The western perspective in Yahoo! News and Google News

Abstract

A large-scale quantitative analysis of geographic news coverage of Yahoo! News and Google News was conducted in the US and India using a content-analysis program to examine overrepresentation of developed countries in the US-based news portal sites. The results of the analysis show an absence of difference in the news coverage patterns of Yahoo! News and Google News between the US and India, which was predicted by the cultural imperialism thesis. However, it was found that Yahoo! News relies on western news agencies for news feeds, so it represents more developed countries; Google News has hundreds of news sources across the world, but its news coverage is highly concentrated and volatile. These findings can be explained by two ways by which western news agencies exercise their global influence on online news: news provision and agenda-setting.

Keywords

Agenda-setting computer-assisted content analysis Google News India international news Japan earthquake news agencies Yahoo! News

Before the Internet was introduced, foreign news in newspapers and on TV were the main, or almost the only, source of information about foreign countries. Foreign news had to be distributed through news agencies across the world, so the international news flows were dominated by western international news agencies (e.g., Reuters, AFP, and AP), and western viewpoints were overrepresented in international news. However, since the advent of the Internet, western international news agencies have become unnecessary for international news distribution because news organizations can distribute their news items to audiences all over the world directly through the Internet. In recent years, news organizations in non-eestern countries (e.g., Al-Jazeera) have been gaining online audiences even in western countries and rapidly increasing their presence in the international news market.

Nevertheless, despite the emergence of non-western news organizations, in the context of cultural imperialism, there are concerns about the global dominance of western media and communications companies. It is claimed that transnational western media and communications companies distribute media contents that justify and maintain their dominance (Schiller, 1991), and that their media contents reflect the interests of western companies and governments (Herman, 2000). In addition to its dominance in the media and communications industry, the US has a great advantage in information and communication technologies (ICT). Boyd-Barrett (2006) states that US media companies can utilize the most advanced technologies to bridge their advantages in traditional media to their advantages in digital media.

This research aims to examine the overrepresentation of western countries in online news services with particular interest in the global dominance of US-based ICT companies. Attention is directed to international news environments not only inside but also outside the US, especially in non-western countries. I compare geographic news coverage by news portal sites owned by US-based companies Yahoo! News and Google News in the US and India using data derived from a computer-assisted content analysis of more than 65,000 news items collected from the news portal sites and online newspapers; geographic news coverage by the news portal sites are compared with online news services operated by local newspaper companies in the respective countries.

Literature review

International news plays very important roles in our understanding of the world (Hachten et al., 1999; Van Ginneken, 1998; Wanta and Hu, 1993). People usually do not have direct experience of foreign events, so international issues, such as conflicts and natural disasters, can only be recognized through international news; however, international news usually underrepresents developing countries. This is because international news agencies distribute more news about western countries and because the editors of national news organizations judge the importance of foreign news events based on the items provided by international news agencies. The editorial process in national news organizations relies heavily on western international news agencies, and thus, the coverage of news in the international news section reflects the western-biased coverage (Boyd-Barrett, 1980). Boyd-Barrett says ‘less than two dozen newspapers around the entire world could make a reasonable claim to independence in the gathering of a comprehensive international news file’ (Boyd-Barrett, 1980: 15). In fact, Galtung and Ruge (1965) found that 87% of the articles about the Congo and Cuban crises in 1960 and the Cyprus crisis in 1964 in Norwegian newspapers were provided by the major news agencies, AP, UPI, Reuter, and AFP. Salamore (in Boyd-Barrett, 1980) conducted a content analysis of the top newspapers in India, Kenya, Lebanon, Japan, and Norway every month between 1961 and 1968, and more than half of the international news items were found to be from the four major news agencies. Matta’s content analysis of 16 daily newspapers in 14 Latin American countries in 1975 showed that 80% of the foreign news items were from these four news agencies (in UNESCO, 1985). Schramm (1978) performed a content analysis of articles about non-Asian third world countries in 14 Asian newspapers in 1977 and found that three-quarters of the news was, again, provided by these four news agencies.

UNESCO’s research in 1979 on newspapers, radio, and TV news in 29 countries showed that the most important determinants of international news coverage are metropolitan centrality, geographical proximity, and former colonial orientation (UNESCO, 1985). Apart from own region, the most covered regions are Western Europe and North America in all of the countries. Chang et al. (2000) conducted a content analysis of TV news in China, Indonesia, Malaysia, the Philippines, and Singapore. They found that in all of these countries, coverage of US news was remarkably high, and they concluded that news coverage follows the position (core/semi-periphery/periphery) in the world system. Sreberny-Mohammadi and Stevenson (Wu, 2000) conducted an extensive international study on international news coverage in 1995; Wu conducted statistical analysis of the data and found that the top 10 most covered countries in newspapers are the US, France, the UK, Russia, Bosnia, China, Germany, Italy, Japan, and Spain, and that the primary predictor of international news coverage is a trade relationship between countries, while the secondary predictor is the existence of an international news agency office.

However, the growing numbers of online newspaper readers¹ created the expectation that online news services would change the news coverage of international news (Berger, 2009). The expectation around the role of the Internet in news delivery and consumption has its roots in the two technological characteristics of the Internet: its capability to communicate with distant places at a low cost and its virtually unlimited capacity to submit content. The former characteristic helps news organizations to deliver news directly to overseas audiences. The latter characteristic eliminates space and time constraints in newspapers and TV programs, and allows news organizations to submit more stories about underreported issues (Reese et al., 2007; Stevenson, 1994).

During the last 10 years, many research studies on online news outlets and international news have been carried out to examine the impact of the Internet on international news coverage (e.g., Himelboim et al., 2010; Paterson, 2005; Wu, 2007). It was expected that online news outlets would be able to carry more international news than offline versions (Stevenson, 1994), but only a small difference in international news coverage was found between offline and online news. Paterson (2005) compared the dependence of the news services of four portal sites (AOL, Yahoo!, Excite, and AltaVista) and five news organizations (MSNBC, CNN, BBC, Sky, and The New York Times) on news agencies (Reuters, AP, and AFP) for international news between 2001 and 2006. He found that there was a significant increase in dependence on news agencies for international news during the study period in almost all the online news services. Wu (2007) conducted research on online news services and the traditional news outlets of CNN and The New York Times and found that the presence of a news agency is a stronger determinant of geographic news coverage in online news than in traditional news outlets. He also found a greater representation of developed countries than developing countries in the online news services of The New York Times and CNN. Middle Eastern countries were more often reported on online news services than print media, but ‘the international news output from the online media does not seem to deviate much from that of their traditional counterparts’ (Wu, 2007: 549).

Van der Wurff (2008) questioned why a large amount of news content online is the same as that in the traditional media and why only a small amount of news content is exclusive to the Internet outlets. He explained that the Internet lowered only the cost of distribution of content but not the cost of production of content, so content providers are not motivated to invest in the creation of original content for the Internet because content is distributed online free of charge. In his previous research on newspapers and online news services in 18 European countries in 2003, Van der Wurff found that, on average, 70% of the most important online news articles are identical to the articles published in printed newspapers (Van der Wurff, 2005).

However, it is too early to conclude that the Internet has only a small impact on international news environments because Wu (2007) and Himelboim et al. (2010) noted a limitation of their research in that they only examined the online services of traditional news organizations, newspapers, and TV news broadcasts. They pointed out the need for studies on popular non-traditional online news services, such as Yahoo! News and Google News.

Yahoo! News and Google News deserve our attention not only because they have a large audience but also because they have three important characteristics. First, neither Yahoo! nor Google is a news organization, so they do not produce (almost any) news items themselves. They instead redistribute news items produced by other news organizations, which are not necessarily limited to traditional news organizations, and they can offer different news items from non-traditional news organizations to their audience. Second, Yahoo! News and Google News directly compete against the news services of traditional news organizations, and they draw financial resources for news production away from traditional news organizations, so they possibly undermine traditional journalism. Third, capitalizing on the international character of the Internet, Yahoo! and Google operate transnationally and develop services targeted toward overseas markets, so the influence of their services is not limited to their home country, the US. Therefore, on the one hand, Yahoo! News and Google News nurture the expectations of change in our news environment; on the other hand, they invoke concerns about the deterioration of traditional journalism and the global dominance of US-based companies on the Internet.

Considering the three characteristics of Yahoo! News and Google News, there are two contradictory forces among contemporary international news environments that I want to focus on: the diversification of online news by online news services and the reinforcement of the global dominance of US-based companies. The conflation of these two contradictory forces may be leading to a complicated consequence for online news environments. On the one hand, because online news services are essentially different from traditional ones, they might be able to represent more underreported countries or regions in international news; on the other hand, because they are operated by US-based transnational companies, they distribute western news items in other countries regardless of the needs of the local market, so the changes in international news coverage created by Yahoo! News and Google News are smaller in non-western countries. In other words, Yahoo! News and Google News can represent underreported countries or regions more than traditional news outlets can, but they do not do so in non-western countries as much as they do in western countries because of their western bias. Accordingly, the research question is as follows:

RQ: How much are western countries represented in the news portal sites owned by US-based companies in non-western developing countries?

Methodology

This research is based on the content analysis of news items submitted to major online news services. Content analysis was conducted by focusing on the geographic (country or region) coverage of the news, and then coverage patterns were compared between online news services operated by newspaper publishers (hereinafter referred to as ‘newspaper sites’) and Yahoo! News and Google News (hereinafter referred to as ‘news portal sites’) both in a western developed country and a non-western developing country.

Hypotheses

Given space limitations for news items on both newspaper sites and news portal sites, they possibly offer more news about developing countries. Yet, as Van der Wurff (2008) explains, newspaper publishers often do not invest more resources to gain more news items, so geographic news coverage on newspaper sites is relatively narrower than on news portal sites that can be provided with news by a diversity of news sources. Accordingly, it is expected that these news portal sites cover developing countries more than newspaper sites do.

H1: News portal sites represent more developing countries than newspapers sites both in the western countries and non-western countries.

However, considering the western bias of US-based media companies, it is expected that news portal sites in non-western countries distribute more items about developed countries, and consequently, the difference between news portal sites and newspaper sites in the coverage of developing countries are smaller in non-western countries than in western countries.

H2: The difference between news portal sites and newspaper sites in the coverage of developing countries is smaller in non-western countries than in western countries.

Case selection

For the above comparison, both the western and non-western country should have (1) a large English-speaking population; (2) a sufficiently big news market; and (3) a large number of Internet users. The first criterion allows comparison under the same availability of news items. The second and third criteria are important for minimizing the influence of economic variables. The US fulfills all the criteria and is also the home country of Yahoo! and Google. India also fulfills all the criteria but it has a significantly different cultural background from the US. As we will see in detail below, India is a non-western developing country, has an English-speaking population of more than 125 million, is home to the world’s largest English-language newspaper, and has 81 million Internet users (Gupta, 2010). Furthermore, India represents the changes in globalization of media, so it is an interesting case in its own right. Thussu says: ‘the most significant recent changes in relation to the globalization of media is the rise of Asia, especially its two large countries. Any meaningful discussion of the internationalization of media studies, therefore, must take into account the rapid growth of China and India’ (Thussu, 2009: 18).

In recent years, the growth of India’s economy has been rapid, especially in the ICT sector. Over the last two decades, India’s GDP has accelerated and the growth rate reached 9.8% in 2006 (OECD, 2010). Boyd-Barrett (2006) says that India is the emerging power in ICT: by 2003, 75% of the world’s top 40 companies were, or were planning to be, operating in India; in 2004, western companies outsourced call centers and back offices to Indian companies to the value of $3.5 billion (3.5 times more than in 2001); in 2005, revenue from software, ICT services, and outsourcing was $17 billion (25% more than in 2003). IBM have been employing 9000 workers in India and additionally acquired one of the largest call center companies with 6000 workers; Hewlett-Packard and Oracle are respectively hiring 8000 and 4200 workers in India. The development of India’s outsourcing industries is due to the low communication costs on the Internet.

India is not only a rapidly growing economy, but also has a large English-speaking population and the high English proficiency of Indian workers enabled call centers to serve US customers (Kobayashi-Hillary, 2004). As mentioned, the English-speaking population in India is more than 125 million.² The population of the US and the UK are 313 million and 62 million, respectively, so India is virtually the second largest English-speaking country in the world. It has 114 English newspapers and the world largest English-language newspaper, The Times of India, with a circulation of 3.14 million and 13.3 million readers (Chaubey and Chandra, 2010).

Two news portal sites and two newspaper sites were selected from both the US and India based on the key distinctions of online news services made by Paterson (2005): (1) the provision of original content and (2) manual or automated selection of news. Yahoo! News and Google News were chosen as online news services that do not provide original content. Yahoo! News and Google News respectively represent news portal sites that manually and automatically select news items. Newspaper sites were chosen as online news services that provide original content: the online services of The New York Times (www.nytimes.com) and The Washington Post (www.washingtonpost.com) were chosen in the US, and those of The Times of India (timesofindia.indiatimes.com) and The Hindu (www.hinduonnet.com) in India. The Indian Readership Survey 2010 reports that The Times of India and The Hindu are the leading English-language newspapers, with a readership of 13.4 million and 6.3 million, respectively (Media Research User Council, 2010).

The services of Yahoo! News (news.yahoo.com) and Google News (news.google.com/news?ned=us) do not only address or are limited to the US market, both have Indian editions in the English language: Yahoo! News India (in.news.yahoo.com) and Google News India (news.google.com /news?ned=in). In India, Google and Yahoo! are the most popular portal sites, with a visitor share of 11.1% and 3.37%, respectively. There are local portal sites, such as Rediff (www.rediff.com/), but they are not as popular as Google or Yahoo! (Experian, 2011). Neither Yahoo! nor Google fully disclose their news selection mechanisms, but Google explains the criteria of its news collection as ‘freshness, location, relevance and diversity’ (Google, 2011).

Data collection

From each of the websites, 65,278 news items³ were collected every 10 minutes 24/7 between 20 February and 19 April 2011 in Pacific Time through RSS⁴ feeds by a computer program developed specifically for this research project. From the RSS feeds, the title, URL, publication date and time, description, and news source of each news item were extracted and stored in a database. Among all the kinds of news items, only those categorized as international news⁵ were collected by the program.

Content analysis

Given the large number of news items collected during the period, dictionary-based computer coding was performed. The advantages of computer content analysis are the capability to process large numbers of texts and the ability to code items reliably (Krippendorff, 2004). Computer content analysis has been used in the social sciences since Stone (1966), but no previous research utilizing computer content analysis for geographic news coverage was found, so the computer program was developed from scratch.⁶

The computer program was created using a script language (PHP) and a database system (MySQL). The program first eliminates unnecessary marks from titles and then searches each title for the words in the dictionary;⁷ if a word is associated with a specific country or region in the dictionary, it is marked as such. In order to develop an inclusive dictionary, computer coding was repeatedly applied to news items collected prior to the data collection period; words associated with specific regions were manually collected from the news items that were not automatically coded, and added to the dictionary. The dictionary was not only composed of adjectival and demonymic forms of place names (e.g., for Egypt, ‘Egypt’, ‘Egyptian’, ‘Egyptians’, ‘Alexandria’, ‘Alexandrian’, ‘Alexandrines’, ‘Cairo’, ‘Cairene’ and ‘Cairenes’) but also contained names of famous figures and words that are associated with the countries covered (e.g., for Japan, ‘earthquake’ and ‘radiation’).⁸ The program was designed only to search titles but not descriptions (body texts) of news articles, because titles are assumed to represent the subjects of the news stories, and there is a lot of noise (names of the newspapers and datelines) in the descriptions that can possibly lead to miscoding by the program.

Of the items, 89.0% were successfully coded by the computer program. In order to check the validity of the words in the dictionary, the agreement between the computer coding and human coding of 300 randomly selected items was calculated; coding was found to be in agreement in 87.3% of the items.⁹ Manual examination of 300 randomly selected uncoded items revealed that 56.3% were about subjects intended to be excluded from the analysis: international economy (10.3%), international issues (5.6%), science (6.6%), culture (14.0%), and sport (19.6%); the remaining 43.6% were not coded due to lack of words in the dictionary (34.0%), lack of traits in the titles of the articles (63.7%), or a system error by the news service (2.2%).¹⁰ In order to confirm the independence of uncoded items from regional coverage (developing vs developed countries), logistic regression analysis on the coverage of developing countries was performed. No significant difference was found in the analysis (p = .28); thus, the existence of uncoded items (11.0%) does not distort the patterns of news coverage in the coded items.

After the coding, since the focus of this research is only on international news, only news items coded as ‘US’ in the US online news services and only those coded as ‘India’ in the Indian online news services were excluded from the dataset.

Findings

The hypotheses were tested by logistic regression analysis: the response variable was coverage of at least one developing country and the explanatory variable was types of online news services (Yahoo! News, Google News, and newspaper sites). The two newspaper sites in their respective countries were treated as baselines with which the news portal sites were compared to find overrepresentation of developed countries in the respective countries. It was expected that the news portal sites would represent developed countries more than the newspaper sites in both the US and India, but the results of the statistical test (Table 1) showed that in the US, Yahoo! News US edition is significantly less likely to cover a developing country and the others were not significantly different from the newspaper sites, so the first hypothesis (H1) was rejected. Even after exclusion of news items reporting about the Japan earthquake on 11 March 2011, which was an exceptionally big news event happening in a developed country during the data collection period, it was found that Yahoo! News is, again, less and Google News is more likely to cover developing countries both in the US and India (Table 2).

Table 1.

Relative likelihood to cover developing countries.

Country	News service	Likelihood
US	Newspaper sites	1.00
	Yahoo! News	0.88***
	Google News	0.97
India	Newspaper sites	1.00
	Yahoo! News	0.98
	Google News	1.10

N = 56,959. Significance codes: *p < .05, **p < .01, ***p < .001.

Table 2.

Relative likelihood to cover developing countries (excluding Japan).

Country	News service	Likelihood
US	Newspaper sites	1.00
	Yahoo! News	0.73***
	Google News	1.28***
India	Newspaper sites	1.00
	Yahoo! News India	0.68***
	Google News India	1.41***

N = 46,825. Significance codes: *p < .05, **p < .01, ***p < .001.

Furthermore, the difference in the likelihood to cover a developing country between the newspaper sites and the news portal sites was not significant either in the US or India. There were differences in the relative likelihood to cover developing countries between the US and India (respectively, 0.68 vs 0.73 times, in Yahoo! News; 1.41 vs 1.28 times in Google News). Nevertheless, these differences do not mean that the relative likelihoods of news portal sites to represent developing countries are different between the US and India as seen in the interaction effects. Further statistical analysis showed that there were no statistically significant differences in the relative likelihood of news portal sites to cover developing countries between the US and India (Table 3): the difference between the news portal sites and newspaper sites were statistically significant (p < .001), and the difference between the US and India was also statistically significant (p = .001), but the interaction between the news portal sites and the countries were not significant. Therefore, the second hypothesis (H2) was also rejected.

Table 3.

Overall and interaction effects (excluding Japan).

Variables	Likelihood	Estimate	Standard error	z value	p value
(Intercept)	5.09	1.628	0.040	40.866	< .001***
Type of online service
Newspaper sites	1.00	–	–	–	–
Yahoo! News	0.73	–0.313	0.048	–6.469	.000***
Google News	1.29	0.253	0.047	5.349	.000***
Country
US	1.00	–	–	–	–
Indian	0.80	–0.223	0.068	–3.277	.001**
Interaction					.001**
Newspaper sites and India	1.00	–	–	–	–
Yahoo! News and India	0.94	–0.060	0.078	–0.776	.438
Google News and India	1.10	0.096	0.077	1.238	.216

N = 46,825; AIC = 42,973. Significance codes: *p < .05, **p < .01, ***p < .001.

The unexpected results were explained by the different patterns of news coverage within news portal sites. Their patterns of news coverage differed in two important aspects: degree of increase in news coverage of a certain country in response to a big news event (responsiveness) and level of concentration of news coverage to major countries of news events (concentration). The differences in responsiveness and concentration of news coverage appeared as noticeable differences in the patterns of news coverage between Yahoo! News and Google News.

The high responsiveness makes the news coverage patterns of Google News very volatile while the low responsiveness makes that of Yahoo! News stable. For instance, after the Japan earthquake, the likelihood to cover Japan increased by different degrees between the two news portal sites: the coverage in Google News became more than 16.1 and 18.9 times more likely in the US and India, respectively, while the coverage became only 5.5 and 7.1 times more likely in Yahoo! News and only 6.8 and 7.9 times more likely in newspaper sites (Figure 1).

Figure 1.

Increase in likelihood to cover Japan after the earthquake.

Regarding concentration of news coverage, Google News has a highly concentrated pattern, whereas newspaper sites and Yahoo! News have dispersed patterns. In the US, Google News covered 118, Yahoo! News 170, and newspapers sites 164 news events (Figure 2). The news coverage of the top three countries (Table 4) was highest in Google News (42.4% and 46.3%) and lowest in Yahoo! News (31.7% and 36.1%), so the Gini coefficients¹¹ were highest in Google News (0.91 and 0.92) and lowest in Yahoo! News (0.84 and 0.87).

Figure 2.

Distribution of news coverage in the US.

Table 4.

Distribution of news coverage.

Country	News service	Gini coefficient	Proportion of top 3
US	Yahoo! News	0.84	31.7%
	Google News	0.91	42.4%
	Newspaper sites	0.88	30.8%
India	Yahoo! News India	0.87	36.1%
	Google News India	0.92	46.3%
	Newspaper sites	0.90	45.0%

Overall, interestingly, the difference in international news coverage between online news services in the US and India was found to be much smaller than the difference between Yahoo! News and Google News (Table 5). The difference between the US (1.00) and India (0.81) is only 0.19, whereas the difference between Yahoo! News (0.71) and Google News (1.35) is 0.64.

Table 5.

Overall effects (excluding Japan).

Variables	Likelihood	Estimate	Standard Error	z value	p value
(Intercept)	4.12	1.42	0.04	39.206	<0.001***
Type of online service
Newspaper sites	1.00	-	-	-	-
Yahoo! News	0.71	-0.35	0.04	-9.237	0.000***
Google News	1.35	0.30	0.04	7.958	0.000***
Country
US	1.00	-	-	-	-
India	0.81	-0.21	0.02	-8.348	0.001**

N= 46,825; AIC= 42,977; Signif. codes: *p < .05, **p < .01, ***p < .001.

Discussion

The traditional formulation of the research questions, comparing countries to examine western bias, in this study did not appear to be really meaningful as the two hypotheses were rejected and as the difference in types of content selection procedures is the stronger determinant of news coverage than the difference in countries. Both in the US and India, the news portal sites were found to have the same trends, so there is no indication of cultural imperialism of Yahoo! and Google themselves. However, behind these news portal services, a heavy presence of western news agencies that leads to a skewed representation of developing countries was found. Yahoo! News and Google News are only a ‘portal’, so their contents are determined by their news sources and contents choice mechanisms. In Yahoo! News, western news agencies have a direct influence on its contents as the news providers, while they exercise influence on the contents on Google News as global news agenda-setters.

Neither Yahoo! News nor Google News produces news items themselves (Table 6), so their news coverage patterns are explained by the news sources they are relying on. Yahoo! News US edition is provided with news items mostly from western news agencies (91.7%): AP (39.4%), AFP (30.9%), and Reuters (21.3%). Numbers of news items fed by Indian news agencies comprised only a half of news items from the western news agencies: Reuters (60.1%), IANS (16.0%), and ANI (14.0%). Not surprisingly, 53.2% of the news items from Reuters appearing on Yahoo! News India were about developed countries. This supports the finding by Paterson (2005) that online news services are dominated by news agencies, and it also explains the rejection of the first hypothesis (H1) that news portal sites would represent more developing countries compared to newspaper sites.

Table 6.

Numbers of external news sources and proportion of news items.

Country	News service	Number of news sources	Proportion of news items
US	Yahoo! News	12	100.0%
	Google News	487	100.0%
	New York Times	15	12.3%
India	Yahoo! News India	5	100.0%
	Google News India	250	100.0%
	The Hindu	6	77.3%

The news coverage patterns of Yahoo! News can be explained by the well-known western bias of news agencies. However, the reason the coverage patterns of Google News are so volatile is more complicated. It is counterintuitive that Google News covered only 118 countries with 487 sources, whereas Yahoo! News and the newspaper sites covered 170 and 164 countries only having 12 and 15 sources, respectively. Indeed, Google News US edition is not dependent on western news agencies: its top three sources are Reuters (9.1%), BBC News (6.8%) and The New York Times (4.5%); its India edition also has 250 sources and Reuters (9.9%), BBC News (8.2%), and AFP (4.6%) are the top sources. While Yahoo! News US edition has no news providers in developing countries at all, 11.8% of news items on Google News US edition were provided from news sources in 22 developing countries. They include Xinhua (China), Aljazeera.net (Qatar), Ha'aretz (Israel), Jerusalem Post (Israel), and The Hindu (India); and even a government-funded Iranian satellite broadcaster, Press TV, was found among the news sources.

The high responsiveness and concentration of Google News are explained by the joint influence of its automated news selection mechanism and global dominance of western news agencies. Western news agencies disseminate homogeneous news stories across the world and their news coverage is aggregated and redistributed by Google News.

It was found by Barabási et al. (2000) that the characteristics of the link structure on the Internet is a highly concentrated ‘power-law’ distribution (logarithm distribution): only a limited number of popular websites have millions of links from other minor websites. They explain the reason for the power-law distribution on the World Wide Web by the concept ‘preferential attachment’: new websites are more likely to have links to well-known, popular websites than unpopular ones. Google’s underlying web search algorithm known as Page Rank is based on analysis of the link structure on the World Wide Web and it locates most frequently linked websites at the top of the search results. As a result of this, people are usually provided with links only to very popular websites by Google Web Search.

The power-law distribution and preferential attachment were not only found in the link structure of the Web. Himelboim et al. (2010) found the same power-law distribution of news coverage on the newspaper sites. They conducted a content analysis of the geographic coverage of news items on 223 newspaper sites in 73 countries and found a highly concentrated distribution of news coverage to major countries such as the US, Pakistan, France, the UK, and Israel. The relationship between the news coverage on the newspaper sites and the concentrated news coverage on Google News found in this research exactly mirrors the relationship between the power-law link structure on the World Wide Web and the links on Google Web Search. This similarity allows us to explain the concentrated distribution of news coverage on Google News – that is, news coverage on newspaper sites across the world as a whole is concentrated to a limited numbers of countries and Google News’s contents reflects this skewed distribution of news coverage; in other words, it is not Google News that overrepresents a certain country, but newspaper sites, from which it collects news items. In fact, a close observation of the news stories collected by Google reveals that most minor online news services are in fact simply redistributing news stories fed by news agencies. Just as only a few newspapers have an international news gathering capacity (Boyd-Barrett, 1980), so too only limited numbers of online news services submit original news stories. Google News creates clusters of news items to estimate the importance of each event and those clusters are presented and a full list of news items are linked from its top page. Among over 1000 news stories in a top news cluster, news stories ranked below 100 are predominantly from low-tier online news services, which rarely have international news production capability.

Most newspaper publishers are dependent on western news agencies for international news items, and their decisions in prioritization of news events and allocation of resources to cover news are largely influenced by news distributed by agencies (Boyd-Barrett, 1980). Today, the situation may be slightly different in a way that not only news agencies but also major western newspaper sites (The New York Times, The Washington Post, The Guardian, etc.) are influencing news coverage on news outlets across the world. However, there is still an oligopoly in international news production that possibly makes Google News’s news coverage pattern so concentrated. The high responsiveness of the news coverage too can be well explained by the global dominance of a handful of news organizations: the sudden changes in news coverage on Google News is a reflection of changes in news coverage on news sites all over the world; such rapid changes in globally dispersed news sites cannot occur without entities that have a global reach.

The international news agenda-setting function is still in the hands of western news organizations and Google’s automated news aggregation mechanism reflects the agenda through newspaper sites across the world. Google itself may not have the intention to represent a western perspective on its news portal sites, but, through its automated news gathering mechanism, its news choice is consequently reflecting the western news organizations’ agenda.

Further, the fact that Google News’s coverage is more concentrated than that of Yahoo! News, which depends more on western news agencies than it does, implies that the skewness of coverage can be intensified by Google News as a result of the automated aggregation. Its social impact might be a misjudgment of the importance of foreign events by its users. Users of Google News are not only news consumers but also news producers, reporters, and editors, so this can be another cause of overrepresentation of major countries in the international news arena. In that case, a news coverage distribution, which initially has low unevenness, is multiplied repeatedly through Google News, and finally becomes a truly uneven power-law distribution.

Conclusion

Yahoo! News and Google News sites per se do not have a western bias, but western news agencies have very strong influence on the news portal sites, because they do not produce news content themselves. Yahoo! News is provided with news items only by news agencies, and it represents developed countries more than newspaper sites because of overrepresentation of developed countries in news items provided by western news agencies. Google News has diverse news sources and its news items are not directly provided from news agencies, but it reflects the global news agenda set by the international news agencies, and thus represents relatively few countries with big news events through its automated news gathering mechanism. The global influence of western news agencies is uniformly spreading across the world on the Internet, so both western and non-western users are exposed to news stories about news events featured by western news agencies when using news portal sites.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Barabási

A-L

Albert

Jeong

(2000) Scale-free characteristics of random networks: The topology of the world-wide web. Physica A: Statistical Mechanics and its Applications 281 (1–4): 69–77.

Berger

(2009) How the internet impacts on international news: Exploring paradoxes of the most global medium in a time of ‘hyperlocalism’. International Communication Gazette 71(5): 355–371.

Boyd-Barrett

(1980) The International News Agencies. London: Constable.

Boyd-Barrett

(2006) Cyberspace, globalization and empire. Global Media and Communication 2(1): 21–41.

Chang

T-K

Lau

T-Y

Xiaoming

(2000) From the United States with news and more: International flow, television coverage and the world system. International Communication Gazette 62(6): 505–522.

Chaubey

Chandra

(2010) Access to online newspapers of India: An overview. In: Gaur

Zarmdt

Gupta

Gaur

(eds) Digital Preservation and Access to News and Views. The Hague: IFLA, pp. 188–194.

Experian (2011) Press release. Experian launches Hitwise internet measurement service for India. Available at: www.experian.com/hitwise/press-release-experian-launches-hitwise-india.html.

Galtung

Ruge

(1965) The structure of foreign news: The presentation of the Congo, Cuba and Cyprus crises in four Norwegian newspapers. Journal of Peace Research 2(1): 64–90.

Google (2011) About Google News. Available at: news.google.com/intl/en_us/about_google_news.html.

10.

Gupta

(2010) Googling news and viewers online. In: Gaur

Zarmdt

Gupta

Gaur

(eds) Digital Preservation and Access to News and Views. The Hague: IFLA, pp. 195–204.

11.

Hachten

Scotton

(1999) The World News Prism: Changing Media of International Communication. Ames: Iowa State University Press.

12.

Herman

(2000) The propaganda model: A retrospective. Journalism Studies 1(1): 101–112.

13.

Himelboim

Chang

T-K

McCreery

(2010) International network of foreign news coverage: Old global hierarchies in a new online world. Journalism and Mass Communication Quarterly 87(2): 297–314.

14.

Kobayashi-Hillary

(2004) Outsourcing to India: The Offshore Advantage. Berlin: Springer.

15.

Krippendorff

(2004) Content Analysis: An Introduction to its Methodology. London: Sage.

16.

Media Research User Council (2010) Indian readership survey. Available at: www.mruc.net/mages/stories/IRS2010Q2%20Topline.pdf.

17.

OECD (2010) Country statistical profiles 2010. Available at: stats.oecd.org/.

18.

Paterson

(2005) News agency dominance in international news on the internet. In: Skinner

Compton

Gasher

(eds) Converging Media, Diverging Politics: A Political Economy of News Media in the United States and Canada. Lanham, MD: Lexington Books, pp. 145–162.

19.

Pew Research Center (2011) The state of the news media 2011. Available at: stateofthemedia.org/2011/online-essay/data-page-7/.

20.

Reese

Rutigliano

Hyun

Jeong

(2007) Mapping the blogosphere: Professional and citizen-based media in the global news arena. Journalism 8(3): 235–261.

21.

Schiller

(1991) Not yet the post-imperialist era. Critical Studies in Mass Communication 8(1): 13–28.

22.

Schramm

(1978) International News Wires and Third World News in Asia: A Preliminary Report. Medford, MA: Edward R Murrow Center of Public Diplomacy, Fletcher School of Law and Diplomacy, Tufts University.

23.

Stevenson

(1994) Global Communication in the Twenty-First Century. Harlow: Longman.

24.

Stone

Dunphy

Smith

Ogilvie

(1966) The General Inquirer: A Computer Approach to Content Analysis. Cambridge, MA: The MIT Press.

25.

Thussu

(2009) Introduction. In: Thussu

(ed.) Internationalizing Media Studies. London and New York: Routledge, pp. 1–10.

26.

UNESCO (1985) International Flow of Information: A Global Report and Analysis. New York: UNESCO. Available at: unesdoc.unesco.org/images/0006/000652/065258eo.pdf.

27.

Van der Wurff

(2005) Introduction: Impacts of the Internet on newspapers in Europe. Gazette 67(1): 5–7.

28.

Van der Wurff

(2008) The impact of the Internet on media contents. In: Küng

Picard

Towse

(eds) The Internet and the Mass Media. London: Sage.

29.

Van Ginneken

(1998) Understanding Global News. London: Sage.

30.

Wanta

Y-W

(1993) The agenda-setting effects of international news coverage: An examination of differing news frames. International Journal of Public Opinion Research 5(3): 250–264.

31.

(2000) Systemic determinants of international news coverage: A comparison of 38 countries. Journal of Communication 50(2): 110–130.

32.

(2007) A brave new world for international news? Exploring the determinants of the coverage of foreign news on US websites. International Communication Gazette 69(6): 539–551.