Uncorking Expert Reviews with Social Media: A Case Study Served with Wine

Abstract

The growth of social media outlets in which individuals post opinions on publicly consumed goods provides an interesting and relatively unexplored area for examination of the role of crowd sourcing amateur opinions in areas traditionally relegated to experts. In this paper we use wine as an illustrative example to investigate the interaction between social media and expert reviews in the market for high end consumer goods. In particular, we exploit a novel data set constructed from the social media website CellarTracker, which is composed of the averaged individual reviews for 355 distinct wines on a quarterly basis from 2004 through 2017, and pair this with a similarly dimensioned panel of average auction prices for these wines as well as the reviews from three leading experts. We develop a signal extraction model to motivate the interaction between amateurs and experts in revealing a measure of the quality of the wine. The model is then used to motivate the adaptation of an empirical panel structural VAR approach based on Pedroni (2013) by embedding the expert reviews as an event analysis within the panel VAR, which is used to decompose information into components that signal the quality of the liquid in the bottle versus other aspects of the wine that are valued by the market. The approach also allows us to decompose the influence of the expert reviews into components associated with what we define as the quality of the wine versus the pure reputation effect of the expert. The results on expert reviews are consistent with the idea that experts can substantially impact prices through channels other than their signals of quality.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 3 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.