What, exactly, is “science?”

What, exactly, is “science?”  And what is “preponderance of evidence?”


Much has been made in recent days about how “science” proves “overwhelmingly” that a lower drinking age reduces alcohol-related traffic fatalities.  The supporting evidence for this claim may be found in a 2002 article by Alex Wagenaar and Traci Toomey, which identifies 52 peer-reviewed analyses that establish a cause-and-effect relationship between drinking age and traffic fatalities.


This is called “science,” when, in reality, it is mathematics, statistics, probability, abstraction, estimates.  Generating these data did not require a laboratory, nor did it require an advanced degree in any of the natural sciences.  These data are the result of “imputation” (try deciphering NHTSA’s definition of that term), assumption, extrapolation, and the plugging of numbers into a formula.  That’s math, folks, not science.  And it’s an estimate.  And there is not a single, identifiable human life associated with any of those numbers.  How after all count wars not fought?  Fortunes not made?  Friends never met?  Lives not lost?


So let’s begin by using the correct term.  This is “mathematics speaking for itself,” not science.


But when a discipline speaks for itself, it requires no filter, though it demands its hearers not listen selectively.


And here’s why.  The same peer-reviewed article that lists 52 analyses showing a cause-and-effect relationship between a higher drinking age and lower traffic fatalities notes that this number is out of a TOTAL of 102 such analyses.  In other words, about HALF the analyses show a relationship … and HALF don’t.


The other side can say, as they have in each of the following publications, that these are simply one man’s opinions (Parade Magazine).  They can question why an historian knows more about science that scientists (KGO radio).  They can dismiss CR’s assertions as untruths (KGO radio, Fox News).  In short, they can continue ad hominem responses.  Or they can explain what the scholar they cite, who is a member of MADD’s own Board of Directors, means by this statement (p. 213):


“Of the 102 analyses, 52 (51%) found a statistically significant inverse relationship between the legal drinking age and crashes;  that is, as the legal age was lowered, the number of crashes increased, and as the legal age was raised, the number of crashes decreased.”


These are not “off-the-cuff musings.”  These are not “opinions masquerading as facts.”  They are peer-reviewed data that any historian – or political science major – can read and quickly grasp.


So the next time you hear about “more than 50 peer-reviewed studies,” keep in mind:  the actual number is 52.  And that’s out of a total of 102.


Let’s let science (er, rather, mathematics) speak for itself.  Unencumbered by emotion.  Without personal attacks.  And in the service of nothing more, or less, than the truth.

 –[CR] Director John McCardell

2 Responses to “What, exactly, is “science?””

  1. John Searles Says:

    Statistics and more generally mathematics are the tools of science. It’s how scientists actually determine if results are more likely to be real or simply be chance findings. How is historical analysis conducted? Is it simply a subjective view or do you try to evaluate data like we do in the sciences? I am surprised that anyone would question the methods of science.

    While 52% of the studies in this very well conducted meta-analytical study that you cite demonstrate an inverse relationshiop between MLDA and trafffic crashes, it is important to remember that only 1 (1%) of the studies found a positive relationship.

    Your distortion of this study and statistical methods in general is disturbing in light of your call for letting “science speak for itself.”

    I would urge everyone to actually read the study which is avaialble at:


  2. Ajax the Great Says:

    I’d say that John McCardell’s review of the study is the better one. The so-called “preponderance of the evidence” is rather slim indeed, with nearly half of even the higher quality studies finding no effect of the 21 drinking age on traffic fatalities, and the majority of studies finding no effect on alcohol consumption. And the Wagenaar review only included studies from 2000 and earlier, thus leaving out the more recent ones such as those by by Dee and Evans (2001), Males (2008), and Miron and Tetelbaum (2009), three studies that cast doubt on the idea of a lifesaving effect of the 21 drinking age.