Andrew Gelman

Andrew Gelman is a professor of statistics and political science and director of the Applied Statistics Center at Columbia University. He has received the Outstanding Statistical Application award from the American Statistical Association, the award for best article published in the American Political Science Review, and the Council of Presidents of Statistical Societies award for outstanding contributions by a person under the age of 40.

Recent Articles

Military officers have different opinions than enlisted personnel

Josh asks about political opinions of U.S. military personnel. Jason Dempsey and Bob Shapiro have done some work on this. Here’s Dempsey’s website, and here’s something he wrote a few years ago that’s relevant to the discussion:

The Military Times released the results of a survey showing that members of the armed services planned to vote for John McCain over Barack Obama by a factor of nearly three to one–this at a time when the Democratic nominee was handily beating his Republican rival in almost all national polls. The survey apparently reaffirmed the long-held conventional wisdom that the U.S. military overwhelmingly backs the GOP. . . .

The truth about the military’s politics, however, is more complex and all too often obscured by narrowly focused polling. Participants in the Military Times survey, for example, tended to be white, older, and more senior in rank–that is, they were hardly a representative sampling of the armed services. . . .

In a study of the Army that I [Dempsey] conducted in 2004 with my colleague, Professor Robert Shapiro, I tried to get a fuller picture of the social and political attitudes of soldiers, producing the first and only random-sample survey to canvass enlisted personnel and junior officers, as well as their superiors. Broadening the survey yielded results that fly in the face of the conventional view. The Army, it turns out, is hardly a bastion of right-wing thought.

It is true that the upper echelons of the military tilt right. My own research confirmed that about two-thirds of majors and higher-ranking officers identify as conservative, as previous studies found. But that tilt becomes far less pronounced when you expand the pool of respondents. That is because only 32 percent of the Army’s enlisted soldiers consider themselves conservative, while 23 percent identify as liberal and the remaining 45 percent are self-described moderates. These numbers closely mirror the ideological predilections of the civilian population. . . .

The political differences between officers and enlisted personnel can be partly explained by a demographic divide. Whereas officers are predominantly white, have at least a bachelor’s degree, and draw incomes that place them in the middle or upper-middle class, the enlisted ranks have a higher proportion of minorities, make less money than officers, and typically enter service with only a high school diploma. Nevertheless, even when controlling for factors like race and gender, officers are significantly more likely than soldiers to identify as conservative. . . .

In addition to its ideological moderation, the Army is not as partisan as popularly portrayed. Whereas 65 percent of Americans think of themselves as either Republican or Democrat, according to the Annenberg survey, my study shows that only 43 percent of the military identifies with one of the two major political parties. Two out of three officers consider themselves either Republican or Democrat, but only 37 percent of enlisted personnel do so.

Officers tend to be not only more partisan, but also more Republican, with GOP affinity strongest among the highest ranks. While I [Dempsey] was unable to fully parse the reason for this, the evidence strongly suggests the pattern is generational. Today’s senior officers entered the Army during the late 1970s and 1980s, a time when the Republican Party had a strong advantage on issues of national defense and the Democratic Party was seen as antiwar if not anti-military. By contrast, junior officers who joined the Army after 2001 are almost as likely to be Democrats as they are Republicans, foreshadowing a possible shift in officer attitudes. . . .

Don't Pull a Tucker Carlson

Charles Murray wrote a much-discussed new book, Coming Apart: The State of White America, 1960-2010.

David Frum quotes Murray as writing, in an echo of now-forgotten TV personality Tucker Carlson, that the top 5 percent of incomes “tends to be liberal—right? There’s no getting around it. Every way of answering this question produces a yes.”

Frum does me the favor of citing Red State Blue State as evidence, and I’d like to back this up with some graphs.

Frum writes:

Say “top 5 percent” to Murray, and his imagination conjures up everything he dislikes: coastal liberals listening to NPR in their Lexus hybrid SUVs. He sees that image so intensely that no mere number can force him to remember that the top 5 percent also includes the evangelical Christian assistant coach of a state university football team. . . .

George W. Bush 2012?

OK, not really. But NYT colleague Brooks does write:

In sum, great presidents are often aristocrats and experienced political insiders. They experience great setbacks. They feel the presence of God’s hand on their every move.

The Politics of Eyeliner

Good catch by Leslie Savan:

Here’s how the New York Post’s Andrea Peyser began a column (“Jobless & Shameless Gal Going for Gold”) on one of the women charging Herman Cain with sexual harassment:

Gold diggers—unite! Sharon Bialek is 50, out of work and, according to one who knows her, she’s a smooth operator living way above her means. From the look of her heavily painted face, she’s also soon to be in acute need of a new tub of eyeliner.

Is “Academically Adrift” statistically adrift?

Jacob Felson points me to this discussion by Alexander Astin of a recent book on college education:

The implications of the study recently released with the book Academically Adrift: Limited Learning on College Campuses, by Richard Arum and Josipa Roksa, have been portrayed . . . in apocalyptic terms: “extremely devastating in what it says about American higher education today” . . . The principal finding giving rise to such opinions is the claim that 45 percent of the more than 2,000 students tested in the study failed to show significant gains in reasoning and writing skills during their freshman and sophomore years.