Manufacturing Claims: Computer-Assisted Analysis of Asbestos Product Identification
77 Pages Posted: 8 Aug 2009
Date Written: August 3, 2009
Abstract
A combination of computer-assisted and manual analysis of depositions given by asbestos claimants approved for payment under the proposed prepack bankruptcy plan for Congoleum, a floor tile manufacturer, reveals the following: only about 8 percent of malignant claims and 3 percent of non-malignant claims provide evidence of potential exposure to asbestos fibers from a Congoleum product; although all of the approved claims contained a signed affidavit averring exposure to Congoleum’s asbestos-containing products, 64 percent of the approved claimants gave testimony that was inconsistent with the affidavit; and, of these, over 40 percent gave testimony that precluded exposure to asbestos from a Congoleum product. The results support the following: (1) fears that large numbers of asbestos claims filed with claims facilities are being approved for payment without individual evaluation are consistent with the results; (2) mass, signed affidavits using the standard language required under the bankruptcy plan are not indicative of substantive evidence of exposure; and (3) on the methodological front, computer-assisted content analysis of a large depository of depositions is a valuable research tool and a practical tool in the preliminary evaluation of massive volumes of claims.
Keywords: text mining, content analysis, asbestos
JEL Classification: K13, K41, C19, C89
Suggested Citation: Suggested Citation