Braer Garland Sovereign Seal

BRAER GARLAND RESEARCH INSTITUTE

← Return to Core Repository Index

Decadal Shifts in U.S. Demographics

Longitudinal Harmonisation and Tabular Analysis of Public Use Microdata (2004–2024)

1. Data & Methodology

This analysis utilizes 1-year American Community Survey (ACS) Public Use Microdata Sample (PUMS) files retrieved from the IPUMS USA database. To guarantee longitudinal comparability across a 20-year timeline, a master orchestration layer was developed in Python (v.3.11.3) linking to a MySQL 9.5 relational backend. This layer excludes group-quarter definitions from the 2014 and 2024 samples to perfectly mirror the 2004 survey universe boundaries.

2. Survey Universe Harmonisation

The total analytical footprint spans a massive microdata substrate. The exclusion matrix systematically dropped 145,066 and 183,206 group-quarter records in 2014 and 2024 respectively to enforce strict structural consistency across all study years.

Sample Metric Status N (2004 Node) N (2014 Node) N (2024 Node)
Total Households Sampled 483,316 1,363,306 1,531,614
Excluded Group Quarters 0 145,066 183,206
Final Comparable Households (N) 483,316 1,218,240 1,348,408

3. The Inversion of Household Headship

The most critical empirical pattern observed is a constant, highly significant decline in the proportion of male householders. Over the twenty-year study period, this structural displacement culminated in an unprecedented inversion, shifting from a clear male majority to a female majority head-of-household society.

Head-of-Household Sex Proportion (2004) Proportion (2014) Proportion (2024) Significance (2024 vs 2004)
Male Householders 55.8% 50.3% 49.0% p < 0.001 (***)
Female Householders 44.2% 49.7% 51.0% p < 0.001 (***)