Aggregation, heterogeneity & volatility
Decomposing aggregate volatility into weighted firm-level variances, covariance terms and the effects of changing composition.
Research programme
Research
I study how fluctuations across heterogeneous firms combine into macroeconomic volatility, and how normalization and geographic units change economic measures.
Core question
When economic data are weighted, grouped or normalized, which relationships follow from the transformation itself, and which require additional empirical assumptions?
Specialization & location quotients
Measurement of specialization under finite samples, probabilistic structure and size-related distortions.
Geographic correlations & spatial support
How normalization, geography and support affect correlations and interpretation in economic data.
Selected outputs
Published peer-reviewed
Measuring Size Distortions of Location Quotients
International Economics, 2021. Explains how observation size distorts location quotients and motivates probabilistic specialization measurement.
Read the DOI recordDoctoral thesis
Essays in the Economics of Aggregation
Doctoral research developing mathematical and computational accounts of how heterogeneous micro-level fluctuations generate aggregate volatility.
Current manuscript
Concentration Is Not Scaling
The current manuscript separates economic concentration from volatility scaling and tests an exact decomposition of variance and covariance contributions in code.
Working paper
The Overlooked Insights from Correlation Structures in Economic Geography
A 2021 working paper examining how correlation structures and spatial normalization affect economic co-location measures. Newer extensions remain under review.
Applied and computational work
Research in progress
Poverty measurement
Survey and geographic measurement work with explicit uncertainty, interpretation and external-validity limits.
See the engineering evidenceSoftware / research artifact
Micro-to-macro reconstruction of aggregate volatility
I derived aggregate-volatility contributions from weighted firm variances and signed covariances, then implemented Python tests of the exact identities. The computations distinguish accounting results from assumptions needed for a scaling law.
See the engineering evidenceMethods and academic record
I use formal derivations, simulations and numerical tests to check mathematical identities; survey and spatial projects add estimation, validation and explicit data-support limits.
The full research portal collects the thesis and related materials. Explore the full research programme