What is it about?
When we study regional incomes, results can change just because we switch from municipalities to districts or counties. Using Hungary’s 2019 household income data, we show how mapping and regression results shift with aggregation, and which local factors (jobs, entrepreneurs, education) matter most at finer scales.
Featured Image
Why is it important?
Policy decisions often rely on spatial models, yet they may be distorted by the “modifiable areal unit problem”. By comparing OLS and spatial econometric models (SEM, SAR, SDM) across municipal, district, and county levels, we show that higher aggregation masks relationships and weakens spatial autocorrelation signals especially at county level. The takeaway is practical: use the smallest feasible units to avoid misleading evidence.
Perspectives
What struck me is how easily “the story” changes with the map’s resolution. At county level, many drivers look insignificant, not because they don’t matter, but because averaging smooths away the variation people actually experience. This work reinforced my preference for high-resolution evidence in regional policy debates and for being explicit about the trade-off between interpretability and information loss.
Tibor Bareith
ELTE Centre for Economic and Regional Studies Hungary
Read the Original
This page is a summary of: The importance of aggregation in regional household income estimates: A case study from Hungary, 2019, Regional Statistics, January 2023, Hungarian Central Statistical Office,
DOI: 10.15196/rs130603.
You can read the full text:
Contributors
The following have contributed to this page
