Published work

Labour market effects of bushfires and floods in Australia: a gendered perspective

with Josiah Hickson Economic Records Volume 98, Issue S1 pages 1-25

We study the labour market effects of bushfires and floods within Australia over the past two decades, focusing on gender as a determinant of vulnerability. Whilst floods unambiguously lifted the labour supply of both genders (creating around 13,500 jobs per year), the likelihood of female employment is particularly vulnerable to bushfires – lowering by 1.6 percentage points (or around 5,000 jobs per year). This effect is partially explained by industry sector, with bushfires lifting overall male employment through industries including mining and transport, whilst reducing more female dominated services-sector participation. We also examine intrahousehold dynamics, finding strongevidence for an ‘added worker’ effect.

Working paper version

Working Papers

Trade and gendered labor outcomes: evidence from changing export demand in Indonesia

with Ridho Izzati & Donny Pasaribu

We present new evidence on the gendered consequences of trade, specifically examining shifts in foreign demand. We construct panel mean-aggregated local labor market outcomes for women at the provincial level. Our Two-Stage Least Squares (2SLS) approach suggests that an increase in foreign demand shock has a limited impact on reducing gender gaps and negatively affects total employment and the share of female-dominant sectors. This adverse employment effect is primarily driven by resource-based sectors, typically male-dominant industries. Further investigation reveals that the most affected women workers are characterized as married and low-educated. These results offer crucial insights into trade and women’s empowerment: relying on resource-basedcommodities does little to empower women and trade policies should not solely focus on export activities in their narratives.

Intergenerational link in women employment: evidence from Indonesia

I investigate the existence of an intergenerational link between women’s labor supply decisions in Indonesia using rich large-scale longitudinal data known as the Indonesia Family Life Survey (IFLS). This study contributes to limited empirical evidence on the intergenerational link in female attachment in the labor market in the context of a developing country. In addition to cross-sectional correlation, I employ a permanent component approach (Chadwick and Solon, 2002; Galassi et al., 2019) to estimate the intergenerational correlation between the mother’s lifetime employment and the daughter’s lifetime employment. I find that a mother’s employment affects her daughter’s employment decision in the future. From the baseline specification result, this intergenerational link coefficient is equivalent to more than two additional years of education. This suggests meaningfulness and the importance of intergenerational belief transfer in shaping the next generation of the female labor force. Finally,I provide evidence that the role-model effects and occupation-specific human capital transfer are the main plausible mechanism behind the intergenerational correlation. The study highlights the challenges of any efforts to improve the female labor supply given a slowly changing social norms society.

Unlucky and scarred: long-term consequences of labor market entry condition in Indonesia

This paper provides empirical evidence of the long-term consequences of labor market entrance conditions using Indonesian data. I exploit the fact that since the Asian Financial Crisis hit in 1997/1998, Indonesia suffered a spike in the unemployment rate that prolonged for almost a decade. I collect and harmonize a long series of Indonesian labor market surveys (SAKERNAS) spanning over 30 years to construct a pseudo-panel cohort of new labor market entrants from 1990 to 2019. Following Kahn (2010) and Oreopoulos et al. (2012), I exploit exogenous temporal variation of the unemployment rate at the national level and province level to test the existence of scarring effects. To deal with endogenous migration issues for the province-level specification, I constructed a migration-weighted unemployment rate (Schwandt and von Wachter, 2019) based on historical inter-province migration patterns from the Population Census. I find evidence of a scarring effect where a 1 percentage point increase in the unemployment rate at the year of labor market entrance causes about 15% loss in probability to be employed full-time and about 26% potential monthly income loss. The negative effects of the unemployment rate in the initial year on employment and income linger up to 11 years after entering the labor market. I find women and men share similar burdens in terms of negative employment effects, but larger negative income effects for women.

Land of the (Un)Fair Go? Peer gender norms and gender gaps in the Australian labour market

with Josiah Hickson | TTPI Working Paper Series

Australian attitudes towards women remain more conservative than in many other OECD countries. We examine the effect of these norms on female labour outcomes and intrahousehold dynamics using a peer effects model and nearly two decades of longitudinal household survey data. Our results indicate that conservative gender norms are costly to individual women and are an important determinant of gender inequality, resulting for women in lower lifetime rates of labour force participation and suppressed lifetime earnings trajectories. Estimated effects are large in magnitude: shifting norms to be one standard deviation more egalitarian would eliminate three-quarters of the gender gap in employment and around two-thirds of the gender pay gap. More egalitarian peer norms are also associated with increased household incomes, a more equitable division of unpaid domestic work, and greater overall life satisfaction.

Mother education and children’s well-being: evidence from four Pacific countries

with Dyah Pritadrajati

We study the relationship between maternal education and children’s well-being in four Pacific countries. We exploit the richness of MICS dataset to investigate such relationships and the potential mechanism behind. We find that mother’s year of schooling positively correlates with likelihood of overweight and Early Childhood Development Indicator score while negatively correlates with the probability of having stunted children. We also find that the results vary across countries where the results are mainly driven by Kiribati and Samoa samples. Our further investigation reveals that years of schooling indirectly affect these outcomes via better caring practice and enrolment to early child- hood education (ECE). We collect evidence that suggest potentially direct effects of year of schooling on stunting outcome. Policy wise, this paper shown that improving access to education to the women population per se has limited effect in the context of a developing country. Our results advocate for disseminating specific maternal and parenting knowledge and improving ECE access as an effective policy to improve child well-being in the Pacific area.

ADBI Working paper version

Why do young females delay labor market participation in rural Indonesia? A cohort analysis using maximum entropy approach

Rural females born in the late 80s and younger in Indonesia were less likely to partici pate in the labor market relative to their older counterparts who were born three decades earlier. The opposite trend however emerges for urban women. The interaction between two opposing trends is arguably the main culprit to the decades-long stagnation in female labor force participation in Indonesia. I construct a synthetic panel of cohorts using 30 years-long of repeated cross-sectional household survey data, namely SAKERNAS and SUSENAS. Focusing on the multi-generation of rural women, I employ the age-period-cohort model (Browning et al., 2012) to shed light on key factors that drive the negative cohort effect of labor force participation. The results suggest at least two important insights. Firstly, albeit more young women opt out the unpaid family job, I do not find evidence that improved education drives women out of the labor market as indicated by previous findings (Schaner and Das, 2016). Secondly, the negative cohort effect is largely driven by the married women with kids group. This suggests that childbearing cost isan important issue over generations in rural areas, despite a significant drop in fertility, as found in other countries (Klasen et al., 2020). My results challenge the adequacy of the current campaign on lifting female labor force participation by only focusing on improving access to education in rural areas and providing access to childcare in urban areas.

Quality-quantity trade-off: an evidence from Indonesia.

This paper provides evidence on quantity and quality trade-off (Becker and Lewis, 1973) existence in Indonesia using variation of province-level total fertility rate by cohorts as an instrument variable. I use the fact that Indonesia’s nationwide family planning policy, known as Keluarga Berencana, became ineffective due to weaker implementation, after an exogenous political shock in 1998 that dramatically turn Indonesia into more decentralized regime. This event unexpectedly have plateaued total fertility rate in Indonesia post-1998. Using the fourth wave Indonesia Family Life Survey data, I investigate the trade-off on children aged less than 15 years old who are born between 1992 and 2006. As contribution to previous efforts (Millimet and Wang, 2011), this paper also investigate such trade-off separately for urban and rural settings. I find trade-off exists only for height-for-age in rural areas. It does exists for year of schooling and has greater impact for urban samples. These results support the argument of potential role of public goods provision in reducing the cost of trade-off (Angrist et al., 2010). My findings robust over alternative instrument variable constructions, birth order effect, and falsification test. The results support rising concern on the need of revitalization of family planning policy in Indonesia.

Book Chapter

Structural Transformation and the Release of Labor from Agriculture.

with Asep Suryahadi and Veto Indrio

Abstract: Structural transformation in Indonesia has been characterized by faster output than employment shifts from agriculture to industry and services. As a result, the ratio of output contribution to employment contribution in the agriculture sector has fallen relative to the other sectors. The finding from the long-term employment transformation matrix (1997−2014) in this study confirms that people who started working in the rural agriculture sector have a lower probability of moving to other sectors, especially to urban-located sectors. Furthermore, despite the continuing new entry of younger cohorts into the labor market, this dynamic in employment transformation has not changed much during the last 2 decades. This phenomenon may have a role in explaining the stagnation in poverty reduction and the increase in inequality in recent years.