"Danger Behind the Wheel," a three-day series that begins today in the Portland Press Herald/Maine Sunday Telegram, grew out of one question: Are drivers with suspended licenses more dangerous than other motorists?

In early 2007, after a series of high profile fatalities involving suspended drivers, the newspaper assigned full-time investigative reporter Kevin Wack to find the answer.

Wack's reporting led him to the Maine Department of Public Safety, which maintains a database of statewide motor-vehicle crashes. The database is public record and contains raw investigative reports from crashes on public roads if they result in an injury or at least $1,000 in property damage. The reports also include information on whether a driver's license is suspended.

The department initially refused to provide the newspaper with a copy of the database, arguing it could be obtained only through InforME, a Web site run by NIC Inc., a Kansas-based company authorized by Maine law to sell public information online. InforME offered to sell the newspaper a copy of the database for $96,000.

The newspaper contended that state law required the Department of Public Safety to provide the data directly, and for a reasonable cost. At the newspaper's request, the state Attorney General's Office intervened in the dispute, after which the department agreed to provide a copy of the database to the newspaper at no charge.

The database contains records on about 188,000 crashes. The newspaper focused its analysis on all crashes from Jan. 1, 2003, to Dec. 31, 2006 -- about 160,000 in all -- because they were the only years with complete data. Experts at The Seattle Times converted the database into a usable format. The Seattle Times Co. owns the Portland Press Herald/Maine Sunday Telegram.

Wack used Microsoft Access, a database-management system, to run scores of queries, including tallies of the number of people killed in crashes involving suspended drivers and the number killed in crashes involving properly licensed motorists.

A fatality rate was established by dividing the number of fatalities involving suspended drivers by the total number of suspended drivers involved in crashes; a separate fatality rate was established the same way with licensed drivers. The rates were then multiplied by 1,000 to determine the number of deaths for every 1,000 drivers involved in crashes. Similar analyses were conducted to establish rates of serious injury and alcohol or drug use.

To determine whether the suspended drivers who were involved in fatal crashes had shown a pattern of reckless driving, Wack obtained from the Maine Bureau of Motor Vehicles the driving records of suspended motorists who were involved in fatal crashes. In more than two out of every three cases from 2003 to 2006, records showed that the motorist had at least one prior conviction for driving without a valid license or a drug- or alcohol-related driving violation.

To check Wack's statistical findings, the newspaper hired Jeff Porter, data library director for the National Institute for Computer-Assisted Reporting, a nonprofit program based at the University of Missouri School of Journalism. Porter verified Wack's research by examining the database's content, reviewing the queries Wack created to ask the database questions, and auditing his statistical conclusions.

To flesh out the database's conclusions, Wack interviewed scores of people, including drivers, victims, family members, police, prosecutors, lawmakers and national experts. He also read research studies on suspended drivers, spent hours in a patrol car with sheriff's deputies and visited police roadblocks aimed at nabbing suspended drivers.

The resulting series is being published today, Monday and Tuesday in the Press Herald/Telegram and at pressherald.com.