Work in Progress
Tightening next door: spatial spillovers of monetary policy on regional unemployment in Chile
Luis Pino-Sandoval
Resumen
This letter asks whether the regional effects of monetary policy depend on the economic structure of neighbouring regions. I combine quarterly unemployment rates for Chile’s 16 regions (2012Q1–2026Q2) with market-based monetary policy surprises, measured as the change in the one-year swap rate around 140 scheduled policy meetings, and estimate spatial local projections of the SLX type with region-by-season and time fixed effects. Exposure is the pre-sample manufacturing share of regional output; spillovers are captured by the same share averaged over contiguous regions. A region’s own exposure does not change its unemployment response. The exposure of its neighbours does: after a 25 basis point surprise, unemployment rises by 0.8 percentage points more six quarters later in a region whose neighbours are one standard deviation more manufacturing-intensive. The combined effect of higher own and neighbours’ exposure is 0.56 percentage points after four quarters (wild bootstrapp = 0.06) and 0.64 after six. For a typical surprise of 4.5 basis points this is about 0.1 percentage points. The pattern holds with inverse-distance weights, without the pandemic years, without the Santiago region and with a broader interest-sensitive exposure, but not with 90-day swap surprises. Studies that rely on a region’s own structure may miss part of how monetary policy spreads across space.