In the battle to lure the new FBI headquarters to Prince George's County, state officials frequently point to a study that finds the largest share of the agency's employees live in Maryland.
The only problem with the statistic: It doesn't take into account where a single FBI employee actually lives.
Developers in Maryland and Virginia are competing for the proposed new headquarters to replace the roughly 40-year-old J. Edgar Hoover Building in Washington. And elected officials, including the state's congressional delegation, are eager to point to the project's potential economic benefit.
In making the case, Maryland lawmakers often hold up a 2012 study from the Department of Business and Economic Development that estimates that 43 percent of FBI employees live in Maryland, compared with 33 percent in Virginia and 17 percent in Washington.
Since so many employees already live in the state, the argument goes, Maryland is a natural fit for the project.
Few mention that the underlying census data used by the state to generate the report obscures statistics on federal employees who work at security-related agencies, including the FBI, so researchers extrapolated the number from other agencies.
The data is suppressed to reduce risks for those agencies involved in homeland security efforts.
State officials studied a census tract that encompasses the FBI along with a large swath of downtown Washington, a roughly 10-block span from near the White House to the Verizon Center. They analyzed employer-household patterns for federal employees in that area — except for those working for security agencies — and applied the percentage to the FBI.
"While it is not certain that FBI employees live in the same areas as these other federal employees, they have access to the same roads and transit system, and are generally paid on the same levels as these other employees, so their commuting and living choices are assumed to be similar," according to an explanation of the methodology provided by the Department of Business and Economic Development.
Outside data experts said there's nothing wrong with the state's analysis. No one keeps official score of where an individual federal agency's employees live, and there's no one correct way to make an educated guess. Further, the agency has been upfront in explaining its methodology.
But it is a guess, a nuance sometimes lost when elected officials discuss the project.
There are also other ways to slice the data to produce different results.
Another well-known data set that measures commuter patterns, for instance, suggests that 41 percent of workers near the FBI headquarters live in Maryland, compared with 40 percent in Virginia — a much closer spread.