The Genuine Progress Indicator (GPI) is a multi-dimensional composite indicator that estimates the quantity and distribution of net benefits of the economic system on supporting social and environmental systems. It has been developed and modified for nearly three decades utilizing mixed methods from macroeconomics, environmental economics, natural resource economics, and other heterodox approaches to environmental, social, and economic accounting. Although originally estimated and contrasted with Gross Domestic Product (GDP) at national scales, interest in state-level adoption has developed in the US to inform and guide policy. As a community of practice has developed, questions have arisen about the quality and legitimacy of the GPI, including concerns about normative assumptions, choice of variables and monetary weights, inclusion of stakeholders, extent of data collection, and statistical rigor. To investigate the quality and design of the GPI, we apply a composite indicator analysis developed by the Organization for Economic Co-operation and Development (OECD) to fifty US state GPI estimates using a consistent method. We focus on a multi-variate analysis of the structure of the composite, sensitivity to weighting and aggregation assumptions, and the statistical relationship with other well-being indicators, including the Gallup Well-Being Indicator, UN Human Development Index, life expectancy, and Ecological Footprint.
Results/Conclusions
Analysis reveals the size and variability of the GPI is dominated by a small number of components. Ten of the twenty-five components contribute to 90% of average GPI, while the bottom twelve contribute less than 5%. A few components contribute to the majority of the variability, for instance with the cost of non-renewable resource depletion responsible for 71% of the variance between state GPIs. Additionally, nine component ranks correlate with state GPI rank at an absolute value of 0.7 or higher. An outlier analysis also revealed six components with low outliers, and nearly all with high outliers. A sensitivity analysis on wage rates, benefits of household labor, and costs of inequality, climate change, non-renewable depletion, and motor vehicle crashes revealed a number of unintended policy outcomes. Finally, the comparison of GPI with other state-level indicators found no correlation with the Gallup Well-Being Indicator and Ecological Footprint, weak correlation with life expectancy, and significant correlation with the Human Development Index. The study suggests steps towards shared GPI governance amongst practioners and researchers, consideration of data parsimony and potential double-counting, and data selection criteria to help fill gaps, prioritize needs, and better articulate the purpose and meaning of GPI.