본문 바로가기
  • Home

Positive Externalities of Local Public Events and Digital Information Diffusion: A DID-Mediation Analysis of the Jeonju Gamaek Festival

  • Industry Promotion Research
  • Abbr : IPR
  • 2026, 11(4), pp.517~528
  • DOI : 10.21186/IPR.2026.11.4.517
  • Publisher : Industrial Promotion Institute
  • Research Area : Interdisciplinary Studies > Interdisciplinary Research
  • Received : September 30, 2026
  • Accepted : October 23, 2026
  • Published : October 31, 2026

Seong-Il Lee 1

1신한대학교

Accredited

ABSTRACT

This study investigates the positive externalities and mediating mechanisms of digital information diffusion a ssociated with the Jeonju Gamaek Festival by combining credit card transaction big data with a quasi-experimental desi gn. Combining an eight-year daily panel dataset from 2019 to 2026 (N = 992) with 2025 micro-spatial commercial big data, we employ a Two-Way Fixed Effects Difference-in-Differences (TWFE-DID) model, concentric buffer analysis, an d an exposure DID mediation model. The empirical findings are as follows. First, the festival significantly increased n on-resident food and beverage expenditures in Jeonju by +4.61% (p = 0.0098); however, in contrast to the substantial i nflux on Thursday and Friday, a potential ceiling effect emerged on Saturday due to commercial capacity constraints. Second, micro-spatial analysis identified the commercial district within a 500m–1km radius (Ring 2) as the primary be neficiary relative to the outer baseline (Ring 3: 1–1.5km), revealing a trade-off between daytime restaurant spillover (+ 25.33%) and nighttime pub cannibalization (-9.87%). Third, the festival triggered an approximately 16.45-fold increase i n online search intensity (exp(2.8003), +1,545%), and portal search trends indirectly mediated 72.78% of the total cons umption impact, supporting the digital information funnel mechanism. The statistical attenuation of the direct effect refl ects the absorption of traditional promotional stimuli into mobile search inquiries rather than the absence of offline pro motion. The robustness of the estimates was confirmed through placebo tests, wild cluster bootstrap procedures, and th e detection of post-festival residual consumption (+0.60%), suggesting the potential extension of tourist stays. This stud y addresses aggregation bias in festival impact evaluations and provides baseline evidence for designing spatiotemporal coexistence policies and targeted digital marketing strategies.

Citation status

* References for papers published after 2025 are currently being built.