Lying in Ponds

In last year’s summary, I wrote about a national political shift away from the Bush administration and toward the Democrats, and the way that pundits on both sides seemed to respond by being somewhat more willing to criticize their own parties. That tendency continued and strengthened this year — even the most partisan columnists had several crossover columns. Nearing the end of the Bush administration, Republican pundits feel free to sharply criticize the administration at times, and also to choose sides in the wide-open Republican presidential race. Democratic pundits who once concentrated their fire on the Republican White House and Congress, are now willing to occasionally criticize the Democratic Congress and their least-favorite Democratic presidential candidates.

Ann Coulter had the highest partisanship score in 2007, because of her off-the-charts contempt for Democrats, with 516 negative references to only 10 positive references. That 51-1 ratio is by far the most extreme of any of our pundits toward either party. Her most common positive Democratic reference (2 of 10) was to FDR; it’s an obvious sign of partisanship when a member of the opposite party that has been dead for over 60 years is high on the praise list. Ms. Coulter has been critical of Republicans on immigration, and she is not a fan of Mike Huckabee. The rest of our Republican pundits have lower partisanship scores, because they also criticize the Bush administration and their criticism of Democrats is less excessive.

Last year, Joe Conason and Paul Krugman were virtually tied for second behind Molly Ivins as the most partisan Democratic pundit, and this year they’re virtually tied for first, with Joe Conason slightly ahead. This year I evaluated Mr. Conason’s columns from both The New York Observer and Salon, and his high partisanship score resulted from a 14-1 ratio of negative to positive Republican references and about a 3-1 ratio of positive to negative Democratic references. The biggest story though, was that both Mr. Krugman and Mr. Conason sharply criticized Democrats in a few crossover columns, particularly at the end of the year when they both attacked Barack Obama. For Mr. Krugman, this represents a departure from four years ago, when he avoided criticism of any of the Democratic candidates.

It will be interesting to see if 2008 will be similar to 2004, when the presidential election seemed to generally increase pundit partisanship, particularly for some of the columnists (e.g. Charles Krauthammer) whose scores spiked upwards. My “free time” to work on this website continues to be nearly non-existent, so I’ll just have to keep plugging away to maintain it at a minimal level.

I think we can safely render verdicts on a a couple of pundit predictions from earlier this year:

1. A year ago Ann Coulter boldly predicted that a Democratic Congress would “instantly set to work enacting a national gay marriage law, impeachment hearings, slavery reparations and a series of new federal felonies for abortion clinic protesters.” But rather than working furiously to implement a “crazy” agenda, this Congress has instead been criticized as a Do-Nothing Congress. For that, we’ll give Ms. Coulter a “VERY WRONG”.
2. The resignation of Attorney General Alberto Gonzales in August came in spite of many predictions that it would never happen. Joe Conason’s prediction in Salon was unequivocal: “No matter how discredited he is — and no matter how much damage he continues to inflict on the Justice Department — this attorney general will not resign.” So we’ll give Mr. Conason a “VERY WRONG” as well.

Brendan Nyhan has pointed out a couple of other Gonzales predictions that were based on the premise that he was “irreplaceable” (Josh Marshall) or “indispensable” (Sidney Blumenthal) to protect President Bush from scandal:

Just a few weeks ago, big liberals were saying that Alberto Gonzales was “irreplaceable” and “indispensable” and promising near-apocalyptic horrors if he were fired, impeached, or decided to resign. So why has the President “grudgingly” accepted his resignation?
. . .
So who gets confirmed? And do the “horrors” come out? Marshall doesn’t acknowledge his previous claim that Bush needs Gonzales in a post on the resignation this morning.

Making an occasional bad prediction is not a terrible thing (I’m a meteorologist, after all!), but pundits very rarely question the premises upon which their faulty predictions were based, or even acknowledge their mistakes at all. Will Ms. Coulter write a column where she admits that Congressional Democrats are not the radicals that she imagined? Mr. Conason followed the Gonzales prediction by speculating (not quite a prediction) that President Bush would nominate Ted Olson to replace him: “If Bush wanted to heal the politicization that has nearly ruined the Justice Department, he would drop Olson and nominate someone else — but then he would be an utterly different kind of president than he is.” Since Mr. Bush instead nominated Michael Mukasey, will Mr. Conason now conclude that Mr. Bush really is an utterly different kind of president?

Last year, Paul Krugman lost the Democratic partisanship championship to Molly Ivins after four winning years (2002-2005). Now Frank Rich has inched past Mr. Krugman in the 2007 rankings, on the strength of his nearly non-stop anti-Bush administration screeds, averaging over 30 negative Republican references per column (965 in only 31 columns). In the past, I used to praise his willingness to offer sharp criticism of Democrats, but this year his occasional criticism of Hillary Clinton has been more than offset by lots of praise for Barack Obama, John and Elizabeth Edwards, Jim Webb, Henry Waxman, etc. Maybe he’ll step up his criticism as the presidential race heats up toward the end of the year.

But it surely must be a sign of partisanship when a pundit’s most frequent positive references to the opposite party are to dead people. Ann Coulter’s most frequent positive opposite references this year are to FDR (The only good Democrat is a dead Democrat). Paul Krugman’s are to Ronald Reagan, and Frank Rich’s top two are Gerald Ford and Reagan (the only good Republican is a dead Republican).

A few observations at the haldfway point of 2007:

• Paul Krugman is holding his lead in the Democratic partisanship rankings, and Frank Rich has passed Joe Conason for second place. Mr. Conason’s ranking dropped in June when he wrote a very strong crossover column in Salon criticizing Hillary Clinton.
• Ann Coulter is again running away with the Republican partisanship title, despite a couple of crossover columns criticizing the Bush administration on immigration. Her 60-1 ratio of negative (299) to positive (5) Democratic references is the highest that it has been since her 2004 record of 70-1.
• Frank Rich’s March prediction of a Scooter Libby pardon was mostly vindicated by President Bush’s commutation of Libby’s prison sentence.
• The name of Marc Rich has been coming up frequently as a comparison to the Scooter Libby commutation, so I was curious if Paul Krugman would mention it. One of the reasons that Mr. Krugman has been able to achieve such high partisanship scores is his meticulous avoidance of Democratic scandals — he has not mentioned the name of Marc Rich in any of his more than 600 NYT columns since the beginning of 2000. True to form, he wrote about the Libby case this past week without mentioning Marc Rich, unlike fellow Times pundits Frank Rich and Maureen Dowd. On the other hand, E.J. Dionne also discussed the commutation without mentioning Marc Rich.

Finally, congratulations to Joe Conason for the two new additions to his family.

In his most recent Salon column, Joe Conason raises the possibility of an impeachment, and makes a clear-cut prediction about the future of Alberto Gonzales:

If the House Judiciary Committee session Thursday starring Alberto Gonzales produced few revelations about the suspicious dismissal of eight (or nine or more) U.S. attorneys, the hearing did clarify a critical political reality. No matter how discredited he is — and no matter how much damage he continues to inflict on the Justice Department — this attorney general will not resign.

We’ll add that to our prediction list and see what happens.

Beginning today, the rankings reflect only 2007 columns. As usual, the rankings can be expected to flop around quite a bit in the first couple of months due to small sample sizes.

Some changes for this year:

1. I’m now evaluating Joe Conason’s columns from both Salon and the New York Observer, to address the question of whether his partisanship score might be systematically different between the two sources.
2. Because retrieving columns from Townhall.com has become more difficult, I’m now trying to get columns by several pundits from Jewish World Review instead.
3. Molly Ivins wrote two columns before her passing; I’ll be looking for another liberal pundit to put on the roster in her place (suggestions welcome). I’m also not sure why Michael Kinsley has not written a column this year.

Molly Ivins has been battling cancer for some time; our thoughts and prayers are with her in her latest bout.

UPDATE (1/31/2007): I’m very sad to hear that Molly Ivins has passed away. Creators Syndicate has a tribute.

In the year-end summary, I pointed out that both Paul Krugman and Joe Conason wrote anti-Lieberman columns which qualified as “crossover columns” — a first for both of them (of the several years of columns evaluated here). Commenters Larryesq and Andrew Myers both made thoughtful objections, arguing that a column criticizing a Democratic heretic should not be considered a crossover.

Certainly they have a point — a column criticizing Joe Lieberman does not indicate non-partisanship to the same extent that one critical of Hillary Clinton or Barack Obama would be. But Mr. Conason’s crossover and the first of Mr. Krugman’s two anti-Lieberman crossovers were written during the primary campaign, while Joe Lieberman was a Democratic senator fighting for the Democratic nomination against another Democrat. In that situation, pundits have a wide range of options. The strongest partisan Democratic approach would be to praise both candidates, and the most partisan Republican approach would be to criticize both candidates. A pundit could praise one and criticize the other (or praise one and ignore the other, etc.), and their partisanship score would be somewhere in the middle. The Lying in Ponds approach is to simply count all of the references in an attempt to (imperfectly) quantify partisanship. Both Krugman and Conason spent much more time criticizing Lieberman than they did praising Ned Lamont, so their columns were counted as crossovers.

I think that designation is appropriate, because they could have easily chosen a far more partisan path. They could have chosen to play down any disagreement within the Democratic party, by lavishly praising Lamont and minimizing direct criticism of Lieberman. That’s exactly the approach Paul Krugman took in the 2004 Democratic primary season. He expressed a strong preference for Howard Dean and Wesley Clark, but avoided direct criticism of the other candidates. Then when John Kerry took the lead in the race for the Democratic nomination, Mr. Krugman was more favorable toward Mr. Kerry than any of the other 33 pundits that year.

Writing a couple of crossover columns doesn’t mean that a columnist has become David Broder. Mr. Krugman and Mr. Conason are both extremely partisan columnists, indicated by thousands of individual references in the hundreds of columns I’ve evaluated here. But they’re not perfectly partisan, and in the case of Joe Lieberman, they chose to directly attack a Democratic heretic rather than simply ignore him or praise his primary opponent. So they get some credit for that.

I do think that I should begin considering Joe Lieberman references to be non-partisan, since he has officially changed his status:

In 2006, Senator Lieberman was elected to a fourth term as an Independent, because of the strength of his record and his accomplishments for the state. He won the general election by more than 100,000 votes. He remains committed to caucusing with Senate Democrats, but will be identified as an Independent Democrat (ID-CT).

I’ll make the change beginning now.

2006 was the fifth year of Lying in Ponds, and the partisanship rankings reflected a significant shift in the national political landscape. Republicans were in complete control of the White House and both houses of Congress, but their political standing fell to such low levels that it clearly affected the behavior of both Democratic and Republican pundits. My interpretation is that this lopsided environment has driven even some of the most partisan Republican pundits past a kind of tipping point, where they consider the White House and Congressional leadership to be a lost cause. These columnists have then been more willing to criticize their allies directly, perhaps hoping to influence the future direction of the party, where they had slavishly defended them in the past. At the bottom of the rankings, Charles Krauthammer and David Brooks actually shifted enough that their partisanship scores ended up slightly on the Democratic side. Similarly, with the Democratic Party in ascendance, even some of the most partisan Democratic pundits have relaxed to the point where they consider it safe to take time to criticize the least favorite members of their own party (e.g. Joe Lieberman), where they would have previously either criticized them tepidly, defended them or avoided the subject.

So with Republicans reeling, some of the most partisan Democratic pundits continued their usual criticism of Republicans, but also became marginally more willing to criticize Democrats. In January, Molly Ivins wrote her first crossover column after 270 columns, sharply criticizing Hillary Clinton. She had a couple of other crossovers, even one praising John McCain, Lindsay Graham and John Warner. But the usual extreme partisanship of the rest of her columns was enough to lead the rest of the Democratic pundits, after finishing near the top in previous years. Her partisanship score was a little lower this year (69) than last year (74), but enough to win as the rest of the field moderated more than she did.

That means that for the first time in the five years I’ve been analyzing columns, Paul Krugman was not the most partisan Democratic pundit. Lightning struck in May, when he wrote his first ever crossover column after more than five years and 592 New York Times columns. His first crossover contained 16 negative references to Joe Lieberman, which are more than his total number of negative Democratic references of all kinds in some entire previous years, and Mr. Krugman later wrote one more anti-Lieberman crossover.

Joe Conason also wrote his first crossover in three years, also directed at Joe Lieberman. In May, Mr. Conason objected to being compared to Ann Coulter, and I responded by looking at the issues of civility and partisanship.

On the Republican side, Ann Coulter seemed to reach the tipping point a year earlier. Her ratio of positive to negative Republican references was an remarkable 7.8 in 2004 (467 positive to 60 negative), but dropped precipitously to 1.7 in 2005 and then increased somewhat to 2.4 in 2006. She wrote five crossover columns criticizing Republicans, but her usual allotment of anti-Democratic rants (458 negative and 21 positive Democratic references) kept her partisanship rating high enough to lead all Republican pundits for the third time (out of four years). In addition to partisanship, Ms. Coulter continued her tradition of wishing for the death of her political opponents, by titling a column “They Shot the Wrong Lincoln” (referring to liberal Republican Senator Lincoln Chafee).

Michelle Malkin bucked the general trend; she criticized Republicans similarly in 2005 and 2006 (frequently on immigration and national security), but her criticism of Democrats was much sharper in 2006. She had 172 negative to 11 positive references (15.6) in 2006, compared to 103 negative to 23 positive (4.5) in 2005. The only explanation I can see for that anomaly is that she wrote a series of especially partisan columns in the weeks before and after the midterm elections (sample titles: “Blabbermouth media, blabbermouth Dems”, “The Democrats’ military disdain”).

Mark Steyn was the most partisan Republican pundit in 2005, based on six months of columns. This year, his contempt for Democrats remained undimmed (216 negative to 9 positive Democratic references), but substantial criticism of various Republicans led to five crossover columns and he therefore slipped to third place in the Republican rankings. His references to Republicans were 112 positive to 31 negative in 2005 (3.6), but only 93 positive to 89 negative in 2006 (1.0).

I enjoyed analyzing the columns as usual, but a severe lack of time led to even less commentary than last year. I’m sorry about that, but with job and family responsibilities as likely to increase as decrease, I’m expecting the struggle to continue. But I’ll keep up the partisanship rankings, and I’ll continue to try to document predictions and corrections. Thanks as usual to readers and friends like the soon to be bridge Grand Life Master Peg Kaplan!

Since I’m so far behind, these will have to be a little brief:

1. Mark Steyn was so certain that Republicans would hold both the House and Senate this year that he predicted it three times (in October 2005, in June 2006, and again in September 2006). The Democratic Party actually gained 30 House and 6 Senate seats, taking control of both chambers. We’ll mark all three predictions as “Very Wrong” in our list.
2. Another Mark Steyn prediction suffered a similar fate — he said in October 2005 that “Baby Assad’s fall will not be long now”. Since Bashar al-Assad is still the president of Syria as far as we know, that’s another “Very Wrong” for Mr. Steyn.
3. Frank Rich boldly predicted that the movie Brokeback Mountain would be a “runaway phenomenon“. The results were somewhat mixed. The movie made lots of money, and won critical acclaim and tons of awards. But the box office results were good but not great. It finished as the 22nd highest-grossing movie released in 2005, between Saw II and Monster-In-Law (Star Wars: Episode III - Revenge of the Sith was #1). Brokeback Mountain was shown in fewer theaters than any of the films with higher grosses, but even on a per-theater basis, it earned about 30% less money than King Kong, the movie that Mr. Rich explicitly used for comparison. So let’s just call the outcome of the prediction “Mixed”.
4. I put Molly Ivins‘ statement that “Then whom do we blame when we lose the republic?” on the prediction list, but upon further reflection, it’s probably not fair to call it an actual prediction. I guess it’s just a rhetorical question, based on the kind of exaggeration one would expect from Ms. Ivins, ranked as the most partisan Democratic pundit this year. So let’s just designate that one “Not Really a Prediction” and forget about it.
5. Frank Rich is building a record of impressively bad predictions concerning the Bush administration — he was certain in July 2005 that Karl Rove was about to resign because of the Plame case. Then in June of this year, he made one-and-a-half election predictions, neither of which have panned out. The first was: “While the Democrats dither about Iraq, you can bet that the White House will ambush them with its own election-year facsimile of an exit strategy, dangling nominal troop withdrawals as bait for voters.” But the number of U.S. troops in Iraq actually increased by about 10% (PDF file) from June to the election. Discussions of future troop levels before the election were mostly put on hold while waiting for the Iraq Study Group report. He also speculated that Karl Rove might push Donald Rumsfeld out of office in time to influence the election (the change was actually announced the day after the election). So we’ll give him another “Very Wrong” rating.

To summarize, I wouldn’t recommend betting the ranch on any Mark Steyn or Frank Rich predictions. I don’t think we should be surprised that pundits often tend to make mistakes when predicting the worst for their political opponents — partisanship and wishful thinking are significant barriers to accurate analysis. The prediction list is in the right sidebar; eventually I’ll try to organize and present it in a better way.

I was surprised when I read today’s John Tierney column ($):

Whatever they do the next two years, I won’t be here to kick them around. This is my last column on the Op-Ed page. I’ve enjoyed the past couple of years in Washington, but one election cycle is enough. I’m returning full time to the subject and the city closest to my heart: science and New York. I’ll be writing a column and a blog for the Science Times section.

I hate to abandon my libertarian comrades here fighting in the belly of the beast, but this is the right moment to leave. After six years of libertarians reluctantly electing Republicans as the lesser of two evils, we’ve finally had enough. We’ve voted out big-government conservatism, and the result is the happy state of gridlock. For now, our work is done. See you in January in a new column on a new page.

It will be interesting to see who the NYT picks to replace him.

In her November 1 column, Ann Coulter explicitly predicted the actions of a Democratic Congress:

Even a dying party has death throes. If Democrats win a slight majority in the House or Senate, Americans will get shrill, insane leadership of the nation in time of war.

Democrats can’t not be crazy. They will instantly set to work enacting a national gay marriage law, impeachment hearings, slavery reparations and a series of new federal felonies for abortion clinic protesters. The only way to get Democrats to focus on terrorists would be to convince them that the terrorists are interfering with a woman’s right to choose or that commercial jetliners exploding in midair are a threat to America’s wetlands.

We’ll check back later to see how well subsequent events conform to her forecast.

One of our favorite bloggers, Brendan Nyhan (formerly of Spinsanity), experienced the full fury of one partisan wing of the blogosphere recently when he dared to offer some principled criticism in the wrong direction. Mr. Nyhan also made a couple of sloppy errors which gave his critics ammunition, but the ferocity of the abuse that resulted was wildly out of proportion to his mistakes. He later noted a couple of interesting aspects of the dust-up.

Mark Steyn originally predicted the outcome of the 2006 elections last year and he repeated it in June. In last week’s column he repeated it a third time:

The foaming leftie routine seems to be a tough sell to a general audience. I see that, a mere three weeks after I guest-hosted for Rush, the widely acclaimed and even more widely unlistened-to Air America is going belly up. Coincidence? You be the judge. But I doubt the “liberal” radio network would be kaput if anti-Bush fever were about to sweep the Democrats to power this November. I think I said a few months back that the Dems would be waking up to their usual biennial Wednesday morning after the Tuesday night before, and I’ll stick with that.