Breach disclosure laws have 'no effect' on identity theft
Red tape failing to stanch losses
Widespread information security breach laws in the US have failed to do much to reduce identity theft. The finding, by researchers at Carnegie Mellon University, comes as calls are growing in Europe to enact laws that would oblige organisations to notify customers in cases where their personal details become exposed.
The researchers based their findings on a state-by-state analysis of data from the US Federal Trade Commission (FTC). They looked at identity theft complaints made to the FTC in the four years between 2002 and 2006, hunting for changes after states introduced data breach disclosure laws. Over recent years, 43 states have followed the lead of California and introduced breach disclosure regulations.
The FTC has invited people to submit complaints about identity theft through its website since 1999. Only a minority of victims make such reports, which are used by law enforcement agencies to look for crime trends. In 2006, for example, 246,035 identity theft reports were submitted to the FTC out of an estimated total of 8.9 million incidents where fraudsters had secured loans, goods or lines of credit under false identities.
The Carnegie Mellon team used freedom of information requests to obtain a state-by-state breakdown on these reports.
The researchers found that factors such as changes in average income and population in a state, as well as overall levels of fraud had a much greater effect on fraud rates, as explained in an abstract to the paper (below):
We find no statistically significant effect that laws reduce identity theft, even after considering income, urbanisation, strictness of law and interstate commerce. If the probability of becoming a victim conditional on a data breach is very small, then the law’s maximum effectiveness is inherently limited. Quality of data and the possibility of reporting bias also make proper identification difficult. However, we appreciate that these laws may have other benefits such as reducing a victim’s average losses and improving a firm’s security and operational practices.
"There doesn't seem to be any evidence that the laws actually reduce identity theft," Sasha Romanosky, a PhD student at Carnegie Mellon and one of three authors of the study, told Computerworld.
Romanosky, who worked in computer security at eBay and Morgan Stanley before deciding to pursue his doctorate, acknowledged that the data sample the Carnegie Mellon analysed is incomplete and based on a self-selecting sample of people prepared to report that they'd become a victim of fraud. Nonetheless he suggested that the methodology of the study is strong enough, and its findings are clear enough, to raise doubts about the conventional wisdom that breach disclosure laws provide a useful means to alert customers that their data has been compromised. The idea is that, once notified, consumers will be more on their guard about possible incidents of fraud.
Disclosure laws are also designed to encourage firms to improve their security policies or else risk the threat that any information security breach would force a public disclosure that might tarnish their brand.
Although the researchers found little benefit in breach laws, at least in terms of reducing identity theft, the Carnegie Mellon team advocate the adoption of federal breach disclosure laws featuring standard notification requirements. Federal laws in the area make compliance easier.
The Carnegie Mellon team's research is due to be presented at a conference on the economics of information security to be held at Dartmouth College later this month. The paper, entitled Do Data Breach Disclosure Laws Reduce Identity Theft?, can be found here (PDF).
Economists and security watchers are split over whether or not corporate data breaches fuel identity theft. Some reckon dumpster diving and phishing scams are a more significant contributory factor to ID theft.
Javelin Strategy & Research, reckons that 30 per cent of known identity thefts in 2005 came as a result of corporate data breaches. Collectively losses from identity theft came to $56bn in 2005, it reckons.
However other research cited by the Carnegie team found that the chances of becoming a victim of identity theft as a result of a data breach is small, around only two per cent.
Separately a survey by UK-based net security firm ClearSwift found that a large majority (87 per cent) of UK-based IT decision-makers don't think the general public ought to be informed if a data breach happens. Over half (61 per cent) of the 400 senior techies surveyed also reckon that it's preferable not to let the police know of breaches either.
Half those quizzed (49 per cent) reckon that data breach notification legislation would push up IT spending by at least 5 per cent. ®