Blogging as Propaganda

Today on the conservative blog, Hot Air, had this story about initial weekly jobless claims. The story's thesis: this new report is further evidence that Obama's employment strategy is failing. The story points out that jobs report was up 25 thousand jobs from the previous week. The story does mention that jobless claims are down but dismisses that as inconsequential.

Put away the streamers and stop the chorus from singing “Happy Days Are Here Again,” at least for a while. Bloomberg reports that initial jobless claims rose again last week by 25,000 as more that 584,000 Americans applied for unemployment benefits for the first time. While the four-week average of initial jobless claims dropped, this increase hints at continued escalation of the national unemployment rate:

...

We may be seeing less firing, but we’re still seeing plenty of it. More than a half-million people had to apply for unemployment last week, which means that we’re not seeing new jobs created.In fact, I’d challenge the notion that we’re seeing less firing. The four-week average hit 559,000, and last week’s number was 554,000. The latest figure will push the average higher in the next couple of weeks, which is what happens when the rate of claims increases.

Now, this story is peculiar on many levels. First, jobless claims is a very technical economic statistic. Unless you cover it regularly, you shouldn't draw any conclusion from it. After all, it comes out every week. How much could something that only measures a week's worth of activity tell us? If you are going to make analysis about jobless claims, you should be reporting on them regularly. That's not what Hot Air does, they only report on jobless claims in random times. No doubt those times match up with bad news for President Obama. Here's what the piece at Hot Air didn't mention. This is the fourth week in a row that jobless claims are under 600,000. It is, however, the second straight week they have increased. So what does that mean? Your guess is as good as mine which is why I wouldn't draw a conclusion from this week's numbers one way or another. So, why would Hot Air draw a definitive conclusion on a random weekly number? It's because the conclusion is much more propaganda than opinion based on fact.

Conservative blogs were doing something similar with the monthly jobs numbers for a while as well. Here were the job losses from February to May of 2009, 710,000, 663,000, 539,000, and 345,000. Now, there's a pattern there and that pattern was that things were improving. (that pattern was of course broken in June when numbers when the Bureau of Labor and Statistics reported a loss of 477,000) If you were to report the jobs numbers honestly, you would have to acknowledge that things were getting better. That's especially true following the May numbers. That's because not only did they show a marked improvement from April, but they beat estimates by over one hundred thousand jobs. How did the same Hot Air cover those numbers? They selectively isolated the unemployment rate, which was higher than expected, and focused on that.

The unemployment numbers for May hit earlier this morning, and it looks like Barack Obama didn’t save many jobs at all. Despite claims from both Obama and VP Joe Biden that the Porkulus package had saved 150,000 jobs, unemployment went up another half-percent to 9.4%, setting a new record for the past quarter century. Unemployment rose across a broad spectrum of demographics, too:

Of course, no one could honestly analyze the May jobs numbers and conclude they were bad or bad for the president. They improved by about two hundred thousand from the previous month and they beat estimates by more than a hundred thousand. Anyone who at the time was pushing a narrative that these numbers were evidence of Obama's failing economic policies was engaging in pure rank propaganda.

Liberal blogs are of course notorious for engaging in rank propaganda. Daily Kos once touted that Olbermann had beaten O'Reilly in the ratings. They never mentioned that on that particular day Olbermann appeared on the flagship NBC station while O'Reilly was on cable. For years, liberal blogs would cite any source, no matter how obscure, that claimed that Bush lied about anything: Iraq, the economy, the GWOT. It was the liberal bloggers that couldn't get enough of the so called "Downing Street Memo" which purported to prove that Bush lied. Of course, this memo was actually the opinion of a single official in the British Intelligence service and not a fact as the bloggers claimed.

The issue of polling is another way in which both sides of the blogosphere engage in propaganda. It's usually done when polls are isolated that show a particular side more favorably. For instance, Republicans have come to treat Rasmussen's polls as the gold standard. That's because Rasmussen's polls have always found Obama's support to lack that of most other polling. Now, that doesn't mean that Rasmussen is wrong, and I've referenced Rasmussen's polls before as well, but isolating a Rasmussen poll and using it as the thesis of a story is totally misleading. So, of course, liberal blogs have taken to attacking Rasmussen.

Some commentators on the right have been pointing to an interesting number that
has been coming from the Rasmussen daily tracking poll, which Rasmussen bills as the "Presidential Approval Index," which Scott Rasmussen only began bringing out
in late 2008. The key questions then are: What is this number, and is it a valid measurement of real popularity? In an interview today with TPM, Rasmussen defended the index's validity against some harsh criticism, saying that intensity of opinion -- the true figure measured by his index -- does indeed matter.

The thing to remember is that this is not simply subtracting all the respondents who disapprove of President Obama from the people who approve. Instead, Rasmussen takes the numbers who strongly approve or disapprove, and then performs this math. As of today, that index number is -10, compared to an overall rating of +1 in Rasmussen's daily tracker.

It would seem at first glance that this number can skew negative -- that is, the people who disapprove of a president are inherently more likely to feel strongly about it, compared to a certain level of lukewarm support for a president. For example, the 2004 exit poll put George W. Bush's strong approval at 33%, to strong disapproval of 34%. But his overall approval was 53% to disapproval at 46%, and he was re-elected 51%-48%.

It's very dubious to have a blogger question the professional judgment of a pollster. Josh Marshall of Talking Points Memo blogs professionally. Meanwhile, Scott Rasmussen polls professionally. I would trust Rasmussen's judgment over Marshall's for the validity of the poll.

To me it's dubious to create a narrative based on polling to begin with. It's even more dubious to focus on one poll, whichever side you are on. Real Clear Politics has a collection of polls that continues to show solid support for the president however it also shows that support has significantly eroded over the last few months. Focusing on polls appears to be beside the point. If health care doesn't pass, the president's poll numbers will fall precipitously. It won't merely be Rasmussen that shows bad poll numbers but everyone. If, on the other hand, the economy turns around, and especially if health care passes, even Rasmussen will show positive numbers. So, why the fixation on polls? It's because there are so many out there that a propagandist can always find a poll that will affirm their pre determined opinion.

Unfortunately in two plus years of blogging I have come to the conclusion that for the most part most bloggers are much more propagandists than they are journalists. Many more bloggers use their blogs to push an agenda rather than to report. They mask opinion that are dressed up to appear as facts. Ironically enough, blogs sprouted in large part because citizens believed that the MSM was engaging in this very behavior. As it turns out, most bloggers are no better than the media they despise only they don't write nearly as well.