About Obama Poll Watch

Obama Poll Watch

ObamaPollWatch.com is a sub-site of our main page here at ChrisWeigant.com. It was created a few months into Barack Obama's term as president, for two reasons. The first was to track Obama's performance -- as measured by public opinion polling -- for the rest of his presidency. The second was to give us something to do until the next presidential election cycle rolled around, because we had so much fun with the 2008 Electoral Math series of articles, which tracked the polls state-by-state during the exhaustive campaign season. In other words, Obama Poll Watch would give us something to occupy our time here, and keep us off the streets, as it were.

As with most of our column series at the site, this was a sort of organic growth, spurred on by one column with a fairly good idea, which took on a life of its own over time. Which is where we find ourselves now.


Source Data

There are a few reputable websites out there who provide polling data to the general public. My favorite, on a personal level, is probably Pollster.com, who used to be an independent site, but has since moved around in the media world until it found a home at The Huffington Post (as my humble columns also have, I should mention for full disclosure reasons). But here at ObamaPollWatch.com we don't discriminate, and have provided a prominent link to two different data sites on every one of our pages (at the top left of our page layout).

Pollster has a flaw in the way they present their data (in my opinion), which is why I do not use it for our source data here. This flaw is a lack of hard data on a day-to-day basis. So I turned to RealClearPolitics.com instead. They have a page which tracks public opinion polling and posts a discrete number each day for presidential job approval and disapproval. This is easier to plug into my own admittedly amateur poll-tracking attempts, which is why I decided to use their data for our raw source data here at ObamaPollWatch.com.

Feel free to view RealClearPolitics' data, to double-check my own figuring, on their presidential job approval rating tracking page.



Our basic polling methodology is to smooth things out as much as possible, in order to discount what are known as "outlier" polls (or those which show results wildly out of whack with all other polls). This is accomplished by a two-part smoothing process, which we humbly call our "poll of polls of polls," in the hopes that such a cool phrase will eventually go "viral" and everyone will start using it.


The first phase of this smoothing takes place at RealClearPolitics.com, in the way they come up with their daily numbers. Their numbers are an average of the most recent polling, as they define it. Without knowing their exact methods, RCP seems to use the polls from approximately the past two weeks' timeline, and only allows the "daily" polls (Rasmussen and Gallup) one solid number per week. That may be an oversimplification of their methodology, but it's the best I can do without access to their actual algorithms.

RealClearPolitics.com thus comes up with a daily number for both "Obama job approval" and "Obama job disapproval" and posts it. We here at ObamaPollWatch.com take those numbers as daily data points, and chart them. Over the course of a month, we average out these numbers to a single data point for the month. This further smoothes the data, but also introduces somewhat of a delay in showing any sharp movements in public opinion.

There are, to be sure, problems with any methodology of poll-tracking. Including Rasmussen's polls, for instance, noticeably skews things, since it has been shown that they exhibit a definite Republican bias. Furthermore, all polling companies are currently struggling with the problem of cell phones. Landline-only polls are also showing about a five-percent skew Republican from people's actual opinions out there, due to the increasing number (over 25% among the younger folks) of people who have no traditional "landline" phones installed in their residence. Most reputable polling companies are trying hard to adjust to this new reality, but federal law currently requires all cell phone polling calls to be hand-dialed, which slows things down (and costs more money) for the polling organizations, so it remains to be seen how much of a disconnect this will prove to be in the near future.


Historical Data

On the ObamaPollWatch.com site, we provide comparison data from past presidents, reaching back to the earliest authoritative public opinion polling data available to the public.

Practically, this means that for all presidential data before George W. Bush, we rely totally on Gallup polling data available at the University of Connecticut's Roper Center site.

This data is incomplete, I admit. There are gaps. We are investigating the reason for these gaps, and hope to present more complete data in the near future here. Eventually, we hope to present comparison charts for Obama versus all presidents back to Franklin D. Roosevelt, who happened to be president when public opinion polling "came into its own" both for scientific statistical reasons and for the fact that by F.D.R.'s time, most people had telephones in their homes.

This site is a work in progress, and will hopefully continue to improve over time. We thank you for your patience in the meantime.