employment security and life satisfaction. We use two of the richest panel data sets in terms of breath (GSOEP with 24 years of data) and depth (HILDA with 6 years of data but broad coverage of issues). They provide evidence for two countries which represent according to the VOC literature (Hall and Soskice, 2001) two different types of market economies: a strongly coordinated ( Germany) and a weakly coordinated, more or less liberal type ( Australia). The main question addressed is to what extent Sen’s CA model is able to explain well-being in these two regimes. We particularly view the separate effects of stocks, functionings and events. The findings strongly support Sen’s capabilities framework for explaining well-being. Capabilities indicated by ‘stocks of social, human and cultural capital’ contribute most to explaining income security and life satisfaction whereas functionings indicated by flows and investments in stocks including life course events contribute most to explaining people’s employment security. Sen’s CA model seems therefore promising for explaining well-being. The results show that the capability approach makes ‘Sen-se’ and they also echo the contended particular features of the so-called coordinated and liberal or unregulated types of market economies.

The effect of ethnic residential segregation on wages of migrant workers in Australia

Mathias Sinning, Matthias Vorell

Abstract
This paper investigates the effects of ethnic residential segregation on wages of migrant workers in Australia. The empirical analysis uses data from the
Household, Income and Labour Dynamics in Australia (HILDA) Survey in combination with regional information from Australian Census data. We account for the non-random sorting of individuals into neighborhoods by instrumenting for neighborhood attributes and considering regional fixed effects. Post code areas in which immigrants can gain additional points when entering Australia through the point system are used as an instrument for endogenous location choices. Our findings reveal that ethnic residential segregation has a significantly positive effect on wages of migrant workers and therefore provide support for the external validity of existing quasi-experimental evidence on refugees.

Housing affordability dynamics in Australia 2001-06

Rachel Ong, Gavin Wood

Abstract
This paper investigates the dynamics of housing affordability in Australia over the period 2001-06 using waves 1 to 6 of the Household, Income and Labour Dynamics in Australia Survey. We utilise a discrete time hazard model approach to examine whether housing affordability stress has increased over the period 2001-06, and whether spells in housing stress are transient or persistent. The results suggest that while most Australians escape unaffordable housing circumstances, there is a minority for whom unaffordable housing circumstances are a long term experience. Panel models are employed to estimate the impacts of socio-demographic characteristics on the probability of exiting from a spell of (un-) affordable housing given (un-) affordable housing in the previous year. The model findings indicate that those with children and have no employment are more prone to persistent housing affordability stress. However, residential moves during spells of housing affordability stress tend to alleviate housing cost burdens. Survival in affordable housing has become progressively more difficult over the 2001-06 timeframe, particularly for owner purchasers. This finding is unsurprising given a hous price boom and rising interest rates over the period of analysis. Residential moves are again influential, but those made by households during a spell living in affordable housing are associated with the onset of housing affordability stress.

Marital loss and mental health: gender and the importance of social support

Belinda Hewitt, Gavin Turrell and Katrina Giskes

Abstract
We investigate the impact of transitions out of marriage self-assessed mental health of men and women and examine whether and to what extent social support moderates the association. Change in the level of social support people have after separation or widowhood has been proposed as an explanation for their poorer mental health relative to married people. Few studies, however, investigate the issue. We examine this separately by gender, because women report poorer overall mental health than men and have been found to experience a greater decline in mental health with the loss of a marriage.

We use Waves 1 to 6 of HILDA, our analytic sample includes all people who were married at Wave 1 and we follow them through to wave 6, regardless of missing waves. Our final sample includes 3,024 men with an average of 4.5 wave observations and 3,225 women with an average of 4.6 wave observations. The use of multiple observations enables us to better control for selection and causality in the effects of the transitions out of marriage on mental health, as this data structure violates the assumption of independent observations we use random effects modeling. Our dependent variable is the mental health scale from the SF-36 measured at each wave. The main independent variable is marital status and lagged marital status. We develop an indicator for social support by summing respondents’ scores (1-7) on 10 questions relating to different types of social support; a positive score indicates a higher level of social support. Controls are included for age, marriage duration, socioeconomic status, ethnic background and children.

Compared to continuously married men, those who separated experienced a significant 10 point decline in their mental health, and widowed men experienced a significant decline of 11 points. Women experienced similar declines in their mental health, with separated women 9 points lower and widowed women 13 points lower than continuously married women. We also interacted social support with marital status and lagged marital status. The results of this model indicated the level of social support people had when they separated or become widowed is very important for their mental health. Men and women who reported high levels of social support had mental health scores similar to those who were continuously married. The level of social support had a slightly stronger effect on men’s mental health after separation or widowhood than women’s, but this difference was not significant.