Eugene Volokh, And the Survey Says . . ., MSNBC's,

June 27, 2003



I like reading and writing about surveys, for two reasons.  First, they're often seriously flawed, and finding and reporting these flaws is fun and, I hope, valuable.


Second, well-conducted surveys do give us an opportunity to actually learn more facts -- and often surprising facts -- about people's attitudes and experiences.  We may think we know what our fellow Americans think, but usually our judgments are just based on our own narrow circle of acquaintances, coupled with prejudices about other circles.  Well-done surveys can give us a broader perspective.  Even the best surveys, of course, have their flaws; but so does the alternative, which is just relying on our intuitions and personal experience.


So, on to two surveys that were recently released.  I'll begin with a Drug Policy Center press release that reports substantial support for legalizing marijuana:

A poll by Zogby International released today found that 41% of Americans agree that "the government should treat marijuana more or less the same way it treats alcohol: it should regulate it, control it, tax it and only make it illegal for children." This represents a striking increase from previous nationwide polls on making marijuana legal. . . .

The poll released today interviewed 1,204 adults chosen at random nationwide.  They were asked to agree or disagree with the following statement: "Some people say the government should treat marijuana more or less the same way it treats alcohol: it should regulate marijuana, control it, tax it, and only make it illegal for children."  The margin of error is +/- 2.9%.

This is a poll based on a random sample, which is good (self-selected surveys, such as most Internet surveys, are bunk, because people who volunteer to answer such surveys are usually wildly unrepresentative of the population at large).  And the question also seems pretty fairly worded.


If you look closely, you see a possible glitch with the question:  Literally, people were asked whether they agree that "Some people say . . . ."  (I e-mailed the Drug Policy Center for the exact text of the question, and it confirmed this.)  Some respondents might have said "Well, yes, some people do say that; I've heard them say it; so I'll respond 'I agree that some people say . . .,' though I strongly oppose the pro-legalization sentiment itself."  So the text of the question is one possible source of error -- but I suspect that most respondents (unlike law professors who, like me, used to be computer programmers) wouldn't be that literal.  And while the 41% is higher than I had expected, it's not that far from the results that some other polls have shown (see, for instance, here, revealing a 34%-58% split on a more generally worded question in Fall 2002).


Now I'm pleased by this 41% result, because I tentatively support legalization of marijuana.  I say "tentatively" because I'm not an expert on the subject, and haven't studied the important practical policy questions surrounding this issue.  Still, my sense is that marijuana really isn't any more harmful than alcohol, and the harms of outlawing it are greater than the harms of tolerating it.  The experience of alcohol Prohibition suggests (though by no means proves) that marijuana prohibition is likely a mistake.


And yet, if this is so, then why are there so few politicians who publicly support marijuana legalization in their campaigns?  True, 57% of the public opposes marijuana legalization -- but maybe that's precisely because politicians haven't really made the pro-legalization case yet, which provides an opportunity for savvy challengers.  So you'd think that in at least some liberal areas, where the support is higher than 41%, pro-legalization candidates would try to make something of this issue.


There are, however, two likely reasons why the picture is less bright for the legalization forces than one might think.  First, I asked the Drug Policy Center for a breakdown of the results not just by agree vs. disagree, but also by the strength of agreement.  (Many surveys, including this one, give people several options, and not just two.)  Here's what that breakdown ends up being:

Strongly agree with legalization -- 23.5%.

Somewhat agree with legalization -- 17.4%.

Somewhat disagree with legalization -- 11.4%.

Strongly disagree with legalization -- 45.3%.

Now thaе isn't good news for the legalizers:  Nearly half the public strongly opposes legalization, and less than a quarter strongly supports it.


In politics, the intensity of feeling is more important than the magnitude of feeling.  That's why substantial minorities that feel very strongly can often block policies that are supported by majorities who feel less strongly.  (We see that phenomenon in gun debates, where gun rights advocates are often able to block modest gun control proposals that seem to have majority support; we also see it as to race preferences, where preferences supporters are often able to defend race-based affirmative action programs that seem to be opposed by the majority.)  So a candidate might be able to win on an issue where 57% of the public mildly disagrees with him -- but it's much harder to win where 45% of the public strongly disagrees with you, and another 11% mildly disagrees.


Second, and this is just a guess on my part, my sense is that opposition to marijuana legalization is pretty evenly distributed throughout the nation.  Some Congressional or legislative districts, or some states, might be much more pro-choice and others might be much more pro-life.  Some might be pro-gun and some might be anti-gun.  But I suspect that opposition to marijuana -- driven largely by concern about its effect on children, and a suspicion that once it's legalized, it will be very hard to keep it out of children's hands -- is much more evenly distributed.  If I'm right, then supporting legalization may be a losing proposition in nearly every state and nearly every district, at least unless one limits legalization to medical uses.


Now, the second survey:  The Center for the Advancement of Women -- an organization run by a former president of the Planned Parenthood Federation of America -- commissioned a Princeton Survey Research Associates survey of American women.  The Princeton survey likewise picked a random sample of 1000 women age 18 and above (this yields a margin of error of +/-3%), and it asked the following question, which I reproduce together with the results:

Which ONE of the following four statements comes CLOSEST to your own view on abortion?

A - Abortion should be generally available to those who want it;  [30% picked this]

B - Abortion should be available but under stricter limits than it is now; [17% picked this]

C - Abortion should be against the law except in cases of rape, incest, and to save the woman's life; [34% picked this] or

D - Abortion should not be permitted at all [17% picked this]

Did not give an opinion on abortion [2%]

That's right -- 51% +/- 3% of American women apparently believe that abortion should either be entirely illegal, or legal only in cases of rape, incest, and to save the woman's life.


Are these survey results accurate?  It's hard to tell.  The results surprised me quite a bit -- the most similar recent poll that I could find (see here, and scroll down to the NBC News/Wall Street Journal Poll, Jan. 19-21, 2003) revealed that 38% +/- 4.5% of Americans (male and female) took this view.  Maybe American women are much more pro-life than American men, or maybe one or both of the polls is inaccurate.


Still, the results seem worth focusing on, whether one is pro-life or pro-choice.  Naturally, they don't tell you anything about what's right and what's wrong; for instance, I generally oppose abortion bans, and would continue to do so regardless of the polls.  But the poll does suggest that the pro-life position may be more politically viable than some might think.  People who follow politics, as well as people who follow abortion policy, should keep that in mind.


And, as blogger Stuart Buck has pointed out, the poll does pretty clearly show one other thing:  One can't honestly condemn pro-life candidates (at least ones who recognize some exceptions for rape, incest, and when the woman's life is in danger) for being supposedly "out of the American mainstream" on abortion.  For better or worse, pro-life views are very much a part of the American mainstream.