Framing the Health Care Reform Debate.

I’ve written a paper about how framing shapes the health care reform debate in the U.S. I think it shows the tip of the iceberg, from what is possible with ‘Digital Methods’, a very interesting approach, that uses software to research society. It also includes some ‘old-fashion’ opinion polls from Harris, with quiet disturbing results. Did know for example that 37 percent of adult Americans believe that “the proposed reforms would create panels that would decide who should live and who should die”?

After a widely applauded presidential campaign, enabled by grassroots and rhetoric, Barack Obama finds himself in a less comfortable position in the ongoing debate around the American health care reform. Linguist George Lakoff (2009a) argues that the decline of support on this issue is caused by an elaborate framing campaign of conservatives, and the lack of framing by the Obama administration. With the help of special software, I examined three spheres, the Web sphere, blogosphere and newssphere. This (limited) research mainly confirms Lakoff’s theories, but also show that Obama’s counter-frame is becoming more effective, which arguably led to mystification and misinformation (Castells, 2009).

You can find my paper here, or click on the read more button (I encountered some trouble with the layout in wordpress, there seems to be a big difference in some browsers)

Barack Obama is widely credited for running one of the best-organized campaigns in the history of political campaigning. His campaign activated millions of voters who were actively engaged around a message of ‘hope’ and ‘change’. Nevertheless since his big victory and inauguration in January 20th, 2009, according to opinion polls his ‘job approval’ decreased to under 50 percent. Especially in one of the most important topics of his campaign, health care reform, support for his plans decreased rapidly during this summer. “How is it possible that the same people who did so well in the campaign have done so badly on health care?” (Lakoff, 2009a).

Opinion polls[1] show that from February 2009 to approximately half June most Americans were in favor of Obama’s plan for health care reform. But since half June, support decreased to a declining majority. Next to the decline, is also a widespread misinformation about a broad spectrum of issues around health care reform[2].

Lakoff analyzed the problem very thoroughly on August 20th, 2009, just after Obama’s speech. According to Lakoff one of the biggest problems is the lack of framing from the Obama administration in almost all of its communication around health care reform. “The answer is simple and unfortunate: The president put both the conceptual framing and the messaging for his health care plan in the hands of policy wonks. This led to twin disasters.” (Lakoff, 2009a). The first disaster is what Lakoff calls ‘The PolicyList Disaster’. Instead of a general plan, the plan is generally communicated as a set of reforms. Instead of a story about how health it is focussed on all the different parts. Secondly the Obama administration primarily focussed on facts in almost all their communication, this is what Lakoff calls, “The PolicySpeak Disaster” (ibid). The idea that people can consciously control their reasoning, and therefore will automatically draw the right conclusion based on facts, is debunked by neuroscientist Antonio Damasio as described in his book Descartes Error (2004). On the other hand the conservatives are very good organized and run a lively campaign (together with lobbyists) against health care reform very effectively.

Since the idea of framing became popular around 2004, it has also been criticized. Rahm Emanuel for instance said, that the Bush campaign didn’t win two campaigns only because they used the right words. And Jesse Walker wrote a cogent critique that shows how hard it is to understand how framing works (Jenkins, 2008). Walker is arguing that Lakoff fell in his own trap, being a linguist and come with linguistic solution to the problem (Walker, 2008). I also acknowledge that in the past presidents as Bill Clinton didn’t succeed, and the Obama administration is closer to health care reform than ever before, even without ‘public option’. The goal of this paper is to get a better understanding of framing, not on the Obama administration’s strategy to get health care reform pass both congress and the senate. The research question is; how does framing, by both conservatives as progressives (in this case the Obama administration) reflect on the Web sphere, blogosphere and (online) newssphere and therefore on society in the case of health care reform?

First I will introduce the basic theory of framing based on research of George Lakoff, Antonio Damasio and sociologist Manuel Castells.  Secondly I will analyse the extensive theory of Lakoff of August 20th, 2009 on how the health care debate is framed. Lakoff also provides a list of terms he thinks are used by both sides, and gives a little theory per term how he thinks this will work out. I will translate this theory to my own research. In my method explanation I argue that researching the three Web spheres tell us something about the nature of framing, and the differences between spheres in the case of framing. With special software provided by Digital Methods Initiative this research is based on search queries in three different Google spheres; the Google Web Search, Google Blog Search and Google News. I researched how many times terms analysed by Lakoff are mentioned in those three spheres. Google’s search engines demarcated the three spheres for me.[3]

Theory of framing
In this chapter I will only briefly explain the basics of (political) framing, required to understand my research. For an extensive theory of framing I refer to Damasio (2003), Lakoff (2004, 2009a, 2009b, 2009c) and Castells (2007, 2009). Lakoff based himself on research done by linguistics, especially in the seventies[4], and recent to very recent neuroscientific research.[5] In their research there are three main aspects that are important for my research in this paper, which I will explain briefly.

First, as written in Damasio’s Descartes Error’ (originally from 1993), we, human beings, are not the rational ‘creatures’, we believe we are since the enlightenment (2003). In fact our brain cannot reason without emotion. Lakoff argues that framing is a natural phenomenon. Framing ensures the possibility of interpreting, and metaphors frame our understanding of the world. Damasio and Lakoff see the brains as physical connections (synapses) between billions of neurons. Those connections are made and strengthened by repetition. Basically there are two basic emotions by which the brain structure itself (via emotional pathways); epinephrine for ‘negative’ emotions and dopamine for ‘positive’ emotions (Damasio, 2003). Which leads to six basic emotions: fear, disgust, surprise, sadness, happiness and anger (Castells 2009). Lakoff describes six major errors in the brain; optimism bias, the fundamental attribution error, the illusion of control, reactive devaluation, risk aversion and the salient exemplar effect (Lakoff, 2009c). Framing therefore is making connections between parts of the brains mainly because of repetition, not only by words but also by images and especially stories. Stories are particular popular in politics, for instance the heroic stories of John McCain in Vietnam. Those stories are emotional narratives, making or strengthening connections in the brain (Lakoff 2009c).

Reagan (and later Bush) understood that despite most people didn’t share his view on almost all his issue’s; they were willing to vote for him. They run their campaigns on five principles;

  1. Values
  2. Clear communication
  3. Authenticity
  4. Trust
  5. The possibility to easily identify with him

Where Kerry and Al Gore ran their campaign on issues.

Secondly, to overrule a frame is very hard and almost always done with the wrong method. Most of the time politicians try to negate the frame, and overrule them with hard facts. Debunking a frame by negating is according to Lakoff just repeating the frame (2009b; 2009c). Lakoff believes that the Republicans are much better in political framing, mostly because they use their think tanks to come up with powerful frames. They have spent over four billion dollars, to repeat their worldview via their media, spokespeople and institutions (2008). Therefore the conservatives actively try to make the frames, and the progressives are normally reactive, and therefore ‘confirm’ those frames unintended. [6]

Thirdly Lakoff (2009c) and Castells (2009) theorize bipartisan is very common (they don’t provide numbers). Because we live in the same society, it is very normal to have both frameworks in your brain. The so-called mutual inhibition takes place when one part of the brain inhibits the other. Therefore, most people are not completely conservative or progressive. For instance, they are for the biggest part conservative but on some values progressive, the moderates (Lakoff, 2009b).

A very good example of how framing is used in practice is the case of the “War on Terror”. According to Amelia Arsenault and Manuel Castells there is still a big misperception around the Iraq War “Iraqi Freedom”. Take for example the now infamous weapons of mass destruction. The Waxman Report reported that the Bush administration made 237 false or misleading statements. Former president George W. Bush admitted on television that there were no weapons of mass destruction in Iraq (Castells & Arsenault, 2006). Still even in November 2008, 37 percent of the Americans believe that “Iraq had weapons of mass destruction when the U.S. invaded”. Only 52 percent thinks that this is not true.[7] Arsenault and Castells (2006) believe that this misperception is socially produced. “At the heart of this process of inducing misperceptions is the interplay between the political and communication establishments – in this particular case, between the Bush Administration and the mainstream media.” (Castells & Arsenault, 2006) (see also figure 1.). Castells concluded later; “In the process leading to the Iraq War, American citizens were submitted to the frames of the war on terror and patriotism” (2009: 186) (see also figure 2.). People tend to believe what they want to believe, they are not able to consciously step out of their frame. Again the frame is physical, and even to date the frame around “The War on Terror” is still very effective. As Lakoff analyzed (2009c) you cannot simply negate a frame, because you will be simply accused for being unpatriotic and naïve. It physically strengthens the frame and the synapses get stronger.[8]

Figure 1. “Social process of the production of misperceptions about the Iraq War.”(Castells & Arsenault 2006: 286).


Figure 2. Social production of mediated perceptions of the Iraq War. (Castells 2009)

Lakoff cofounded the progressive think-tank “Rockridge Institute”[9], which stopped at april 2008. Lakoff states that he achieved many goals; one of them is that the progressives now also know and adapt the theory of framing [10]. As I already briefly mentioned, Obama is applauded for using frames (unintended?), for instance the caring and empathy frame (2008).

Framing the Health Care Reform debate according to George Lakoff
The biggest flaw in the Obama’s administration’s mindset is that they believe they just have to tell the truth, according to Lakoff (2009c). The progressive say, ‘the truth is on our side’, let’s talk about the truth. That idea is fundamentally wrong according to Lakoff (ibid). You should tell the through, but the conservatives are not all lying on purpose, nor do you convince only with the truth. The first problem is that the administration communicates the policies separately (2009a). There is no real narrative, which covers it all. Therefore it is much easier to take down those points separately than to take down the plan as a whole. Legislators are lobbied to be against particular features, and it gets much easier to find the unfavorable plans (ibid).

Policy talk versus framed language.
Lakoff criticize important speeches by Obama and senior advisor David Axelrod for not mentioning ‘public option’ (a national-funded health care system) at all, at their speeches on August 16, 2009. We would later see that Obama strategically left public option out as an must have, and in December 2009 Nancy Pelosi states that public option will not be included in the final reform[11].

Nevertheless public option is the wrong term in the first place (Lakoff, 2009a). There are very well known stories about health care coverage, people who have to wait and insurance company’s are very inefficient. Instead the Obama administration communicates dry facts about costs and a policy list.  On the other hand the Republicans understand very well that health care reform debate is about morality, and use terms like ‘government take-over’ and ‘death-panels’. “Ask yourself which is more memorable: “Government takeover”, “socialized medicine” and “death panels” – or Axelrod’s 24 points?” (ibid).

Harris Polls in September 2009 show that a majority of 58 percent believe that President Obama’s plans would create a “government-run health care system”, and 22 percent is not sure.[12] In the Harris Polls of October 2009 numbers even increased, to 65 percent who believe in government takeover. The same poll under 2293 adults show that 37 percent believe that “the proposed reforms would create panels that would decide who should live and who should die”. Despite of the death panels 43 percent believe that “the system we have now is better than what the president is proposing”.[13]

My Method
Lakoff wrote this on August 20th, 2009, and thinks that is not too late, to change the game. Since then Obama held a speech at congress and a hardened campaign for his health care reform. Since August a lot is said about health care reform, and I believe the Obama administration (especially in Obama’s speech for congress in September 2009 and speech in January 2010) did a little better than before. I will do a comparative media analysis in the Web sphere, the blogosphere (via Google Blogs) and the online newssphere (via Google News).

Digital research methods; researching Google or society through Google?
According to Hewson et al (2003), the Web offers many possibilities as a research tool. Observational studies can be done, without disturbing subjects, and limit bias. Internet research is generally cheaper, and it is much easier to get enormous datasets. Internet research also offers possibilities to research to special-interest populations and minority groups (ibid).

Richard Rogers argues that social research cannot only be empowered with the Web, but can also be done on the Web (Rogers 2009). In its early days, the Web was seen as a separate space. Ideas of hyperspace and later cyberspace demarcated a division between the real and the virtual (ibid). Internet research would tell to researchers more about the cyberspace (for instance embodied in the numerous user studies), but not about society. Rogers suggests the ‘inauguration’ of a new era in Internet research, where one could not just study the Web with the Web, but also society (ibid, 8). The Digital Method Initiative calls it online ‘groundedness’, which arguably happened around 2007 (Rogers, 2009). Earlier studies were focussed on the digital divide (Castells 2004), and recent studies show that still a part of the world is (partly) excluded from the Web. Although a fast increasing part of the world population do have access to the Internet, in one way or another, online censorship is still a problem. Nevertheless I don’t think there is reason to believe that this is the case United States. According to a recent study of Nielsen, The United States has 195 million active Web users.[14] The OpenNet Initiative reported some Internet filtering incidents, but the United States is not technically filtering their Internet.[15]

Rogers was inspired by a journalist research done by the Dutch newspaper NRC Handelsblad in 2007 (2009). A journalist from NRC Handelsblad investigated the language of extremist right-wing groups, and used the Internet Archive [16] .The journalist compared the language of ‘regular’ right-wing groups and explored if the language was hardening over time.

“Thus the findings made about culture were grounded through an analysis of Web-sites. Most significantly, the online became the baseline against which one might judge a societal condition” (ibid, 9).

Researching the Web sphere?
This research is based on the same logics. With the help of tools offered by The Digital Methods Initiative, I investigate the Web sphere, newssphere and blogosphere to confirm or debunk Lakoff’s framing theory applied on the health care reform debate in the United States (2009a). The special software asks Google what kind of terms the different spheres adopt, in an automated process.

I argue, on the basis of research done by Hindman (2008) and Grimmelmann (2008/2009), that a Google query gives a good impression of the spheres, for two reasons. First, the web is vast; in 2008 Google had indexed a trillion unique pages.[17] It also lacks any kind of organized categorization (Grimmelman, 2008). Given the demise of the directories as the Google Directory (Rogers, 2008), it seems to be very hard to do well. According to Grimmelman, as a result “we rely on search engines’ rankings to make the Internet useful” (2008). When you search for information, you can go to websites you already know or find new sites. You can find new sites by links from the sites you already know, or use a search engine. It is very unlikely to find the information by ‘randomly’ typing addresses in your browser’s address-bar (ibid).

Search engines, especially in the Western World, are in most countries dominated by Google.[18] In the United States, according to Hitman, 71,57 percent of the searches is done through Google and 15,39 percent by Yahoo. Matthew Hindman did a comparative research on Yahoo and Google, using Hitman software (2008). The Hitman software combined data from multiple Internet Service Providers. The search engines (Yahoo and Google) shared 90 percent of their top 10 results in political queries and 4 of their top 5 (Hindman, 2008).

It is hard to say what percentage of incoming visitors comes directly from a search engine. For instance, a part of the visitors who directly came to a website, might have found that website through a search engine at the first time. For instance, the website of was ranked first in Google and second in Yahoo. It got 80 percent of their visitors from search engines (Hindman, 2008).

The second reason why I believe Google can give a good reflection of the Web sphere lies in Google’s algorithm. Google ranks their search results through a system what they call PageRank. Google counts the number of inlinks (the number of sites that have a link to a particular site[19]). It also weights the inlinks by popularity, a link from the CNN homepage gives you more value than a link from a website with PageRank 1[20]. It remains uncertain how this is exactly in practice. Although the stakes are high, and there is much speculation and analysis[21], the exact Google recipe is still unknown, and even changes over time. Google is to some sense a black box, but it provides some information about how webmasters can improve their search ranking.[22] Overall, sites that are ranked higher in the search results also have more links to their site on average. Yahoo and Google use a different algorithm and methods for crawling the web, their top 10 results are nearly the same, although slightly in a different order (Hindman 2008). Sometimes the results are gamed (Grimmelmann, 2008/2009), but I believe this is not the case with “health care reform”. To sum up; you can get a good impression of the Web sphere by a search-query in Google.

Personalization, my Google sphere might be different from your Google sphere.
Google is constantly tweaking their results, which makes it sometimes harder to compare results. Since December 4th 2009 Google started with personalizing their search engine results. Previously Google personalized only logged-in members with Google accounts. On the basis of a cookie (a little local file on the computer) Google connects you now with the databody they collected about your search preferences. Google use the word personalization for registered users and customized for ‘anonymous’ users. [23] However, it is possible to disable the customization. Google also tells you if and how they customized the search query.[24] . During this research I actively checked if Google personalized my results, and deleted my cookies before I checked one of the Google Spheres. Therefore, I believe my data is taken from the most neutral point of view possible. On the other hand, queries on “health care reform” returned by Google might be customized for other users. My Google spheres might be different from yours. For example if you regularly surf to the website of, CNN might be favored over other news sources. Google is not totally clear how it will work in practice (and use rather vague terminology in their explanation). The customization started in the beginning of December, I never encountered any customization, and very little is said about it since. Therefore, I don’t think it is that influential jet.  Nevertheless it might be a challenge for future research with Google.

Pro-health care websites are dominant in the Web sphere
The research conducted by Hindman (2008) showed a substantial overlap between Google and Yahoo. According to Hindman, this reflects a winner-takes-all pattern (ibid.: 68). Hindman’s research confirms usability-research, as Steven Krug’s Don’t Make Me Think (2000). Although the web is very vast, and Google has indexed approximately 40.500.000 websites for the query “health care reform” [25], only a few users get past the first page (the first 10 results) (Hindman 2008). Therefore, arguably, there is a dark or hidden web, with sites that are only found by a very small percentage. Those sites have a very limited influence on a debate. When they get more attention, they almost automatically get higher search rankings.

Hindman concludes that many users find the same information. Search engines make it much easier to find smaller websites, but make it even easier for users to return to known sources (ibid). As a result some pages barely show up high the results. Hindman shows that Pareto’s power law applies in his results. According to the power law, it is the top 20 that is responsible for 80 percent of the market (ibid). To conclude, I argue that to get heard it is especially important to appear in the top 10 results. Eye tracking studies reveal that users pay the most attention to the triangle in the upper left, the so-called golden triangle (Krug, 2000). In a Google specific eye tracking study people paid the most attention to first 3 results. People spent most time looking at the golden triangle, and then scanned the page in a F-shape, with a quick glance at the advertisements on the right.[26]

Cross-spherical analysis
The aim of this research is explicitly not to map all the political activity online in the debate around health care reform, nor is it the purpose to make claims about the Obama administrations strategy. This research is particularly aimed at getting a better understanding of how framing works, in this case in the three different spheres. Although the aim is not to examine the spheres as such, I would like to briefly explain my understanding of those spheres to contextualize this research.

Arguably, the three spheres are connected, and constantly adapt each other’s logic. Media imitate other media (Bolter and Grusin, 1999). But they are also in the same space of fast number of growing information on the one hand, and decline of attention in the other hand. As Rogers puts it; “They compete for inclusion as well as prominence in all manner of information spaces
” (Rogers, 2004).

There are quite contradictory theories about the blogosphere, and it seems to be a matter of perspective. Henry Jenkins, a professor in fan studies, see bloggers as actively (and critically) engaged. They also have a quite prominent voice in debates (2006). Lovink is quite critical and claims that although a very small percentage seems to be really engaged, the majority of the blogosphere provides a nihilistic voice (2007).  The newssphere is a sphere that combines the established broadcasters with some online broadcasters. And is editorially conducted by Google in this case. But it is different from traditional media because it is not ranked by ‘authority’, but chronologically. According to Lovink, the blogosphere is parasitic on the newssphere and drains it; the blogosphere is not an addition or alternative to the newssphere (2007).

But the differences between the spheres are not just a matter of perspective. They are not just concurring spheres on the Web but in a larger media landscape as well. As Katherine Hayles puts it; “The temptation to think of text on screen as essentially identical to text on a printed page, simply because the words are the same, is all the more the seductive because the computer is the most successful simulation machine ever created. It is crucially important, however, to recognize that the computer can simulate so successfully only because it differs profoundly from print in its physical properties and dynamics processes” (Hayles, 2004: 71). The more the context shapes to the medium, the more successful it gets.

Ultimately, Rogers sees cross-spherical analysis as “a digital method for measuring and learning from the distances between sources in different spheres on the Web (2009).

The methodology of this research mainly follows the logic and steps defined by ‘The Digital Methods Initiative’[27].

  1. Determine the basic query.
  2. Let Google demarcate the three different spheres on the topic of “health care reform”.
  3. Harvest the URL’s of the different spheres with the help of the Harvester
  4. Determine the terms to research
  5. Use the Google Scraper, Google News Scraper (in combination with the Google Scraper) and the Google Blog Scraper to scrape the blogs.
  6. Analyse the results

Step1, “Health care reform”
Because Google search engines demarcate the spheres; the basic query is very important. When choosing the wrong (biased) entry-point, the research as a whole is biased. The basic query in my research is “health care reform” because it is simply most used by both parties. Obama later used the term “health insurance reform”[28], and conservatives more often speak of “health care reform bill”. Some speak of “health reform”, but they all use the term “health care reform”. Google trends show that “health care reform” is by queried the most[29] (see figure 3), and the most neutral term.

Figure 3. Google trends on different terms

Step 2. Demarcating the spheres.
Following Hindman’s research (2008), I conduct my research over the top 3, top 10 and top 30 results. What I found in my research is the fact that the results of a query in Google (in all the different spheres) are different almost everyday. The Web sphere of December 29th, 2009 is a different sphere than the Web sphere of December 30th, 2009. I will later return to this issue, but therefore I have chosen to use the first 50 results instead of the first 30 results. I noticed that especially the top 20 to top 50 results constantly change. When choosing for the top 50 results, Google will combine the results from the same website, and place the result directly under the first result, in two different ways. Therefore the exact number of the first 50 results will not always be 50 unique websites.

When querying Google for the “health care reform” I always used the following settings;

  • Region: United States
  • Language: English
  • Results per page: 50
  • Not logged in with a Google account
  • I used the same computer with same ip address (a Dutch ip address)
  • The browser was Safari
  • Before I requested a query I always emptied my cookies and my browsing history.
  • Google never reported it customized my results in any way.

To get a better understanding of how framing works, we first have to analyse what kind of websites frame the Web sphere. Categorizing the top 50 results will give a better understanding of what voice is dominant. I categorized the websites manually as news or non-news, and labelled the non-news issues with pro health care, anti health care or neutral. If there was no (clear) statement for or against health care reform I labelled them as neutral. For instance some of the websites were quite vague about their opinion. For instance, “The United States Conference of Catholic Bishops” claimed to be pro-health care plans of Obama, but they were against abortion funding. Despite the official health care reform website of The Whitehouse clearly writes that abortion will not be funded by the health care plans.

The query on “Health Care Reform” at November 9th 2009 and November 13th 2009, as I described above gave me very remarkable results. As we can see in the Appendix (1,2 and 3) twenty-seven out of fifty were non-news sites. Eleven of those were clearly positive towards the Health Care Reform, thirteen were unclear or neutral, only two of them were clearly anti and one website didn’t respond.

When we take a closer at the top 10 ranking, three of the results are pro Health Care Reform. Two websites are neutral; one of them is the international Wikipedia lemma of Health Care Reform. The other neutral website is “The United States Conference of Catholic Bishops” website. There is one clip on YouTube, against Health Care Reform, they are now the 9th result, but they are rising. Four of the first ten results are news website, The New York times, CNN twice and CBS. I cannot mark them as pro or anti health care.

The GOP websites, and are not in the top 50 results. was ranked at the 66th place in the beginning of the week, and moved to the 88th place in the end of the week. The website is cannot be found in the top 500 at all. The websites of the Obama administration and are ranked first and second. Also is not present in the important top, although the second time I queried, the have a page on the 13th place. I’m not sure what causes the low ranking of the GOP websites, pagerank 5[30] is clearly not high, but there are multiple other websites in the top 50 with pagerank 4. Also the number of inlinks is not that high; 51[31], but it is higher then other websites in the top 50.

Although I used the exact same query and search engine settings, I got different results with only five days between the queries. Most of the big changes are news-websites, but in the top 10 results there are some other differences as well. The Wikipedia lemma was boosted from the 7th to the 4th place, the Bishops also climbed two places and the anti Health Care Reform Movie on YouTube made it to the top 10.

Because geo-location is very important for advertisement targeting through Google’s AdWords[32], it is hard to say if there is much advertisement on Health Care Reform. I found only one advertisement, the website of The Financial Times ( Nevertheless I presume there is more advertisement on such a frequently used search-term. Google AdSense is a program that centralizes advertisements for smaller websites. It enables website-owners to get advertisements on their sites, and therefore income. Because it is centralized, it also enables advertisers to put their advertisements on a lot of different websites. used Google AdSense to get their anti-Obama advertisement on websites that used terms around health care reform[33]. Further research is needed to see what is the result of advertisements like this.

How did the top 50 results evolve over time?
When we compare the research results from November 2009, to January 9th 2009, we can see some differences. The top 10 basically show a same pattern, but it now boosts their own News search engine. Above the results, a link refers to Google News and shows the first three results of the news sphere. At the first place in the results, the official webpage of the Whitehouse dedicated to the issue is still on the first place. But we now find a link to the Yahoo News service dedicated to health care reform on the second place, with the first story right beneath. On the third place we find Wikipedia, first on the international version of “Health care reform” and than on the “Health care reform in the United States”-lemma. The top 50 is now more balanced, with 7 anti websites and 9 pro health care reform websites.

One way or another, the pro-Health care reform voice (more or less) dominates the Web sphere, at least they represent a big percentage of the first results, especially the top 3 and top 10. You would say that their language and terminology would be more dominant too. But as we will see in the cross spherical analysis this is not the entirely the case. Over time we can conclude that news websites are increasingly important. I will later return to how the results evolve from November 9th, 2009 to January 9th, 2010.

Step 3. Harvest the URL’s.
By putting the source code of Google results from the spheres, the Harvester[34], the software extracts all the links from the spheres. By excluding links from Google and YouTube, and double links it extracts only useful links. Still, the URL set is imperfect; I will address those problems later.

Step 4. Query design, defining the right keywords
To determine a solid list of framing terminology, I follow Lakoff’s logic. Lakoff defined a clear list of terms used by both, conservatives [35] [36] [37] [38] and progressives, the Obama administration [39] [40] [41] [42] . He also defined clear rules how the progressives should reframe or now counter frame their terminology, to shape public opinion especially with the ‘truth’ (Lakoff, 2009). Next to the terminology he elaborately explained, I analysed some important communication since Lakoff’s article from August 20th, 2009. With special software[43] I triangulated the most important terms in their communication. With the help of Lakoff’s model I extracted additional terms manually from the most used terms in those communication. Together a list was conducted of 16 terms for both conservatives as the Obama administration.

Lakoff proposed to use a more engaging terminology (2009). The Obama administration should tell a very simple and powerful narrative, as he proposed;

Insurance company plans have failed to care for our people. They profit from denying care. Americans care about one another. An American plan is both the moral and practical alternative to provide care for our people.

The insurance companies are doing their worst, spreading lies in an attempt to maintain their profits and keep Americans from getting the care they so desperately need. You, our citizens, must be the heroes. Stand up, and speak up, for an American plan.

Insurance company plans have failed to care for our people. They profit from denying care. Americans care about one another. An American plan is both the moral and practical alternative to provide care for our people.

The insurance companies are doing their worst, spreading lies in an attempt to maintain their profits and keep Americans from getting the care they so desperately need. You, our citizens, must be the heroes. Stand up, and speak up, for an American plan. (ibid)

To sum up, according to Lakoff, Obama should focus on patriotism, and call the reform the American Plan, referring to the other plans, i.e. The Canadian Plan or Dutch Plan(ibid). It can be framed next to the care-frame Obama successfully rolled out during his 2008 campaign. He should also point directly at the “villains” (basically the insurance company’s with their lobbyists and conservatives that spread false information on purpose) Lakoff argues (2009). The terminology should come in a “common parlance” (ibid).

I conducted the following lists:

Obama administration:
“Reality check”
The Obama administration’s reality check website is aimed at “get the facts about the stability and security you get from health insurance reform”[44]. Lakoff argues that this won’t help the debate, because he doubts this will convince people (ibid). Nevertheless, who are these facts adopted by the spheres? It might not directly influence people, but it might indirectly influence them, by being an ‘official’ source. I can image journalist and bloggers use it.

“Public option”
According to Lakoff, public option is just boring language. “Yes it is public, and yes it is as option, but it does not fit to any frame (ibid).

The Obama Plan
Official websites of the Obama administration all use term The Obama Plan.

“Consumer protections”, “Insurance protections”,“Affordable coverage”, Uninsured
Secure, Healthier.
These words tap into the ‘care’ frame on the one hand, and on the urgency on the other hand. The words also more frequently used.

“Insurance company bureaucrats”, “Insurance company failure”, “Status Quo”, Lobbyists, Inefficient
As Lakoff argues, speak about the villains, tell people what goes wrong. Don’t only hint to it, but actually say it out loud.

“Donut Hole”
A fancy term for one of the major coverage gaps.

“Health insurance reform”
The official term of the reform by the Obama administration[45].

“Death panels”
Although none of any bills for health care reform, included even a hint to death panels, Sarah Palin ‘is afraid’ that when the bill passes through the House and the Senate, government will decide who will life and die.[46] The Obama administration tried to clarify on this issue by negating it. As mentioned earlier 37 percent of the Americans believe ‘death panel story’.

Abortion, Rationing, Euthanasia, “Gun regulation”
This falls in the same category as the death panels. While the Obama administration repeatedly told that abortion funding, euthanasia and health care rationing are not a part of health care reform[47], according to a Harris poll, 25 percent thinks euthanasia will be promoted, and 41 percent believes health care will be rationed.[48] Gun owner lobbies actively lobby against health care reform, because they are afraid it will lead to gun regulation.

This seems to be an extension of “HillaryCare” a conservative name for the failed attempt of the Clinton administration to reform health care.

According to Lakoff, the conservatives understand very well that health care is a matter of life and death (ibid).

“Government takeover”, Socialized, Socialist, Communist
These are not just lies, but tap into the deeper conservative frame. This frame is build since the 1970’s, and deeply fears anything near socialism. I personally, think this also illustrates ideological differences from Obama who tries to link patriotism to caring. I used both the term socialist as socialized, which is used in multiple terms, for instance socialized medicine.

Hitler, Nazi
Glenn Beck linked Nazism and Hitler to Obama.[49] How does this resonate in the different spheres?

“The liberal elite”, “Big Brother”, “Extreme left-wing”
Are all used multiple times in speeches.

It is remarkable to see that some of the terminology proposed by Lakoff, is now actually used by the Obama administration. On the other hand, public option is now sacrificed, and the plan is reframed to health insurance reform instead of American plan. Although he still tends to use a fact-based terminology, he, arguably, also use a more engaging terminology. It is interesting to see differences if there is difference in the adaption of more fact-based terminology and more story-based terminology.

It also struck me that there is less overlap in terminology than in the conservative terminology. The administration uses different terms for the same issues; this is the opposite of the “common parlance” Lakoff was talking about. The Obama administration also provides multiple facts over the number of un- or under-insuranced. It not only makes it harder to research, but also to remember. When applying Lakoff’s theory about framing, the Obama administration did better, and even adopted some of the terms Lakoff proposes. The bottleneck according to Lakoff’s theory is inconsequent terminology and the lack of patriotic language. On the other hand you might argue that for Obama patriotism is caring in the first place.

Conservatives on the other hand are more supported by business logics. According to Lakoff, they got their tactics from marketing theorist and may not directly know about brain research and linguistics. The conservative message does not come directly from their leaders, partly because they don’t have real leaders at the time. The republicans, according to a study of Harris, see McCain as the most influential leader[50]. In fact, McCain even “admitted” that Obama is not a socialist. Again, negating a frame only makes it stronger, and is totally different from debunking it. And although Sarah Palin speaks of death panels, and Glenn Beck compares Obama with Hitler, most of the language came from what Lakoff calls, the five-twenty-sevens (2008). Therefore I enhanced the conservatives list provided by Lakoff only with a few terms.

Some terms are used by both ideologies. For instance, bankrupt, which is used by conservatives as nationwide bankrupt. Obama on the other hand, use it as ‘personal’ bankrupt. With this method it is problematic to contextualize, and therefore I can only use clearly defined terms.

Step 5. Scraping the spheres.
The Google Scraper is server sided software, developed by the Digital Methods Initiative. In fact it uses a feature of Google, and automates analysing its results. By Google typing a term and than “site:” the web address, for instance [Obama”] in Google, Google counts the number of hits on the website. By using a specific website and not only the domain address, for instance you can search if a term is used. When you only define the domain, Google will search the whole domain. It is particularly important for the Web sphere and blogosphere to keep this in mind. The Google search engine (for the Web), seems to prefer domains over specific websites. I speculate this is mainly due to its algorithm, that is (partly) based on inlinks, and people seem to link to domain instead of specific sites. Google Blog Search has a simular problem, because the settings of some blog software, it sometimes links to the homepage of the blog. Because of its different design, Google News Search does not have this problem. Google cannot handle question marks; it breaks the url after it. Some software use question marks, especially Content Management Systems, which are quite popular for news sites, and also some blog software. This means that for instance ‘’ will search the whole domain for the requested term. Luckily the most up to date software often redirects via its htaccess-file, which solves this problem when done well. To sum-up, when using the Google Scraper (or Google Blog Scraper or News Scraper) you have to be aware of these issues while analysing the results.

Step 6. The results
The results are divided in three parts, first a sphere comparison, second a term analysis and third a time comparison.






Sphere comparison.
The Web sphere[51]


Issue cloud – issues for all sources (hosts, cumulative, retrieved by Google scraper)

“Reality check” (11) “Public option” (31) “The Obama Plan” (13) “Consumer protections” (12) “Insurance protections” (4) “affordable coverage” (17) “insurance company bureaucrats ” (5) “Uninsured” (36) “Secure” (17) “Healthier” (19) “Status Quo” (16) “Donut Hole” (10) “inefficient” (15) “insurance company failure” (1) “health insurance reform” (22) “lobbyists” (22) “Death panels” (14) “Obamacare” (16) “Pro-life” (9) “Government takeover” (14) “Abortion” (25) “Socialized” (17) “Socialist” (13) “Communist” (7) “Euthanasia” (11) “Big Brother” (6) “Hitler” (11) “Nazi” (9) “The liberal elite” (3) “Extreme left-wing” (2) “Gun regulation” (2) “rationing” (16)
The blogosphere[52]

Issue cloud – issues for all sources (hosts, cumulative, retrieved by Google scraper)

“Reality check” (4) “Public option” (10) “The Obama Plan” (4) “Consumer protections” (5) “Insurance protections” (3) “affordable coverage” (5) “insurance company bureaucrats ” (0) “Uninsured” (14) “Secure” (5) “Healthier” (5) “Status Quo” (4) “Donut Hole” (4) “inefficient” (5) “insurance company failure” (0) “health insurance reform” (6) “lobbyists” (5) “Death panels” (6) “Obamacare” (10) “Pro-life” (4) “Government takeover” (5) “Abortion” (10) “Socialized” (5) “Socialist” (7) “Communist” (2) “Euthanasia” (3) “Big Brother” (2) “Hitler” (4) “Nazi” (4) “The liberal elite” (0) “Extreme left-wing” (0) “Gun regulation” (0) “rationing” (6)

The News Sphere[53]

Issue cloud – issues for all sources (hosts, cumulative, retrieved by Google scraper)

“Reality check” (10) “Public option” (19) “The Obama Plan” (8) “Consumer protections” (8) “Insurance protections” (3) “affordable coverage” (13) “insurance company bureaucrats ” (13) “Uninsured” (28) “Secure” (14) “Healthier” (10) “Status Quo” (11) “Donut Hole” (6) “inefficient” (10) “insurance company failure” (0) “health insurance reform” (11) “lobbyists” (13) “Death panels” (8) “Obamacare” (14) “Pro-life” (10) “Government takeover” (10) “Abortion” (15) “Socialized” (10) “Socialist” (14) “Communist” (10) “Euthanasia” (9) “Big Brother” (10) “Hitler” (10) “Nazi” (10) “The liberal elite” (1) “Extreme left-wing” (1) “Gun regulation” (1) “rationing” (13)

When can conclude that at January 9th, 2010 the terminology used by both the Obama administration and conservatives, is almost equally used in the three different spheres, with a little edge for Obama. The language of the Obama administration is slightly more used in the Web sphere. On the other, as mentioned earlier, the Web sphere is dominated by pro-Health Care websites. These results clearly show that the pro Health care sites adopt some of the terminology of the conservatives and simply negate it. The anti health care website are more careful with adapting the Obama administrations terminology but are simply outnumbered. The ‘neutral’ websites adopt but terminologies.

It is quite remarkable that the News Sphere, filled by professional journalists also adapts most of the language of both sides, although it slightly prefers the Obama administration’s language. This is mainly because it quotes both conservatives and progressives. Terms as “Hitler”, “Government Takeover” and “Death Panels” are almost equally as less radical terms. Notice that Fox News, not appeared in the first 50 results, and therefore is not included for example. In the time comparison this issues is addressed more elaborately.

The blogosphere show a somewhat same pattern as the newssphere and Web sphere  in general, but that seems to follow the general trend of hot topics. The blogosphere is from the three spheres the most focussed on one side of the story. A lot of the blogs adapt only terminology from the Obama administration or conservatives. It is also interesting that the blog en large simply do not adapt any, or one of the terms. Interestingly enough especially the pro Health care reform websites use most of the terminology, by both sides.

Term analysis
The results confirm Lakoff’s framing theory in general, but with a few exceptions. The terms that are most powerfully connected to a frame are used the most, especially around the care-frame of Obama. Topics as abortion and death panels are still very lively in the all the spheres. The Obama administrations ‘villainizing’ of the industry seems to effective, at least the words lobbyists and ineffective, but it less effective on the other terms. Lakoff’s theory that the ‘real’ debat is not about issues but on values is also confirmed by this research, with the exception of Public Option. This is mainly because Public Option is sacrificed in the reform, which is the topic of on ongoing debate.

Time comparison
One of the major problems with a time comparison using the Google Scraper is that Google does not provide search results from the past. Therefore it can be done by manually save the results, which makes it harder to make a solid comparison. I started this research in the beginning of November, and only got limited data (on a limited number of keywords) from November and December 2009 and January 2010. Therefore we can only see the end of the trend, but it provides some useful insights.

The data is from; November 2009; 27th [54] [55] [56] ,December 2009; 26th [57] (blogosphere) [58] (newsphere) [59] (Web sphere), 27th [60] [61] [62], 29th [63] [64] [65], 30th [66] [67] [68], 31st [69] [70] [71], and January 2010, 1st [72] [73] [74], 9th [75] [76] [77]. The data-set is to tiny to make general claims about trends, but it can tell us more about the data previously analysed. The data basically confirms the theory by Lakoff (2009a), and my analysis above.

Still there are some interesting insights. First of all, although I used a smaller group of terms, the conservative terms are slightly more adapted in November. But after the bill passes on the 25th of December, you see the language changing; the spheres even adopt the official terms as “Patient Protection and Affordable Care Act”, as to expect especially in the news sphere.

The blogosphere is compared over time the most pro-health care reform, at least they adopt most of the language of the Obama administration. On the other hand the also adapted language like Hitler and Socialism more easily than the newssphere. The newssphere is also more clustered around a less terms, but the terms change over time. The Web sphere got more dominated by news and especially blogs over time. In November the Web sphere tended to link more to official sources. While in December and even more in January they link to news-sources and especially more to blogs. This indicates that blogosphere gets more dominant in the debate.

On the basis of my (limited) results, we can say that the conservative frame has been very effective, and still resonates in the three spheres. The fact that the pro Health care reform sites outnumber the anti health care sites in the most important Web results in November 2009, does not result in a lack of conservative terminology. In January 2010, there are more anti health care websites, still the pro health care are the majority. On the basis of this research we can also clearly see that the Obama administration adept a more powerful frame, that terminology is now even more used than the conservative terms. The blogosphere seemed to adept the least of the terminology. Although there are differences between the spheres, they all show the same pattern; the language of both sides is reflected on the all spheres. They all use the administrations language a little more.

Still, one could argue, that this is not directly reflected in the opinion polls (see figure 6). This is not only reflected by the number oppose or favor the health care reform plans, but especially with misinformation. Castells analysed that once a powerful frame is established, it is very hard to overwrite it. The counter frame leads to misinformation and mystification (Castells, 2009). The socially produced misperception of the War on Terror is a powerful example. Those misinformation strongly correlated with their opinion about “the war” (ibid).

Figure 6. A summary of the official opinion polls on health care reform.

More research is needed to get a better understanding of framing in the spheres, especially non-text based. In fact framing is not word-based, but exists of mental images (Lakoff, 2008). Jenkins (2009) and Duncombe (2007; 2009) share Lakoff’s theory (2009a), but from a fanstudies perpective. This is for instance embodied in the socialist Joker, as analysed by Whitney Phillips (Jenkins 2009).


Castells, Manuel (2009), Communication Power. Oxford: Oxford University Press.

Castells, Manuel and Amelia Arsenault. (2006), Conquering the Mind, Conquering Iraq. < > accessed at November 10, 2009

Damasio, Antonio. (2003). Descartes’ Error. Pinguin

Grimmelmann, James (2008/2009), The Google Dilemma. New York: Law School Law Review, 53: 939 – 950.

Hayles, Katherine. N (2004), Print is flat, code is deep: The importance of media-specific analysis. Poetics Today. 25 (1): 67-90

Hewson et al. (2003), Internet research methods: a practical guide for the social and behavioural sciences. London: Sage, 26-55

Hindman, Matthew (2008), The Myth of Digital Democracy. Princeton: Princeton University Press, chapters 3-4

Jenkins, Henry (2006), Fans, Bloggers and Gamers: Exploring Participatory Culture. New York: New York University Press

— (2009), Unmasking the Joker. < >, accessed at November 29, 2009

Lakoff, George (2004), Don’t think of an Elephant. New York: Penguin

— (2008) George Lakoff on the Political Mind < >, accessed at December 26, 2009

— (2009a), The PolicySpeak Disaster for Health Care, < > accessed at November 2, 2009

— (2009b) Politics of Language, < >, accessed at December 26, 2009

— (2009c), The Political Mind. New York: Penguin

Lovink, Geert. (2007) Zero Comments: blogging and critical Internet culture. New York: Routledge

Rogers, Richard (2004), Information Politics on the Web.  London: The MIT Press

— (2008), < > accessed at December 26, 2009

— (2009), The End of the Virtual. Amsterdam: Amsterdam University Press

Appendix 1 complete list

From left to right, ranking in the search query on 13-11-2009[78] at 15.00 (ranking in the search query on 9-11-2009 at 8:29[79]), url, source (News or Non-News), category (Pro, neutral or anti Health Care Reform). Please note that the search query on 9-11-2009 only included the first 40 results.

1 (1) Non-news Pro
2 (2) Non-news Pro
3 (3)…/health/diseasesconditionsandhealthtopics/health…care/ health_care_reform/index.html News
4 (7) Non-news Neutral
5 (4)…health.reform…/index.html News
6 (5) News
7 (9) Non-news Neutral
8 (10) News
9 (13) Non-news Anti
10 (11) Non-news Pro
11 (12) Non-news Neutral
12 (-) News
13 (-)…/health-care-reform-funded-billions-penalties-uninsured News
14 (30) Non-news Neutral
15 (-)…/house-passes-health-care-reform/ Magazine
16 (15) Non-news
17 (21)…/final-senate-panel-approv_n_318921.html Blogs
18 (16) Non-news Anti
19 (18) Non-news Anti
20 (17) News
21 (19) Non-news Pro
22 (30) Non-news
23 (25)…/healthcare/…/health_care_reform Non-news Neutral
24 (20) Non-news Pro
25 (24) Blog
26 (26) Non-news Pro
27 (-) Non-news 404
28 (-) Non-news Pro
29 (-)…/the_idiots_guide_to_health_car.html Magazine
30 (23) News
31 (31) Non-news Pro
32 (28)…/SB10001424052970203946904574300482236378974. Html News
33 (34) Magazine
34 (-) Non-news Neutral
35 (-)…/tampa-town-hall-on-health_n_253478. Html Blog
36 (35) Non-news Neutral
37 (-) Blog
38 (38) News
39 (39)…Health…Health+Care+Reform… Non-news Neutral
40 (-)…/b4155030836539.htm Magazine
41 (-) Magazine
42 (33) Non-news Neutral
43 (27) Non-news Neutral
44 (-) News
45 (27) Blog
46 (-)…health_care…/health-care-reform-lite-.html Blog
47 (-) Non-news Pro
48 (-) Blog
49 (-)…/health_financing.html Non-news Pro
50 (-) Non-news Pro
88 (66) Non-news Anti


Appendix 2. News sources

From left to right, ranking in the search query on 13-11-2009 (ranking in the search query on 9-11-2009), url, source, number of pages in the top 50.

3 (3)…/health/diseasesconditionsandhealthtopics/health…care/ health_care_reform/index.html News 2
5 (4)…health.reform…/index.html News 2
8 (10) News
13 (-)…/health-care-reform-funded-billions-penalties-uninsured News 1
15 (-)…/house-passes-health-care-reform/ Magazine 1
20 (17) News 2
25 (24) Blog 1
29 (-)…/the_idiots_guide_to_health_car.html Magazine 1
30 (23) News 1
32 (28)…/SB10001424052970203946904574300482236378974. Html News 1
33 (34) Magazine 1
35 (-)…/tampa-town-hall-on-health_n_253478. Html Blog 1
37 (-) Blog 1
40 (-)…/b4155030836539.htm Magazine 1
41 (-) Magazine 1
44 (-) News 1
45 (27) Blog 1
46 (-)…health_care…/health-care-reform-lite-.html Blog 1
48 (-) Blog 1

Appendix 3 Non-news sources

From left to right, ranking in the search query on 13-11-2009 at 15.00 (ranking in the search query on 9-11-2009 at 8:29), url, category (Pro, neutral or anti Health Care Reform).

1 (1) Pro
2 (2) Pro
4 (7) Neutral
7 (9) Neutral
9 (13) Anti
10 (11) Pro
11 (12) Neutral
14 (30) Neutral
16 (15) Neutral (international)
18 (16) Anti
19 (18) Anti
21 (19) Pro
22 (30) Neutral
23 (25)…/healthcare/…/health_care_reform Neutral
24 (20) Pro
26 (26) Pro
27 (-) 404
28 (-) Pro
31 (31) Pro
34 (-) Neutral
36 (35) Neutral
39 (39)…Health…Health+Care+Reform… Neutral
42 (33) Neutral
43 (27) Neutral
47 (-) Pro
49 (-)…/health_financing.html Pro
50 (-) Pro
66 (88) Anti

Appendix 4 Time comparison

From left to right, ranking in the search query on 9-1-2010 (ranking in the search query on 13-11-2009), url, category (Pro, neutral or anti Health Care Reform).

1 (1) Pro
2 (-) News
3 (-) News
4 (3) Neutral
5 (-) Neutral
6 (2) Pro
7 (-) News
8 (-) News
9 (-) News
10 (-) News
11 (47) Pro
12 (10) Pro
13  (36) Pro
14 (-) Pro
15 (-) Neutral
16 (-) Anti
17 (-) Anti
18 (-) Neutral
19 (-) News
20 (-) News
21 (-) News
22 (-) News
23 (-) News
24 (-) Pro
25 (9) Anti
26 (18) Anti
27 (19) Anti
28 (-) Pro
29 (-) Neutral
30 (-) Pro
31 (-) Anti
32 (40) Magazine
33 (21) Pro
34 (-) Anti
35 (-) Neutral
36 (-) Pro
37 (-) Anti
38 (-),false,false,n,n,n:null; News
39 (-) Pro
40 (-) News
41 (-) Neutral
42 (-) Neutral
43 (-) Neutral
44 (-) Neutral
45 (26) Pro
46 (-) News
47 (-) News
48 (-) Neutral
49 (-) Pro

[1] accessed at January 9th, 2010

[2] accessed at January 9th, 2010

[3] Unfortunately, there are some technical limitations, which I will further discuss in this paper.

[4] For example theory from Charles Fillmore

[5] I should mention that they are all employees to the University of South California, and closely connected in their theories. Castells’ knowledge is mainly based on that of his colleagues but from a different perspective (2009: 137)

[6] Obama seemed to realize this, when he countered critique about Reverend Wright with a speech on race.

[7] accessed at January 9th 2010

[8] Obama refused to use this term, which led to critique of Dick Cheney. Political analysts seem to favor Obama’s approach, but there is no opinion poll result jet.











[19] You can check the number of inlinks by quering link: *webadres* in Google

[20] PageRank 10 is for the sites with most inlinks and PageRank 0 is for the sites with the least inlinks





[25], accessed January 7, 2009





[30] (accessed 13-11-2009 10:37)

[31] (accessed 13-11-2009 10:35)















































[78] (accessed at November 13, 2009 8:29)

[79] (accessed at November 9, 2009 15:00)

1 comment Write a comment

  1. Pingback: Review: U draait en bent niet eerlijk | Han Rusman

Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Comment Spam Protection by WP-SpamFree