Living with(in) Socialist Heritage Space: Everyday Modifications and Social Dynamics in China’s Soviet-Style Housing

China’s socialist workers’ housing occupies an uncertain position between everyday residential use and potential heritage conservation. This study examines long-term modifications to Soviet-style workers’ housing in Harbin’s Village of Powers (VoP), where original collective facilities have been adapted to changing household and economic needs. Drawing on historical drawings, available evidence of early occupancy, architectural surveys of 17 buildings with comparable spatial-area data, interviews, and on-site observations, this study compares reconstructed historical usage conditions with those documented in 2020–2021. It examines spatial redistribution, modification practices, and their social consequences. Building-level comparisons show that the proportion of private space increased by an average of 6.69 percentage points in nine dormitories, compared with 0.21 percentage points in eight family apartment buildings (two-sided exact permutation test, p = 0.00045). However, the difference in total usable floor-area growth was not statistically significant. The documented cases identify improving living conditions and pursuing economic opportunities as two recurring motivations for modification, while the social consequences varied. Some interventions restricted access to collective facilities and were associated with tensions over shared-space use, whereas certain commercial spaces accommodated social interaction under specific conditions. Situated within the transformation of China’s Danwei system, these findings reveal how everyday modifications reshape the relationship between private and collective space. They provide socio-spatial evidence for heritage-sensitive regeneration that considers historical spatial characteristics, contemporary living needs, and the interests of different users. Although the quantitative analysis covers 17 buildings with comparable historical and 2020–2021 spatial-area data, the study is confined to a single purposively selected case.