Toujours dans la lignée de précédents articles sur l’état des lieux de la sécurité informatique, et précisément du vol d’identité, il importe de se baser sur des études qui n’ont pas forcément intérêt à dire que tout va mal et qui, à certains égards, n’ont pas à profiter de l’insécurité ambiante.

C’est pourquoi la comparaison faite par Javelin Stategy and Research est intéressante, autour du billet « ID Theft/Fraud : Rising or Declining ? », mettant en perspective leur propre étude et la FTC’s Consumer Sentinel Study qui vient d’être publiée la semaine passée (« Consumer Fraud and Identity Theft Complaint Data : January - December 2007 » (pdf - 92 pages)).

On peut donc lire sur le blogue de Javelin les justifications sur les différences qui apparaissent :

« A few have asked how Javelin can show that ID fraud dropped across the nation when the FTC shows complaints rising. The answer is simpler than you think : people complain more as problems get worse.

The FTC measures complaints rising by 20%, Javelin shows crime incidents dropping by 5%, with average per-victim losses going up by 25%. Logic says that you’re more likely to file a complaint when your losses rise, and assuming the ratio of increase is a straight line, 20+5%=25% right on the button.

Regarding the FTC’s $2.1B figure, it reflects out of pocket costs, and again only for those who filed a police report. Javelin’s report shows out of pocket costs in the $5B range, which of course is much bigger than the FTC figure because the FTC report only reflects those that filed a criminal complaint, whereas the Javelin study is a representation of everyone in the US that was a victim of ID fraud.

Other differences : The FTC CS study accounts for crimes where the selling party was not who they claimed to be, or did not provide what was advertised. This “merchant fraud” (in Javelin parlance) is outside the scope of Javelin’s identity study. »