Update to Limits to Growth: Comparing the World3 Model With Empirical Data
MetadataShow full item record
CitationBranderhorst, Gaya. 2020. Update to Limits to Growth: Comparing the World3 Model With Empirical Data. Master's thesis, Harvard Extension School.
AbstractFor more than three decades, the authors of the bestseller Limits to Growth (LtG) warned that a pursuit of continuous growth would result in a sharp decline (i.e., collapse) of global human welfare levels within the 21st century. The authors published three LtG books between 1972 and 2004, in each of which they studied interactions between global variables of a model called World3. With World3, which was updated for each book, the authors generated different scenarios for global developments by varying assumptions about technological development, amounts of natural resources, and societal priorities. Their "business as usual" (BAU) scenario contained no assumptions on top of historical averages. BAU showed a halt in the increase of global welfare levels around 2020, and a collapse starting around 2030. Not all scenarios led to collapse; the LtG team identified a set of assumptions that produced a “stabilized world” (SW) scenario in which decline was avoided and welfare remained high. But independent empirical data comparisons since then, most recently from 2014, indicated that the world was still following BAU.
The objective of my research was to examine whether this still was the case based on data available in 2019, and whether there was opportunity left for society to align with the SW scenario. My research objectives were to i) conduct a data comparison between the current global state and scenarios made with the latest version of World3, and ii) determine how close each scenario compared with observed data. I hypothesized that BAU would align more with the data than other scenarios, and do so closely for most or all variables. I collected data for real-world indicators of the World3 variables population, fertility, mortality, pollution, industrial output, food, services, non-renewable natural resources, human welfare, and ecological footprint. This data came from academia, (non-)government agencies, United Nations entities, and the World Bank. I used four LtG scenarios with underlying assumptions that span a range of technological, social, and resource conditions: BAU, SW, “comprehensive technology” (CT), and “business as usual 2” (BAU2). CT represents the technologist’s belief in humanity’s ability to innovate out of environmental constraints. BAU2 assumes double the resources as in BAU and depicts a pollution collapse, including from CO2 (i.e., climate change). Both scenarios indicate a halt in growth within the next few decades, but BAU2 shows a sharp decline while CT shows a moderate one. To measure alignment of empirical data with scenarios I used: value difference, rate of change difference, and normalized root mean square error.
My research revealed an overall close alignment of empirical data with each of the four scenarios, which is a testament to the accuracy of World3. SW was followed least closely, then BAU, and both BAU2 and CT aligned closest. My hypothesis was rejected, but this could change with an update of the comparison because for several variables the scenarios only diverge significantly after 2020. This is especially so for BAU2 and CT, which is why it was not possible to differentiate between them. It’s thus unclear whether a future decline can be expected to be moderate or sharp, but both scenarios indicate society will run into limits in the medium term. The close alignment of scenarios and their lack of divergence means that the identification of BAU2 and CT as closest fits could be nullified or even reversed with a few years’ extra data points. It also means that it is not too late to change course. Although SW tracks least closely, a deliberate trajectory change is still possible. That window of opportunity is closing fast.
Citable link to this pagehttps://nrs.harvard.edu/URN-3:HUL.INSTREPOS:37364868