Research by UAB criminologists Tomislav Kovandzic, Ph.D., and Lynne M. Vieraitis, Ph.D., shows no evidence that prison releases lead to a significant rise in homicide rates. In fact, in a separate study, Kovandzic, Vieraitis, and fellow UAB researcher John J. Sloan III, Ph.D., found that in states with three-strikes laws, homicide rates increase an average 14 percent within three years of passage and 24 percent within five years. The results translate to about 1,300 additional homicides over five years in the states that have such laws.
“Several highly publicized murder cases and a rising crime rate in the 1980s and 1990s brought about the three-strikes laws,” Kovandzic says. “They were meant to make criminals think twice before committing new crimes. But they are not a panacea for the nation’s violent-crime problem, and, according to a growing body of scientific research, they may actually exacerbate the most serious crime: homicide.” The reason, he speculates, is that to felons on their second strike, the difference in the sentence for a third-strike crime and murder would be minimal. “When committing a nonlethal crime, a criminal might kill victims or others at the crime scene to reduce the chances that he or she will be apprehended and convicted,” says Kovandzic.
http://main.uab.edu/show.asp?durki=78147 Declines in drunk driving
The death rate from alcohol related traffic accidents has dramatically declined since the 1970s, according to the National Highway Traffic Safety Administration (NHTSA). According to NHTSA, alcohol related deaths have declined from 26,173 in 1982 to 16,694 in 2004. While some of this decline may be attributable to improvements in car safety and emergency medicine, MADD's supporters claim the group's efforts have brought about this decline, because alcohol-related fatalities declined much more significantly than non-alcohol-related fatalities. Declining death rates in alcohol-related car accidents may have saved hundreds of thousands of lives in the past 25 years. The declining number of alcohol-related car accident deaths found by NHTSA is particulary dramatic because total U.S. population and the number of drivers has increased substantially during the past 23 years.
The statistics kept on traffic fatalities were originally based upon "alcohol-caused" deaths. Critics point out, however, that these statistics were later quietly changed to "alcohol-related." This meant that a sober driver who hit and killed an intoxicated pedestrian, for example, would be involved in an "alcohol-related" incident. Similarly, a sober driver who is struck by another sober driver carrying an intoxicated passenger was counted as another "alcohol-related" death. Further, if the officer believes the driver to be intoxicated but chemical tests show he is not, the death is nevertheless reported as "alcohol-related;" if the tests indicate the presence of any alcohol at all, say .02%, the fatality is counted as "alcohol-related".
In 1999, the federal General Accounting Office (GAO) reviewed the claimed fatality figures from NHTSA and issued a report stating that they "raised methodological concerns calling their conclusions into question ". The statistics, the GAO report said, "fall short of providing conclusive evidence that .08% BAC laws were, by themselves, responsible for reductions in alcohol related fatalities."
http://www.answers.com/topic/mothers-against-drunk-drivingNote at the madd link there is some suspicion madd might be going madd: "Common criticisms of the organization deem it as using junk science to further its goals, and that it is neo-prohibitionist, ageist, and in favor of the creation of a nanny state — all under the pretense of preventing deaths due to drunk driving."
EDIT:suspicion