Having recently read Joe Richie's remarks on A Disturbing Coincidence about a brief confrontation between New York Senator Hillary Rodham Clinton and Pennsylvania Senator Rick Santorum over their differences regarding child-rearing (13 July 2005), I have come to the conclusion that these two Senators ought to be observed more closely, and therefore will be the next two major subjects of the LEO test.
However, before we go into a detailed analysis of Clinton and Santorum, both potential presidential contenders in 2008, it is worth noting that the LEO test is not merely limited to political officeholders or candidates. Mark has expressed interest in applying the test to the opinions of Supreme Court Nominee John Roberts, an analysis of which should appear as soon as I have collected the samples for testing.
Meanwhile Kevin at TechnoGypsy expressed dismay that James Taranto of the Wall Street Journal's online Opinion Journal placed political pundit Andrew Sullivan in the "conservative" camp ("Best of the Web." 29 July 2005).
Knowing that one is ill-advised to adopt someone else's conclusions as one's own, especially when it comes to someone with whom I am completely unfamiliar, I decided to find out for myself by submitting a sample of Sullivan's writing to the LEO test. However, unlike previous case studies, which involved using a single test table to measure the frequency of key ideological indicators, I decided to measure the Sullivan sample four times, using four distinct test tables of ideological indicators. The first test table constitutes the original, while the second is a slightly expanded table. The third table is greatly expanded, while the fourth table a reduced variation of the third, but including different indicators that did not appear in the first table.
The Aggregate score of these four tests should provide a sharp image of the ideological preference indicated in the Sullivan sample.
The sample itself is relatively short. It includes the full text of three articles attributed to Sullivan that appear on his website, AndrewSullivan.com. These articles are "Not Dead Yet: An Apology", "This I Believe: An American Creed", and "Still Pro-War: Despite the Flaws". These articles were pasted into a single text file and submitted to the four measures described above.
In none of the above tests did predefined conservative indicators lead the sample. On the contrary, conservative indicators in all four tests trailed the other two, suggsting that Order, the principal concern of the conservative, is actually the least important of the three pillars of American politics for Mr. Sullivan. The test results above indicate a generally libertarian rather than a conservative position, and the aggregate score, i.e. the sum of raw scores from each test, suggests that Andrew Sullivan could be described best using the familiar term 'left' libertarian. A comparison of the adjusted scores reveals the general extent of the libertarian preference in the Sullivan sample (Click to Enlarge):
This chart shows that for Mr. Sullivan, liberty is generally of principal importance. Even for the two tests that suggest some form of egalitarianism, libertarian indicators nevertheless remained a major component of Sullivan's rhetoric--more than 1/3 for both test tables. In particular, the third test of the Sullivan sample yields a signature similar to the 19th Century liberal thinker Thomas Hill Green, who demonstrated an equal preference for liberty and equality, and who advocated changes to the established order to enforce equality for the sake of greater liberty.
The distribution of ideological indicators suggests that by and large, Sullivan maintains a moderate position. Even so, one table does suggest a latent liberalism, specifically a preference for equality characteristic of a liberal ideologue. Granted, this one test alone may simply have failed to capture every ideological indicator present in the sample, yet when placed alongside the other test results, the picture remains quite clear that Andrew Sullivan is simply not a conservative.
If Mr. Sullivan is still labeled a conservative by anyone, it is either because those who believe Sullivan to be conservative subscribe to the linear ideological model, or because the Reagan Coalition in the Republican Party has effectively blurred the distinction between these two clearly different ideological preferences.