The walking economy: Pedestrian counts as a real-time indicator of regional economic activity

Central business district (CBD) pedestrian counts can offer timely signals of regional economic conditions given the reporting lags in standard macro indicators. We test whether daily pedestrian counts in Melbourne’s CBD, Australia’s second largest city, are useful for nowcasting Victoria’s monthly retail sales turnover. Using counts from ten fixed footfall sensors from January 2010 to December 2024, we implement easy-to-replicate Mixed-Data Sampling (MIDAS) models providing early signals ahead of official releases. To assess how shifts in work and mobility have altered the information content of footfall, we compare predictive performance across three phases: pre-COVID pandemic (Jan 2013–Jan 2020), COVID (Mar 2020–Feb 2022), and post-COVID/inflationary (Mar 2022–Dec 2024). We find that CBD footfall provides informative, timely signals for regional retail activity and improves out-of-sample accuracy relative to benchmark models across the full sample. Gains are largest for discretionary categories (cafés and restaurants, take-away, clothing, personal accessory retailing) compared with non-discretionary spending (e.g., groceries). Results demonstrate that CBD footfall accurately reflects regional retail sales, maintaining reliability even during periods of economic disruption. Overall, our findings position pedestrian counts as a low-cost, replicable indicator for near-real-time monitoring, offering practical lead information for policymakers, planners, and businesses while informing debates about evolving urban usage patterns.