OK, this is admittedly a shallow observation, since I know neither of these guys personally, but this just seems like the intellectual equivalent of a boxing match between Muhammad Ali and Punky Brewster.

 Baldwin? I could have mistaken that for a SNL parody....

Glenn Greenwald has been working on a 50 page paper for the Cato Institute that he will present on April 3 at noon at Cato Institute in Washington.

In 2001, Portugal became the only EU-member state to decriminalize drugs, a distinction which continues through to the present. Last year, working with the Cato Institute, I went to that country in order to research the effects of the decriminalization law (which applies to all substances, including cocaine and heroin) and to interview both Portuguese and EU drug policy officials and analysts...Evaluating the policy strictly from an empirical perspective, decriminalization has been an unquestionable success, leading to improvements in virtually every relevant category and enabling Portugal to manage drug-related problems (and drug usage rates) far better than most Western nations that continue to treat adult drug consumption as a criminal offense.

Great work, Glenn. See his link for more info...

Baldwin's gateway drug argument is pathetic. When he makes completely unfounded assertions such as "marijuana leads to doing worse things. That's just a fact.", I can only conclude that he is a liar.

Without looking at the research on the issue, I can confidently state that he is unjustified in making strong assertions about the gateway drug hypothesis:

1) We can't do controlled experiments on human drug use. We cannot take one group of kids, give them pot, and then come back 10 years later and see if they are using harder drugs. Selection bias will seriously complicate any study examining the relationship between marijuana use and other drugs. Reserachers can try to account for this by using fancy statistics, but it is almost impossible to draw any conclusions.

2) Even if we could examine the impact of pot smoking on other drug use in America, we would still have no information about whether this relationship would hold up if marijuana were legalized while leaving the other drugs illegal.

So, Baldwin has no basis for making his claims. The fact that he followed his comment with "I don't care what anyone says" indicates that he has no empirical basis for his claim. He's just making it up.

I did a PubMed search for "marijuana gateway drug ":

Here's what I found:

There is a well established correlation between early use of alcohol, tobacco, and marijuana, and later use of other illicit drugs. Of the studies that I looked at, few actually tried to separate out causal factors of later drug use

Understanding the association between adolescent marijuana use and later serious drug use: gateway effect or developmental trajectory?

Rather than supporting the interpretation that earlier marijuana use "triggers" later hard drug use, these results suggest that the longitudinal pattern of drug use that has been interpreted as the "gateway effect" might be better conceptualized as a genetically influenced developmental trajectory.

Reassessing the marijuana gateway effect .

CONCLUSIONS: Marijuana gateway effects may exist. However, our results demonstrate that the phenomena used to motivate belief in such an effect are consistent with an alternative simple, plausible common-factor model. No gateway effect is required to explain them. The common-factor model has implications for evaluating marijuana control policies that differ significantly from those supported by the gateway model.

This later article lead to a series of comments by other researchers , some seemed to be frantically re-emphasizing that the existence of an alternative to the gateway hypothesis does not automatically invalidate the gateway hypothesis. Others proposed ways to distinguish between the gateway hypothesis and the common-factor hypothesis. I got the impression that there really hasn't been much serious research on the gateway hypothesis.

One of these researchers (MICHAEL LYNSKEY) pointed out the following item of interest:

In the Netherlands it was argued that the link between cannabis use and progression to other drug use arose from the social context in
which cannabis was used and obtained. The de facto
legalization of cannabis use was therefore seen as a strategy to
reduce ‘hard’ drug use by separating the hard and soft drug markets. To the extent that the probability of cocaine use conditional on prior cannabis use appears
lower in the Netherlands than in the US, there is some evidence to suggest that this strategy may have been at least partially successful (MacCoun & Reuter 2001).

In the end, I guess the issue is whether you are comfortable throwing people into jail due to indirect problems that their behavior may cause, especially when the causal relationship between their current behavior and the future problems is largely speculative.