"After applying these methods, there is very little remaining evidence of any causal gateway effect. For example, even if soft/medium drugs (cannabis, amphetamines, LSD, magic mushrooms, amyl nitrite) could somehow be abolished completely, the true causal link with hard drugs (crack, heroin, methadone) is found to be very small. For the sort of reduction in soft drug use that might be achievable in practice, the predicted causal effect on the demand for hard drugs would be negligible. Although there is stronger evidence of a gateway between soft drugs and ecstasy/cocaine, it remains small for practical purposes. My interpretation of the results of this study is that true gateway effects are probably very small and that the association between soft and hard drugs found in survey data is largely the result of our inability to observe all the personal characteristics underlying individual drug use. From this viewpoint, the decision to reclassify cannabis seems unlikely to have damaging future consequences."

Source

Pudney, Stephen. Home Office Research Study 253: The road to ruin? Sequences of initiation into drug use and offending by young people in Britain. London, England: Home Office Research, Development, and Statistics Directorate, December 2002.