Most drug theories are based on reserch data or even empirical observations .then a bord of scientists forms a theory (more or less accurate) which can be applied and explain the phenomenon. If im not wrong, the Partnership for a Drug-Free America, in cooperation with the National Institute on Drug Abuse (NIDA) and the White House Office of Drug Control Policy, (year1995) was the first that announced a new anti-drug campaign that specifically targets marijuana. Instead of featuring horror tales of marijuana-induced insanity, violence and birth detects, this campaign was built upon the premise that reducing marijuana use is a practical strategy for reducing the use of more dangerous drugs.
The primary basis for this "gateway hypothesis" was a report by the center on Addiction and Substance Abuse (CASA), claiming that marijuana users are 85 times more likely than non-marijuana users to try cocaine. This figure, using data from NIDA's 1991 National Household Survey on Drug Abuse, is close to being meaningless. It was calculated by dividing the proportion of marijuana users who have ever used cocaine (17%) by the proportion of cocaine users who have never used marijuana (.2%). The high risk-factor obtained is a product not of the fact that so many marijuana users use cocaine but that so many cocaine users used marijuana previously.
For the nearly 70 million Americans who have tried marijuana, it is clearly a "terminus" rather than a "gateway" drug.