Archive for September 28th, 2012

Meet Dean Chambers, The Virginia Republican Who Is ‘Unskewing’ The Polls


Meet Dean Chambers, The Virginia Republican Who Is ‘Unskewing’ The Polls.

4
inShare
share
David Taintor September 28, 2012, 4:38 PM 1870

With Republicans up in arms over poll after poll showing President Obama ahead, conservatives have latched on to a new polling site that promises to “unskew” the data, correcting what the site believes are polling samples with too many darn Democrats. Rick Perry has tweeted his approval of the site — which shows Mitt Romney ahead by 7.4 points — and even Stephen Colbert expressed his faux appreciation.

Here’s what the site’s founder, Dean Chambers, does. He changes the baseline assumption on how much of the electorate is Republican and how much is Democratic. Initially, he used Rasmussen’s real numbers on party identification to re-weight various polls. Rasmussen’s numbers break down to 37.6 percent Republican, 33.3 percent Democrat and 29.2 percent Independent. As of Thursday night, Chambers began using party identification numbers from his own web-based poll.

Chambers’ project started in July after he noticed an ABC News/Washington Post poll that “just didn’t look right.” An ABC News/Washington Post poll conducted this month showed Obama up over Romney 49 percent to 48 percent. “Unskewed,” however, after applying Rasmussen’s numbers on party ID, Romney leads Obama 52-45 in the poll. It’s like magic. But Chambers insists he isn’t “changing” or “making up” data. “The only thing I’m doing is weighting.”

But that’s exactly what most pollsters don’t do. “We don’t have any preconceived notions about the party breakdown of a poll before we conduct it. The only things we make any adjustments for are gender, race, and age,” Democratic-leaning Public Policy Polling’s Tom Jensen told TPM in an email. “It makes sense that as support for Obama increases, more people also identify themselves as Democrats. I know conservatives want to think it’s more Democrats in the poll causing Obama to do better, but it’s actually Obama doing better causing more Democrats in the poll.”

“The reality is that (Republicans are) losing, they can’t accept it, and they’re going to find some reason to dismiss every poll that makes them unhappy no matter what its composition is,” Jensen added. “This isn’t really about Party ID, it’s about hardcore denial.”

In every poll Chambers has reworked — save for a recent “unskewed” Fox News poll that has Obama up 2 points — Romney leads the incumbent. While Chambers, 45, of Duffield, Va., is a Romney supporter and longtime Republican, he said he is simply reporting the numbers “as they are.”

It doesn’t quite work that way, though. Scott Rasmussen told BuzzFeed this week: “you cannot compare partisan weighting from one polling firm to another.” Different firms ask about party identification differently, he explained. It’s not apples to apples. Rasmussen added:

“Some ask how you are registered. Some ask what you consider yourselves. Some push for leaners, others do not. Some ask it at the beginning of a survey which provides a more stable response while others ask it at the end.”

Missing from Chambers’ model is the fact that party identification is not a static metric. The current PollTracker average of party identification, which tracks the broader samples of American adults, shows 33.3 percent of citizens consider themselves Democrats, 22.1 percent Republicans and 34.3 percent independents. In 2010, when Republicans swept the House of Representatives and made gains in statehouses across the country, Republican party identification was much higher, around 31 percent. Based on a web poll he is currently conducting on his site, Chambers found only a 0.4 point spread between Democrats and Republicans today, with Democrats holding a narrow edge.

Still, Chambers said he believes pollsters aren’t skewing the data with malicious intent. They are just operating under faulty assumptions, he said, believing there are many more self-identified Democrats in the country today than Republicans. But the media, Chamber said, are over-reporting the polls showing Obama ahead. “That’s driving the analysis,” he said. “If one were to believe all these polls, you would think (the election is) over. It’s not over.”

Meanwhile, here’s the PollTracker average of where the presidential race stands today:

Advertisements

Census Bureau clarifies poverty numbers – U.S. News


Census Bureau clarifies poverty numbers – U.S. News.

Officials at the U.S. Census Bureau moved Friday to clarify widely reported figures meant to estimate the number of Americans living in poverty.

Dueling Census reports – one based on official poverty estimates that was released just last week and another based on an experimental calculus used in November – differed from one another by 20 percentage points regarding the number of people viewed as living in poverty. The widely reported figure showed that one out of two Americans are in poverty or are low-income. Other Census figures put the figure closer to one out of three Americans.
Advertise | AdChoices

That’s because the experimental measure, a supplement to the official poverty figures meant to take into account such factors as whether a family is receiving food stamps and how much people pay in taxes, uses a poverty level of $24,343 for a family of four instead of the $21,113 used by the official measure.

Read the original story on NBCLosAngeles.com

In expensive states like California and New York, the supplemental measure classified families making as much as $32,869 as impoverished.

However, Kathleen Short, the Census Bureau economist who spearheaded the supplemental report, said it would be wrong to extrapolate from those numbers that Americans are falling into poverty at greater rates.

In fact, she said, the experimental calculation indicated that poverty among children is actually lower than the official poverty rate shows.

On Thursday, reports in multiple news outlets suggested that people making roughly twice the poverty level under the experimental program were “scraping by” and should be considered low-income.

The Census Bureau does not support that interpretation of the data, Short said.

“Below 200 percent of the poverty threshold is the lower end of the distribution,” Short said. “But we would not call it low-income per se.”

A number of news reports on Thursday correctly said that more people fell under the definition of living in poverty under the experimental calculation, but then went on to say that people making twice that would be considered low-income.

However, the practice of using such figures as “150 percent of poverty” or “200 percent of poverty” to determine whether people were of moderate or low income is a practice that grew up around the older, traditional method of identifying poverty, which uses a lower threshold.

More from NBC4 on the Census reports: Assessing poverty

In parts of California under the supplemental approach, a family that owns its own home and earns about $66,000 per year would be earning 200 percent of the poverty level, and not necessarily be considered low-income.
Advertise | AdChoices

NBC4, relying on figures and analysis from the Los Angeles office of the Census, reported the newer, official poverty figures on Thursday, and questioned reports that used the supplemental figures.

A widely distributed news story by the Associated Press (and published by msnbc.com) relied on the supplemental report.

“We did not misunderstand the data,” AP spokesman Jack Stokes said in an email. “The AP story was vetted by the Census Bureau in Washington.”

However, Short said that she did not agree with the news service’s conclusion that the report showed one out of two Americans to be low-income or impoverished.

Short stressed that the supplemental measure was a work in progress, and that it should not be considered a replacement for the official poverty rate.

The AP’s calculations were correct, Short said, “but I’m not agreeing with any adjectives that are placed on being in that category.”

“In fact we stressed the Census Bureau does not have a definition for low income.” Short said

In the supplemental report, “we’re not characterizing what it’s like to be below 200 percent of the poverty line,” Short said. “We don’t have any information to characterize what that would be like.”

Follow NBCLA for the latest LA news, events and entertainment: Twitter: @NBCLA// Facebook: NBCLA

More content from msnbc.com and NBC News:

Romney’s missing hard drives raise questions
Lingering shortage of ADHD drugs unravels lives
Buyer found for priciest bank-owned home
Reporter’s notebook: Return to Cuba