Maddison‐style estimates of the evolution of the world economy: A new 2023 update
Level 5 - mechanism / opinion, no new human data
Level 5 by design analogy (methodological survey and historical database revision framework, not clinical CEBM).
OpenAlex W4393946461 · doi:10.1111/joes.12618
What was done
The authors surveyed the historical national accounting literature and analyzed how relative income benchmarks affect long-run economic estimates in the Maddison Project Database (MPD). They evaluated the biases of three benchmark approaches—the 1990 benchmark, the 2011 benchmark, and the Penn World Tables (PWT) multiple-benchmark method—specifically for pre-1940 estimates, and developed a selection framework to pair time series with optimal benchmark anchors.
What was found
The abstract provides no empirical economic figures. Methodologically, the authors determined that the MPD should maintain the 1990 benchmark as its primary anchor while implementing two structural updates: integrating the 2011 benchmark for the post-1990 period and incorporating a 1909 historical benchmark for US/UK comparisons to fine-tune pre-1940 estimates, moving the dataset toward the PWT multiple-benchmark framework.
Why it matters
This paper establishes the methodological foundation for the 2023 Maddison Project update, improving the consistency of long-run historical GDP comparisons across centuries and countries.
Limits
The abstract reports no quantitative error margins, sensitivity tests, or specific country-level data coverage. Long-run historical national accounts necessarily depend on surviving historical records, price index assumptions, and proxy estimates that carry inherent measurement uncertainty.
Cited by
- context Between 1945 and 2000, the United States accounted for one-third of global economic growth while comprising only 5% of the world's population.