The ID Analytics findings are consistent with other studies when it comes to highlighting risky areas, Coggeshall said. However, unlike previous research based largely on consumer victim reports, the ID Analytics study is based on an analysis of attempted fraud using compromised data.

That distinction is important because an overwhelming majority of identity fraud results from the creation of "synthetic" identities rather than "true name" identity theft, an ID Analytics white paper noted. As a result, there is often no consumer victim to report the crime.

The identity-level data elements used for the analysis came from ID Analytics' customers in the financial services, retail, health care, telecommunications and other industries. The data included names, addresses and Social Security numbers culled from account applications and other transactions monitored by ID Analytics for fraudulent activity.

Fraud rates were calculated by dividing the total number of reported identity fraud incidents by the number of applications to credit grantors during 2005 through June 2006, Coggeshall said. The approach allowed for a better comparison of fraud rates across areas with different population densities, he said.

"We do see stability across time," Coggeshall said.