Journal of Economic and Social Dynamics

Abstract

This paper uses transaction-level data from Registers Iceland covering 225,000 residential sales across 60 Icelandic municipalities (58 retained in the merged panel) from 2006 to 2024, to examine the association between immigration and house prices. The central finding is a null result: in first difference panel regressions with year fixed effects, neither the capital region nor outside capital municipalities show a statistically significant association between changes in the immigrant share and changes in the price per square metre. A secondary finding is methodological: a positive and significant association found in aggregate regional price data disappears once year fixed effects absorb common time trends, illustrating how omitted variable bias from common time trends can produce spurious associations in aggregate regional data. One likely explanation is thin transaction markets outside the capital, where 63% of municipality-year cells contain fewer than ten sales, making annual mean prices too noisy to reveal a reliable price signal.