Many countries are experiencing health workforce shortages [1, 2]. Medical workforce issues have become increasingly prominent on the national policy agenda in Australia, and have led to significant increases in medical training places[1, 3]. Shortages are most pronounced in non-metropolitan areas. To address this, a variety of rural health workforce policies have been implemented in the last decade[4, 5]. Career choice by medical graduates is also of increasing interest to policy makers. In the short term, specialties perceived (for various reasons) as being less attractive are experiencing difficulties in recruiting sufficient entrants to their training programs. In the medium term, vastly increased numbers of medical graduates are likely to create different problems for vocational training programs.
Research has a key role to play in the development of effective policy responses to challenges such as these. There is increasing recognition of the need for and value of close connections between health services research and health policy, through making policy more evidence-based, and making research more policy-relevant[8, 9]. This paper describes a prospective longitudinal study of Australian doctors' workforce participation decisions and their determinants, as the basis for generating evidence of direct relevance in developing more effective medical workforce policies.
Decisions made by doctors about where, when, and how much to work have profound effects on health systems, including equity, capacity, and quality of care. Doctors' labour supply decisions are influenced by a complex mix of factors, including their own preferences about work, leisure, family and lifestyle; economic and non-economic incentives embedded in health system financing and organisation; the culture of medical practice; and longer term trends in demand for health care, demographic change and the composition of the medical workforce.
Existing research has described some of the key trends in Australian medical labour supply, many of which are similar to those in other developed countries. There are notable changes in the gender composition of the workforce, with 34% of Australian doctors in 2006 being female, compared to 18% in 1981[10, 11]. Females have comprised approximately 50% of Australian medical graduate cohorts since the late 1990s, and thus the proportion of females in the workforce is expected to continue to rise. This changing gender balance impacts significantly on workforce supply levels, because on average female doctors work fewer hours than male doctors and take more time out from the workforce[10, 12]. Workforce participation also varies with age, and the age profile of the Australian medical workforce is changing as well. More than half of the workforce in 2001 were of the 'baby boomer' generation, and therefore are now approaching retirement age . This profile is influenced by historical trends in medical workforce supply policies, with a 'demographic hump' created by the last boom phase in supply growth in the 1970s progressively moving into the older age categories. Graduate numbers in Australia have been cyclic over the decades, with static output levels during the 1980s and 1990s, and the current growth phase commencing around 2000 . This will see graduate numbers increase by 81% between 2005 and 2012.
There are a number of important changes associated with these trends, such as decreases in average working hours and differences between cohorts in preferences about workforce participation. Average working hours for female doctors have decreased by 4.6 hours per week (11%) since 1994, while those for male doctors have decreased 9.3 hours per week (17%)[10, 11]. This partly reflects cohort effects, with younger doctors of both genders less inclined than their older colleagues to work the long hours traditionally associated with medical practice[13, 15]. Another key cohort difference is increased mobility, with younger generations having an increased propensity to move not only between jobs, but also between locations (both within and between nations), and across traditional occupational boundaries[15, 16].
While there is a considerable body of research and data describing trends in Australian medical labour supply, there is little research, nationally or internationally, examining the determinants of labour supply decisions for the medical workforce. Although many of these trends are influenced by differences in the preferences of younger cohorts of doctors, they are also influenced by the institutional structures of the health care system, and can therefore potentially be influenced by government policies. This is especially the case for workforce participation, hours worked, workforce distribution, and specialty choice. Although global health workforce shortages suggest that more doctors need to be trained, this is expensive and has very long time lags. There are many other potential policy responses that could increase the productivity of the existing workforce, change the distribution of doctors, or shift tasks to other health workers. Longitudinal data and appropriate statistical analysis are essential for exploring the links between labour supply and its determinants, and large longitudinal data sets covering the entire medical workforce (not just recent graduates) that can be used for this purpose are scarce. The dynamic nature of medical labour markets (with factors such as demand for services, price of services, and government policies playing a role) makes the use of longitudinal data especially important.
We have commenced a longitudinal survey of medical labour market dynamics in Australia, focusing on several key outcome measures including the number of hours doctors work, and decisions to change job, move location, enter a particular medical specialty, or leave the medical workforce. The Medicine in Australia: Balancing Employment and Life (MABEL) study is designed to describe and understand key determinants of these outcomes, including working conditions, job satisfaction, family circumstances, and incentives.
As noted before, the MABEL study was designed to provide research evidence of direct relevance to identified policy issues. The study is advised by a national Policy Reference Group which is a key mechanism to support research-policy linkage. This group comprises representatives from Commonwealth and State government health departments other key organisations such as Rural Workforce Agencies and the Australian Institute of Health and Welfare. The Group helps to ensure that the study is informed by current policy issues and priorities, and to assist in the translation of findings into the policy context.
Implementing a national longitudinal medical workforce study such as MABEL is not without significant challenges. The aim of this paper is to describe the methods and the baseline cohort of the MABEL study. We provide a detailed examination of representativeness, through analysis of response rate and response bias.