To determine the change in probability of accounting fraud, Lee used two prediction models that identify unusually high accruals, inflated sales, overstated inventory, and other conditions often associated with fraud.
The deterrent effect could have been noted by audit firms: their fees tended to fall by 5 percent after clients were exposed to state FCAs, Lee finds. When there’s a lower probability of fraud, there’s also reduced potential auditor liability.
Lee constructed her first sample using US public-equity holdings of state pension funds in jurisdictions that had enacted an FCA with a qui tam provision. Whistle-blowers in those states could obtain financial rewards by reporting financial fraud of any company, regardless of its location, whose shares were owned by the state’s pension fund. Lee then used within-company variations in state-pension-fund ownership as well as variations in state FCAs, such as the year of passage and the scope of coverage, to identify companies that were subject to whistle-blowing laws to some degree.
The second sample consisted of companies subject to the SEC whistle-blower program. Here, Lee’s control group consisted of companies that were previously exposed to state FCA provisions. Companies that had not previously been exposed to state FCAs tended to have a more marked response to the new federal whistle-blower provision.