Lying in Ponds

Some readers have questioned whether the Lying in Ponds methodology of counting references can really measure partisanship. The difficulty of evaluating the partisanship of an individual column is shown by the latest Claudia Rosett column. It is a creatively vitriolic attack on Hillary Clinton, written in the form of a letter from Ms. Clinton to Martha Stewart, offering advice concerning Ms. Stewart’s recent difficulties:

Girl, eight years ago I was on the ropes, dressed in my little pink suit, defending my measly $100,000 windfall and my down-home real estate projects, and (here’s another handy phrase) just “trying to find my way through.” Today I’m worth millions, I’ve got a seat in the Senate, and my eye on the White House (if I can just keep Bill on ice for another two years). Martha, look and learn. Or the day will come when you realize you mighta said shoulda, coulda, woulda–but you didn’t.

Although she presents a litany of criticism against Ms. Clinton worthy of a Michael Kelly or Robert L. Bartley column, the style of the column results in only two negative Democratic references. While the partisanship index of the column appropriately has a value of -100 (Republican), the effect of the column on Rosett’s overall index will be less than it should be. My take is that it may be impossible to avoid this kind of thing, and that it shouldn’t be too bad as long as it random rather than systematic.

Spinsanity’s Bryan Keefer sharply criticizes two columnists in our rankings:

The last few days have seen several jargon-filled attacks on the Bush administration from prominent commentators, including Maureen Dowd and Paul Krugman of the New York Times and filmmaker and author Michael Moore. All three have used loaded, manipulative rhetoric to suggest the same thing: President Bush is attempting to make himself a dictator.
…
If Dowd, Krugman and Moore wish to disagree with Bush’s policies, they should do so on the merits, rather than attempting to discredit them by suggesting Bush wants to appoint himself dictator. Lurid claims about “paranoid nightmares” and fictitious threats are not substantive contributions to the debate.

Only five points separate third place from ninth place on the Lying in Ponds Top Ten. After five Republican-leaning columns in a row, Brendan Miniter has risen to sixth place. His ranking is mostly based on his 44 negative Democratic references — half of them to Bill Clinton.

Thanks to aboutpolitics.com for making Lying in Ponds their political site of the day yesterday!

With his latest anti-administration screed, Paul Krugman retakes the lead in negative Republican references from Mary McGrory. Krugman’s 15 references gives him 355 for the year, compared to 348 for McGrory in the same number of columns (49). McGrory however, has also made 137 positive Republican references, more than seven times more than Krugman’s 19.

There is now a permalink attached to Each day’s boxscore and comments; the permalinks lead to a monthly archive page. At the end of the boxscores/comments (left side) column are links to all of the archives pages, as well as some of the “greatest hits”.

There was more movement in the Lying in Ponds Top Ten today. Michael Kinsley writes a non-partisan column about George Stephanopoulos and drops from second to fourth. Peggy Noonan slips out of the top ten after her fourth consectuve not-very-partisan column, and is replaced by Wall Street Journal colleague Brendan Miniter.

Readers may wonder why mentions of George Stephanopoulos in the Kinsley column are not considered Democratic references. The approach is to consider references to Stephanopoulos in his role as a Clinton White House staffer as Democratic, but not to his current role as a member of the media. Former Reagan speechwriter Peggy Noonan is on the same list of ambiguous references. Another twist is former Republican Senator and Clinton cabinet member William Cohen — the classification of each reference to him depends on the context.

Another FAQ entry:

Why don’t you evaluate the partisanship of reporters instead of columnists? Reporters are the ones who are supposed to be unbiased.

While it would be interesting to evaluate reporters, I’ve decided to focus on pundits. Because of limited time, I can’t even evaluate nearly as many pundits as I would like.

With today’s nonpartisan column on recent corporate scandals, Michael Kelly’s partisan score continues to fall. Kelly was once as high as second place in the Lying in Ponds Top Ten, but has now slipped to ninth — he hasn’t used the word “Clinton” in almost a month.

I’ve begun to work on a Frequently Asked Questions (FAQ) page. Some readers respond to Lying in Ponds as Chris Horton did last month:

Your premise is ludicrous. What matters in credibility of pundits is not partisanship, but truthfulness and accuracy. Pundits are expected to have a bias, they are supposed to have opinions, and if they can defend their bias in a factual manner, they are doing readers a service.

So here’s a proposed FAQ entry:

What’s wrong with pundits being partisan? They are expected to have a bias.

Lying in Ponds is trying to draw a distinction between ordinary bias and excessive partisanship. Pundits are expected to have a bias toward one or the other party — nowhere is it suggested that they should be perfectly balanced between the two parties. Instead, the hypothesis is that a very high partisanship index over a significant period of time suggests that a pundit is going beyond ideological bias to portray the world in a simple way — one party as the good guys and/or the other as the bad guys. If you accept the premise that the world is more complicated than that, then a very partisan pundit is not offering a truthful or accurate picture, even if they are careful to be accurate in the particulars of their arguments. Lying in Ponds expects pundits to be independent, not neutral. Over time, any truthful and accurate pundit will be forced to take note of positive things about the other party and shortcomings in their own, and the ranking system is an imperfect attempt to measure that.

Today’s William Raspberry column contains part of his recent North Carolina State University commencement address, but it doesn’t include the section I had highlighted in a comment from last month. The Raleigh News & Observer carried the full text of his speech, including:

Yet we seldom take the effort to recruit allies from among those who hold views different from our own. Conservatives are (to liberals) people who don’t care about minorities, or women, or “the little people,” not decent men and women who have a different view of what works. Liberals are (to conservatives) people who want only to tax and spend the country into bankruptcy, not thoughtful men and women who want America to work for everyone. Each sees the other as enemy.

Surprisingly often, I have discovered, a focus on the problem, rather than on political enemies, could disclose common interests and lead to innovative solutions.

Thanks to Yahoo for featuring Lying in Ponds on their What’s New page on Friday.

The transcripts from the NPR On the Media interview have been posted.

POPULARITY AND PARTISANSHIP?: There was an interesting question asked in the NPR interview which was edited out of the final segment. Bob Garfield noted that many of the pundits with the highest partisanship scores are also the best-known or most “popular” (e.g. Krugman, Kinsley, Noonan, Kelly). I fumbled around and agreed with him, but I’m not sure after looking at the numbers. First, I don’t really know how to tell which pundits are the most popular. Maybe there is some kind of list of which columnists are the most widely-syndicated or something like that? If anyone knows of such a thing, let me know.

But it seems like the big names are fairly well-distributed throughout the list — Dowd is in the middle, and Broder, Friedman, Dionne, Will, and Safire are near the end. So count me skeptical about any strong relationship between partisanship and popularity.