Models used to develop the project
The Happiness Project is a student-led, staff-supervised collaboration between Bond's journalism students, who collected and drove analysis of the data, then wrote stories contextualising the results, and interactive media and design students, who designed an interactive map to let readers explore the results.
The methodology for the Happiness Project was developed after comparing international quality of life measures such as the Organisation for Economic Co-operation and Development's Better Life Index or the Social Progress Imperative's Social Progress Index.
Australia scores well on these international measures, however, the factors that contribute to this — for example, health, education and job opportunities — vary across the country.
How this project is different
While we considered existing international models of measuring quality of life, adjustments were made based on the availability of (local-level) data and consideration of the Australian context.
It's also important to note that as a data journalism project, which aims to inspire debate and understanding (as well as student learning), the Happiness Project's motives and frameworks are somewhat different to that of the international indexes.
The project provides a lens through which to assess a range of quality of life factors across Australia, but it is not the only way to measure such things.
Quality of life categories
We used nine categories to develop the project, and each of the categories relied on between two and four data sets.
Of course, a concern with any project like this is the potential for selection bias in the factors that are chosen to assess the variable of interest. For this reason, our project shares many of the categories the OECD uses to assess quality of life, with the exceptions of Environment (water quality is less varied across Australia than it is internationally, and air quality data wasn't available at a regional level), Civic Engagement (because we're not measuring a range of different nations, there's less variance in these results, particularly because Australia has compulsory voting) and Life Satisfaction (no data available).
We've been as careful and considered as we can be in selecting the factors that make up our project, but are open to including other measures in potential future iterations.
For the Housing affordability category, we considered: households requiring extra bedrooms, mortgage and rental payments in relation to wages, and low-income housing stress.
For the Education category, we considered: level of educational participation and achievement (including trade qualifications), young adults in schooling and ‘learning or earning’ rates.
For the Community category, we considered: rates of volunteering and cultural tolerance.
For the Jobs category, we considered: personal income, long-term and current unemployment rates and labour force participation.
For the Wealth category, we considered: household income, access to emergency funding and socio-economic disadvantage.
For the Accessibility category, we considered: remoteness, internet access and residents with transport barriers.
For the Work-life balance category, we considered: rates of overwork and unpaid domestic labour.
For the Safety category, we considered: available data around annual homicide and assault rates and modelled estimates of personal safety.
Safety was actually the most difficult category in the Australian context, because we had to collect the data from each state or territory's police force.
In all instances we used the latest available annual crime rates (in some cases calendar years, sometimes financial years or other 12-month reporting periods).
Where LGA-level regional data was not available, larger regional breakdowns were used.
It is also important to note that states define and record crimes like homicide and assault differently, so comparisons in this area are indicative only.
For the Health category, we considered: barriers to healthcare, poor self-assessed health, the median age of death, premature mortality and major health risk factors (high alcohol consumption, smoking, obesity and lack of exercise).
The overall Quality of life score was calculated as an average of the nine category scores for the region.
We collated raw data from: ABS data (Census 2016, SEIFA Indexes); data from the Public Health Information Development Unit's Social Health Atlas series; Australian Health Policy Collaboration data in Australia's Health Tracker Atlas (2017); and public access data from each of Australia's state police forces.
However, raw data is difficult to understand without context, and we needed a way to compare regions across the country, so we developed a composite measure for comparative purposes.
Once the Happiness Project data was finalised, our Investigative Journalism students — who helped design the conceptual frameworks for the project, and collated data — contextualised the findings with a series of interview-driven stories.
Our Interactive Media and Design students have created the visualisation, which allows readers to interact with the data, find their LGA's results and compare regions. The visualisation also provides national heat maps, showing the geography of quality of life, and of each of the categories as a clear visual narrative.
We think journalism brings about debate and change — and we hope the Happiness Project does both.