Correcting Measurement Bias in Online Auction Research: A Relative Value Benchmark for Strategic Underpricing

Empirical research in e-commerce and online auction dynamics frequently evaluates 'strategic underpricing' using category-level quantile cutoffs (e.g., opening bids in the bottom 10th or 25th percentiles). In this paper, I demonstrate that threshold-based quantile indicators introduce severe measurement error by confounding low-value products with strategically underpriced products. To resolve this measurement bias, I introduce the Relative Underpricing Index (RUI), a continuous metric benchmarked against item-level expected market reference values (Vc). Applying heteroskedasticity-robust (HC3) Ordinary Least Squares (OLS) regressions and structural path mediation models to 628 completed eBay auctions across standard product categories, I show that quantile thresholds yield misclassified, spurious estimates. Under the benchmarked RUI metric, strategic underpricing exhibits a strong, statistically significant negative association with net revenue realization (β = -117.42, p < 0.001). Furthermore, while relative underpricing lowers entry barriers and increases unique bidder participation (α = +5.71, p < 0.001), a formal Sobel mediation test confirms that the resulting indirect revenue boost (+51.18, p < 0.0001) fails to recover the direct revenue loss incurred from opening bid discounts. This study provides a corrected econometric framework for marketplace researchers and warns sellers against relying on naive low opening bids.