Urban sustainability – a segmentation study of Greater Brisbane, Australia

Abstract

Setting universal goals for sustainability is problematic and may hinder the adoption of sustainable pathways as different sectors of society often have differing opinions on not just what sustainability means for them, but also what is of priority to them. This paper tests a set of psychographic, behavioural, lifestyle and social identities to segment the public on sustainability. We evaluate general knowledge, apply social-choice tools to identify public priorities, and then apply segmentation to reveal broad strata of community profiles around these choices. We discuss our findings in the context of moving beyond knowledge on sustainability and general public choices, to more nuanced messaging and engagement that respects differences in sustainability orientations. We suggest that by focusing on what matters most for different segments of society, there is potential to design effective processes to engage with people and acquire better ownership of sustainability.

Keywords:

• sustainability
• social choice
• segmentation
• lifestyles
• social identity

Acknowledgements

Steve Hatfield-Dodds is acknowledged for project leadership and support in the formative stages of this work. Cindy Gallois for her critical review is acknowledged. Geri Craven's organisation and support in the survey phase is appreciated. Thanks are also due to Tom Keenan and two other independent reviewers.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Component factor analysis is concerned with patterning all the variation in a set of variables, whether common or unique.

2. The R² indicates how well an indicator is explained by the latent cluster variable. For ordinal, continuous, and counts, these are standard R²measures. For nominal variables, these are Goodman-Kruskal tau-b coefficients, representing a weighted average of separate R² measures for each category treated as a separate dichotomous response variable.

3. The BIC statistic adjusts the L² to control for the sample size and degrees of freedom, allowing us to compare models directly.

Additional information

Funding

The authors would like to acknowledge the Commonwealth Scientific and Industrial Research Organisation (CSIRO) for financing this research.