This paper analyses the Web sites of 82 national daily newspapers in eight Western European countries, in order to determine their online revenue models. Overall, 80.5 percent of our newspapers try to monetise (part of) their online content in direct ways. In doing so, the bulk of the paying sites rely on ‘traditional’ online subscriptions rather than unbundled access options, although 43.9 percent of all sites offer both, and in this way target regular as well as occasional readers. However, the low cutoff points between the online subscriptions and payperview seem to indicate that our newspapers have embraced payperview only halfheartedly. We also find that newspapers that adopt mixed bundling have a higher fear of channel spillover between the online and print subscription. The same is true for smaller newspapers. An important finding is also that for all the strategic decisions that we analyse, there are dramatic intercountry differences. This suggests that local market circumstances — such as the relative importance of advertising revenues, Internet penetration, and even mimicking behaviour play a major role.
2. The strategic decisions of online newspapers
3. Data collection
4. The free or fee dilemma
5. To bundle or not to bundle The degree of aggregation
6. The archive
7. The price of the access options
The objective of this paper is to analyse how Western European newspapers price their online content. Firstly, there is the question whether the Web sites are free or not. In other words, do newspapers try to monetise their online content directly (by charging for it) or indirectly (by relying on advertising revenues)? For the paying sites, we also took a closer look at the subscription formulas and alternative access possibilities offered. Here, our main interest was the extent to which newspapers play it safe and mainly offer traditional subscriptions, or are expanding their digital horizon, for instance by unbundling the news and offering it on a payperview basis.
In particular, we analysed the offering of online news (free versus paid content) and the access options (subscription formulas, PDF versions, day pass, payperview, etc.) for 82 newspaper sites in eight countries (Belgium, the Netherlands, Luxembourg, France, Germany, Italy, Spain and the United Kingdom). We also checked whether a bundle of the online and print version was available, potentially at a reduced price. In our analysis, we paid special attention to and tried to measure the fear of cannibalisation that might exist between print and online subscriptions on a first level and between online subscriptions and payperview on a second. By looking at all these strategic dimensions, the paper documents the differences in online revenue models and price setting between the newspapers in our dataset.
The remainder of this paper is structured as follows. Section 2 first provides a brief literature review and discusses the core strategic decisions of online newspapers. Section 3 describes the way data was collected. The following Sections then contain an extensive overview of our results, after which conclusions are presented in the eighth and final Section.
2. The strategic decisions of online newspapers
As was already stressed in the Introduction, newspapers need to make several important strategic decisions when entering the Internet arena. Firstly, they need to choose between a free and a (semi)paidfor site (Noam, 2006). The recent moves by major newspapers demonstrate the uncertainty related to this choice. For one, in September 2007, the New York Times decided to end its TimesSelect subscription service, a move that will make its top columnists and its archive freely available online. The rationale behind the decision is that the company expects to earn more money through advertising sales from increased traffic on its Web site than it did by asking readers to pay a fee for its premium content . Following its acquisition by News Corp., the Wall Street Journal is also reexamining its 11yearold paidsubscription Web site . Again the goal is to lift online ads from an anticipated jump in readers. More recently even, on 1 October 2007, FT.com announced an innovative charging system, again fuelling the debate about newspapers online revenue models. From midOctober, the site of the Financial Times would pioneer a new approach whereby articles and data would be free to users up to a total of 30 views a month. They would then be asked to subscribe for access to more material. The Financial Times believes that this new model will allow it to have their cake and eat it. Obviously, opening up the online material should increase traffic to the site, and thus also advertising revenues. But the Financial Times is convinced that many of the additional FT.com visitors will appreciate the quality of the journalism, and will join the ranks of the existing subscribers .
Clearly, given the downward spiral of print revenues, most newspaper executives today feel they have no choice but to grab as big a slice of the onlinead pie as they can . Then again, Vin Crosbie of Digital Deliverance, a consulting firm, estimates that newspapers need between 20100 readers online to make up for losing just one print reader . Also, the oncetorrid growth in onlinenewspaper ad revenue is slowing as newspapers face increasing online competition from Web portals and TV networks. According to the Newspaper Association of America, the rate of growth of onlinenewspaper ads dropped to 19.3 percent during the second quarter of 2007, down from 33.2 percent during the second quarter of 2006. A final consideration is that online ads still make up only a small portion of total newspaper revenues, just seven percent of the US$11.3 billion total print and onlinenewspaper ad revenues during the second quarter .
The recent examples mentioned above illustrate the difficulties related to choosing between or trying to marry direct (i.e., charging for content) and indirect (i.e., advertising) revenue sources. Another strategic concern for newspapers is that at all times they need to avoid their printed version being cannibalised by the online version. In general, it is widely assumed that the Internet is cannibalistic . Several authors argue that offering digital content will cannibalise magazines and newspapers print sales (Deleersnyder, et al., 2002; Pauwels and Dans, 2001). However, the empirical evidence on whether online news and news onpaper are independent goods, complements or substitutes is mixed . Chyi and Lasorsa (2002) conducted a randomsample telephone survey in Austin, Texas, to investigate the publics response to newspapers print and online editions. They find that there might, to some extent, be a complementary market relation between the two editions. Kaiser and Kongsted (2005) assess the impact of Web site visits on the demand for 42 German magazines and find that Web site visits actually increase circulation. Deleersnyder, et al. (2002), for their part, look at 85 British and Dutch newspapers that establish Web sites. They find that offering a site has no effect on print sales. In the same line, Kaiser (2006) assesses the impact of a magazines Web site on the demand for its print content and for its advertising space in a small sample of German womens magazines. He finds that having a Web site has no effect on the demand for either product. On the other hand, Filistrucchi (2005) finds evidence of substantial cannibalisation for four Italian newspapers that offered Web sites. Gentzkow (2007) uses microdata from the Washington D.C. market in the period March 2000February 2003 and finds that the online version of the Washington Post reduced the Posts print readership by 27,000 readers per day (i.e., 1.5 percent), at a cost of US$5.5 million per year in lost print profits. Finally, Simon and Kadiyali (2007) also find strong evidence that digital content cannibalises print sales. They analysed a sample of U.S. consumer magazines from 1996 to 2001 and find that, on average, a magazines print circulation declines about threefour percent when it offers a Web site. Offering digital access to the entire contents of the current print magazine even reduces print sales by nine percent.
In trying to avoid the possible cannibalisation of the print by the online version, pricing is obviously crucial, as this determines the relative attractiveness of the two options. However, bundling decisions also enter into play. Indeed, when newspapers opt to offer online content against payment, this is possible in various degrees of aggregation (Bakos and Brynjolfsson, 1997). At one end of the continuum, newspapers could for example offer a package that contains the printed as well as the online version at a special price (Bakos and Brynjolfsson, 1999) an approach we will call metabundling. At the other end, newspapers can unbundle the content entirely and offer articles and/or columns on an individual basis too (Fishburn, et al., 1997; Fishburn and Odlyzko, 1999).
However, the latter approach raises cannibalisation concerns on a lower level, namely between the online subscription options and the payperview option. When payperview is priced too low, this may result in fewer subscriptions, while setting the price for individual articles too high makes them unattractive for occasional visitors (Stahl, et al., 2004). Again, careful price setting will be needed.
Summing up, when extending their business online, newspapers have to make several important and difficult decisions in response to multiple strategic challenges. The rest of this paper documents the decisions taken by our 82 national newspapers. Section 3 first describes the way data was collected. The free or fee dilemma is dealt with in Section 4, after which the bundling concept and the observed level of aggregation are discussed in detail in Section 5. Section 6 discusses how newspapers handle their online archives. Finally, we analyse the pricing of the different access options in Section 7. At each time, the results are preceded by an overview of our definitions, in order to clarify our approach to classification.
3. Data collection
During the period JuneJuly 2006, the Web sites of 82 newspapers in eight countries were examined. More specifically, this analysis included all national newspapers with a daily circulation in Belgium (9 sites), Luxembourg (6), the Netherlands (8), Germany (10), the United Kingdom (10), Italy (20), Spain (8) and France (11) . Regional and local newspapers were not included because of their specific orientation and generally smaller market share . Newspapers aimed at an extremely specific audience (for example horselovers, children, etc.)  also fall outside the scope of our research. The same applies to freely distributed newspapers such as the Metro. They typically only generate revenue through advertising and thus have no strategic decisions to make for their Web sites on this subject. In choosing the countries, their geographical position was an important factor since we wanted to be able to compare certain (neighbouring) countries and hence highlight interesting differences. Additionally, we were limited by our linguistic capacities. We were, for example, unable to include a Scandinavian country in our analysis.
Importantly, only the national daily newspapers, as certified by the national coordinating organisations, are studied in our research. This number differs considerably from country to country. For example, we have no less than 20 Italian newspapers in our dataset, whereas in the Netherlands only eight newspapers are defined as national daily newspapers. One could therefore argue that Italy is overrepresented. However, we have always calculated the average of the country averages, and checked if this figure varied from the unweighted average of all newspapers. The divergence remains limited to one or two percentage points at most. To conclude, we emphasise that when collecting the data, we operated in a selective though exhaustive way; selective in the sense that we made choices regarding the type of newspapers and countries, but exhaustive in that we did not take a random sample but examined the total population within our selection.
4. The free or fee dilemma
As explained in Section 2, the highest level strategic decision for online newspapers is to decide whether to offer their content for free or against payment. However, this is not an allornothing assessment since it is perfectly possible to offer content partly for free and partly against payment. For instance, newspapers can decide to offer recent news in HTML format for free, while charging for archived articles. Due to these and other gradations, the distinction between free and fee becomes slightly blurred, and hence it is not straightforward to categorise newspapers as being free or fee.
Our approach which is explained in Box 1 primarily hinges upon the following question: Do newspapers try to monetise their online content directly or indirectly (i.e., by means of advertising revenues)?. Put differently, we focus on the nature of the revenue model. As a result, the category free contains newspapers that see their Web site merely as a promotion vehicle and/or that purely aim at advertising revenues .
Box 1: An overview of the definitions of free and fee.
In our taxonomy, a free Web site denotes that all the news , columns and archived articles (if available) can be consulted for free on the site.
In contrast to free, fee means that some content is charged for on the Web site. This may be limited to a few columns, the archive or ultimately all content. We further divided this category into fee light and fee+. Indeed, some sites offer content completely for free in HTML format, and solely try to monetise their online content by means of a chargedfor PDF version of the paper of the day.
This category contains the newspaper sites where all the news, the columns and the archived articles (if available) can be consulted free of charge, and where solely the PDF version is offered against payment.
These sites also offer something against payment, but either something other than a PDF version, or something more.
Figure 1 shows that, overall, 16 out of the 82 Web sites (19.5 percent) are completely accessible free of charge, while 66 (80.5 percent) belong to the category fee. Within this second category, 19 newspapers (23.2 percent of all sites) only have a chargedfor PDF version on the Web site, while 47 newspapers or 57.3 percent are part of the category fee+.
Figure 1: The global proportion of FREE vs. FEE.
Figure 2 shows the proportion of free vs. fee light/fee+ per country. Spain, France and Belgium lie with 0, 9 and 11 percent respectively clearly below the average concerning the number of free Web sites. On the other hand, free Web sites are strongly present in Luxembourg (33 percent) and Italy (30 percent). In terms of the classification fee light/fee+, France, in particular, is an outlier, with 82 percent of the Web sites belonging to the category fee+. This will become even clearer when we discuss the results for payperview (see Section 5.2). It is also remarkable that 75 percent of the Spanish newspapers and 60 percent of the British newspapers try to monetise their online content in novel ways. That is, in a different way than merely via PDF versions; for example, by charging for columns. A study by Herbert and Thurman (2007) of the U.K. online newspaper market confirms the latter finding. They conducted qualitative research interviews and a survey between July and August 2006, in order to analyse online business models at the 10 national newspapers as well as at two popular regional newspapers in the U.K. Herbert and Thurman find that the selected newspapers all charge for something on their Web sites , and are expanding their range of commercial services. This said, none of the selected newspapers charge for their most popular area of content: generalinterest news. In most cases, this is also true for archived content. However, a walledgarden approach is strongly present for columns and for business news and analysis. Returning to our own results, in other countries it is not uncommon for a Web site to be freely accessible except for the PDF version, which is charged for. This fee lightstrategy is particularly popular in Germany, the Netherlands and Belgium.
Figure 2: The proportion of FREE vs. FEE+/FEE LIGHT per country.
Summing up, the bulk of our newspapers (80.5 percent) try to directly monetise at least some of their content. This said, in doing so, about 30 percent of the paying sites resort to what one could call a rather traditional approach: the print content is digitalised as is, and offered in PDF format as the only chargedfor option. However, more than half of the paying sites 71.2 percent to be precise (also) try something else.
5. To bundle or not to bundle The degree of aggregation
If we now focus on the paying sites, it should first of all be stressed that the nature of the direct online revenue sources of newspapers is closely related to the level of aggregation or bundling of content. Newspapers can opt for a strong form of bundling by offering a package that contains the print and online version at a discount. At the other end of the continuum, newspapers can also decide to completely unbundle their content by offering individual articles on a per unit basis (i.e., payperview). Of course newspapers can also choose to offer both bundles of content as well as payperview (Shapiro and Varian, 1998; Bakos and Brynjolfsson, 1999).
In our analysis of the European online newspaper market, we focus on two dimensions of bundling, namely bundling of goods and bundling across time. The first dimension involves combining multiple goods to offer them in one package to the customer. The second dimension is about offering goods or services to the customer for a certain period. The third dimension of bundling bundling of consumers is not considered here. In the case of newspapers, this would amount to site licensing, so that for instance a whole campus gets access to the newspapers Web site for a fixed price.
When analysing Web sites, we thus checked in what packages and/or for which time period online newspapers offer their content. In what follows, we try to present these options in order of decreasing intensity of bundling. More specifically, this classification goes from (bundles of) subscriptions to entirely unbundled access alternatives.
5.1: Positioning the online subscription versus print
Firstly, we investigate how newspapers with a chargedfor Web site position their online subscription compared to the printed version. In doing so, we discern three possibilities, as defined in Box 2. In order to distinguish this bundling decision from the bundling decision on the level of copies and articles (that is, the choice between subscriptions and payperview), we talk about metabundling here, since subscriptions are bundles in their own right. Indeed, as is explained in Box 2, under both pure and mixed bundling, newspapers actually offer a bundle of bundles.
Box 2: An overview of the metabundling possibilities for newspapers.
The print and online versions are only available as a package. In other words, there is no separate subscription to the online version available . For example: bundle (online + print) for €320 annually.
The print and online version are offered for sale individually, but can also be bought as a bundle at a discount. For example: print for €250 annually, online for €150 annually, and a bundle for €320 annually. When print subscribers get free access to the online news, we also consider this to be mixed bundling (albeit with a very large discount); on condition of course that the online version is available as a standalone product as well.
The online subscription is only available as a standalone product. This means that the print subscribers also pay the full price for the online version; in other words, they are not entitled to special discounts. For example: print for €250 annually, online for €150 annually.
Figure 3 gives an overview of the metabundling strategies per country. Newspapers that simply do not offer any online subscription formula are categorised under no online subscription . It goes without saying that newspapers in this category do not have to take any metabundling decisions. Figure 3 shows that in Italy (40 percent), the Netherlands (37.5 percent) and France (36.4 percent) there is often no online subscription available, while online subscriptions have the highest penetration in Spain (12.5 percent no online subscription) and the United Kingdom (20 percent no online subscription).
Only one (Belgian) newspaper (1.2 percent) applies the pure bundling principle. Twentyseven newspapers (32.9 percent) opt for a mixed bundling strategy, and 28 newspapers (34.1 percent) choose no bundling. On a per country basis, it is remarkable that in the United Kingdom 70 percent of the newspapers belong to the category no bundling, and that only 10 percent choose mixed bundling. Similarly, in Spain, 62.5 percent choose no bundling and only 25 percent opt for mixed. In Germany, however, mixed bundling strategies (60 percent) have a much stronger presence than no bundling (10 percent). In the Netherlands (50 percent), France (45.5 percent) and Belgium (44.4 percent) too, mixed bundling is more popular than no bundling (12.5 percent, 18.2 percent and 11.1 percent respectively).
Figure 3: The metabundling decisions per country.
When explaining the various meta-bundling possibilities (in Box 2), we mentioned a situation in which print subscribers get free access to the online version (which is to be paid for by others) and how this is also a form of bundling; more specifically mixed or pure, depending on the availability of the online version as a standalone product. Obviously, this tochargeornottocharge question also reflects a difference in strategy. Hence it is worthwhile making this distinction too.
Figure 4: Percentage of the bundling options whereby print subscribers get free access to online facilities.
When looking at the percentage of newspapers that give free access to print subscribers (Figure 4), we find that newspapers that rely on mixed bundling, grant free access to print subscribers in 63 percent of the cases. In the Netherlands, Luxembourg and Spain even all the newspapers that apply mixed bundling adopt this strategy. In Germany (66.7 percent) and Belgium (50 percent) this strategy is also quite popular. Finally, our only (Belgian) pure bundling newspaper gives free access to print subscribers too.
5.2: Unbundled access options
Nowadays, when one wants to consume news, it is often not a prerequisite anymore to have a subscription. Newspapers frequently offer various methods of consuming online articles in smaller packages, even piecemeal. In what follows, we give an overview of these alternatives, largely in order of decreasing degree of bundling over time and/or the number of articles.
Box 3: A description of the unbundled access options.
DAY PASS AND XDAYS PASS
A day pass gives complete access to the Web site for 24 hours. One can thus consult all the online articles for one day . This is a convenient way to search the archive for relevant articles. Note that this is not the same as the purchase of one PDF version of the newspaper. In the latter case, one merely buys the digital version of the newspaper for the day in question, which can afterwards be browsed offline on the screen or which can be printed out. This is something completely different from the surfing on the site principle. A variant on the day pass method is the Xdays pass. This option gives access to the site for X number of days. When following the logic of decreasing degree of bundling, the X-days pass actually precedes the regular day pass.
CREDITSCARD (OR WALLET)
Here the visitor of the site buys X credits at a certain price. With this card  he subsequently pays one or more credits per article (often depending on the length of the article), after which the amount is deducted from the cards balance. This access option permits the reader to select individual articles just as in the case of real payperview but since one has to buy a minimum of X credits, one in fact has to buy a bundle of articles. However, the reader does decide on the composition himself. This is in fact an example of customised bundling  of articles .
This access option gives the opportunity to purchase and pay for articles on an individual basis. The reader for instance screens the archive, finds the relevant article and pays for it on a perunit basis. Another example is the purchase of a column on a perunit basis. Another way to explain the difference between a creditscard and payperview is that the latter implies payasyougo, whereas a creditscard is in fact (partly) prepaid.
PAYPERVIEW IN A BROAD SENSE
When newspapers offer payperview and/or a creditscard, we consider this to be payperview in a broad sense. In fact, in both cases articles are purchased on a perunit basis; the only difference is that in the second case one pays upfront for a certain amount of credits.
As Figure 5 illustrates, 11 percent of all newspapers offer a day pass. Strikingly, this option is simply nonexistent in Luxembourg, the Netherlands and Italy. On the other hand, the day pass has a strong presence in Belgium (55.6 percent). The possibility of accessing the site for X days via a pass is offered by 13.4 percent of the newspapers. The Xdays pass is not present in Luxembourg, the United Kingdom or Spain, and is most popular in the Netherlands, where 25 percent of the newspapers offer this access option.
Figure 5: Frequency of the day pass and the Xdays pass (with regard to all newspapers).
When we narrow our focus to the category fee, we can see in Figure 6 that overall the day pass is offered by 13.6 percent of the chargedfor Web sites. Belgium is with 62.5 percent a big outlier. In the Netherlands and Italy, 33.3 percent and 28.6 percent respectively of the paying sites offer an Xdays pass. They also clearly lie above the average.
Figure 6: Frequency of the day pass and the Xdays pass (with regard to the category fee).
In Figure 7, 23.2 percent of all newspapers offer a creditscard. This access option is most strongly represented in France (45.5 percent), Italy (35 percent) and Luxembourg (33.3 percent), but is nonexistent in the Netherlands. Overall, payperview is offered by 19.5 percent of the newspapers. In Belgium and Luxemburg this pure payperview option is nonexistent. It is most popular in France (54.5 percent) and Spain (37.5 percent), and the same goes for payperview in a broad sense with 54.5 and 62.5 percent respectively.
Figure 7: Frequency of the creditscard, payperview and payperview broad (with regard to all newspapers).
Figure 8 shows that a creditscard is offered by half of the chargedfor Web sites in Luxembourg, France, as well as Italy. Payperview has amongst the paying sites a penetration degree of 60 percent in France and 50 percent in the Netherlands. Finally, it is remarkable that in Spain, 62.5 percent of the chargedfor sites offer payperview in a broad sense.
Figure 8: Frequency of the creditscard, payperview and payperview broad (with regard to the category fee).
To sum up, payperview and the creditscard are most frequently offered in France, and the day pass in Belgium .
5.3: A general picture
In the two preceding subsections, we have documented how newspapers position their online and print subscriptions (5.1), and we have given an overview of the type of unbundled access options they offer (5.2). In other words, we have taken a look at the two extremes of the bundlingunbundling continuum. In this subsection, we try to present a more general picture by analysing how many newspapers offer online subscriptions, how many offer unbundled access options, and whether this is an either/ordecision, or not. Concretely, we calculated what percentage of the newspapers in a given country offer at least one unbundled access option (either payperview, a creditscard, a day pass or an Xdays pass), at least one type of subscription formula (either a general online subscription or a PDFsubscription), or both.
Overall, some threequarters of our newspaper sites (76.8 percent to be precise) offer online subscription formulas. Concretely, this means that about 95 percent of the paying sites offer a type of online subscription. As for the unbundled access alternatives, we find that 47.6 percent of all newspapers (i.e., 59.1 percent of the fee sites) offer at least one such an option. But again the intercountry differences are striking: in the United Kingdom (70 percent), Luxembourg (66.7 percent) and Italy (65 percent) the majority of the newspapers still do not offer any unbundled access possibilities at all. France, on the other hand, is clearly the most diversified country since 9.1 percent of the newspapers there even offer all the unbundled access options at the same time.
Figure 9: Frequency of the access alternatives overview on a percountry basis.
Figure 9 combines the results just presented and shows that, overall, 43.9 percent of the sites offer subscriptions as well as unbundled access alternatives; 32.9 percent limit themselves to offering only subscriptions, while a mere 3.7 percent offer innovative access options and nothing else. In other words, about half of the paying sites (54.5 percent) offer both subscriptions and unbundled options, which seems to indicate that when newspapers opt to monetise their site directly, many of them decide to charge both regular and casual users. On a percountry basis, it stands out that in Spain, Belgium and France, more than half of the newspaper sites offer both subscriptions and unbundled access possibilities. In the United Kingdom, Germany and Italy, on the other hand, many newspapers still rely on the traditional subscription approach. Only in the Netherlands, France and Spain, we encounter (a limited number of) newspaper sites that opt to entirely unbundle their content.
6. The archive
When collecting the data, we discovered that newspapers not only handle their current articles in a variety of ways, but that the same is true for their archive. Obviously, this is also a crucial dimension of newspapers online behaviour. In some cases, access to the archive is completely free for everybody. In other cases the archive is only accessible for subscribers to the online and/or print version. Another possibility is free archival access for subscribers, but charges for occasional visitors. A final option is that the archive is charged for to everybody. In our analysis, we only examine the splitup free for everybody vs. chargedfor, the latter thus being broadly defined.
Figure 10: Free vs. chargedfor archive overview on a percountry basis.
Figure 10 shows that 47.6 percent of the newspapers provide free access to the archive, while it has to be paid for in 42.7 percent of the cases. Almost 10 (9.8) percent of the newspapers do not provide an online archive yet. In the Netherlands (62.5 percent), the United Kingdom (60 percent) and Germany (60 percent) a large number of newspapers opt for a free archive . It is notable that 55.6 percent of the Belgian newspapers have a chargedfor archive, while this percentage even reaches 63.6 percent in France. As an aside, it is also striking that in the United Kingdom 20 percent of the newspapers do not provide an online archive. In France this proportion is also above average, with 18.2 percent.
It is also interesting to link these results with those reported in the previous Section. Indeed, one would expect newspapers with a chargedfor archive to offer unbundled access options more frequently than newspapers with a free archive. For payperview and the creditscard, we indeed found a significant positive relation with the existence of a chargedfor archive (resp. χ2 = 13.236; ρ < .05 and χ2 = 12.391; ρ < .05). There is also a significant positive relation between the Xdays pass and a chargedfor archive (χ2 = 6.177; ρ < .05); in contrast, there is no significant correlation between the day pass and the archive (χ2 = .280; ρ > .05) . A possible explanation for these findings might be that payperview and the creditscard are specifically designed for purchases on a perunit basis, while the day pass has a more general character in the sense that it enables readers to visit the whole Web site for a day. The substantially lower χ2value for the Xdays pass seems to support this interpretation.
7. The price of the access options
As explained in Section 2, careful price setting is one way newspapers can try to avoid the danger of cannibalisation that exists on multiple levels. In this Section we therefore briefly examine and compare the prices of subscriptions and unbundled access options.
7.1: The price of subscriptions
Even though we focus on online content in this paper, it is nevertheless interesting to first have a look at the price of print subscriptions, as this provides a benchmark. In our dataset, the minimum price of an annual subscription to the printed version amounts to €104 and the maximum price is €518.7. The average price is €253.8, the median is €240 and the mode €235. Print subscriptions are most expensive in Spain, France and Germany (on average €311.3, €308.3 and €285.7 respectively) and cheapest in Luxembourg and Italy (on average €153.8 and €209.2 respectively).
The price of an annual subscription to the online version varies from €50 per year up to €370 per year at most. On average it costs €161.9 with the median being €126 and the mode €80. Online subscriptions are thus clearly cheaper than print subscriptions. Lower production costs probably play a role here, but it also has to be stressed that the content is not necessarily the same. The Netherlands and France are relatively expensive for online subscriptions (on average €181.3 and €224 per year respectively). Spain is the cheapest country (€65 on average); this contrasts sharply with its prices for print subscriptions (cf. supra).
A subscription to the PDF version varies from €50 to €291 annually. It costs €134.7 on average per year, with the median and mode both amounting to €120 . PDF versions of newspapers are quite expensive in the United Kingdom and France (on average €143.3 and €200 respectively) and cheap in the Netherlands and Belgium (on average €102.5 and €107 respectively).
In order to check whether the print version really is a benchmark, we calculated the correlations between the price of an annual print subscription on the one hand and the price of an annual online subscription and a subscription to the PDF version on the other. As expected, there is a strong positive relationship between the first two variables: the higher the price of the print subscription, the higher the price of the Internet subscription (ρ = .734). However, no significant relationship was found (ρ = -.013) between prices for annual print subscriptions and annual PDF versions. This is surprising, as the PDF version is actually closer to the print version than a subscription to the Web site. The first finding, however, at first sight suggests that newspapers that are more expensive have a higher fear of cannibalisation. A competing explanation is that these newspapers might have more/better online content that justifies a higher price for the online subscription. Our result might also be driven by the fact that some newspapers and/or countries are simply more expensive overall, i.e., both for print and online. In order to remedy the latter problem, we also calculated ratios between online and print prices (see Table 1) .
Table 1: Ratios of annual subscription prices. Country averages Annual subscription prices (online/print) Ratio annual subscription prices (PDF/print) Belgium 61.7% 46.5% The Netherlands 55.4% 45.4% Luxembourg N.A. 88.6% France 75.1% 71.4% Germany N.A. 32.8% Italy 66.7% 48.1% Spain 28.1% 42.4% United Kingdom N.A. N.A. Overall 57.4% 53.6%
Apparently, a low fear of cannibalisation exists in Spain, where the online subscription costs on average 28.1 percent of the print version. On the other hand, newspapers in France and Italy fear channel spillover the most, as the online version costs on average 75.1 and 66.7 percent of the print version, respectively. As for the PDF version, strong fear of cannibalisation is present in Luxembourg (88.6 percent) and France (71.4 percent). Furthermore, it is remarkable that Spain is the only country where the fear of channel spillover is higher for PDF versions than for online subscriptions (judging from a ratio of 42.4 percent, compared to 28.1 percent for online subscriptions). In all the other countries analysed, the danger of cannibalisation is more prevalent for online subscriptions than for PDF versions.
Figure 11 plots the ratio between the prices of the online and print subscription formulas which we take to be an indication of the fear for channel spillover against the offline size of the newspaper (measured as a proportion of circulation to population), and reveals a negative relationship (ρ = -.600) . Concretely, the smaller a newspaper, the higher the fear of channel spillover (i.e., the higher the ratio).
Figure 11: Relationship between offline size and ratio of the annual subscription prices.
Finally, we also examined the relationship between newspapers bundling strategies and their fear of cannibalisation. More specifically, we compared the ratio of a newspapers annual subscription prices (online/print) and the decision to opt for mixed, respectively no bundling. Table 2 shows that for those newspapers adopting mixed bundling (i.e., offer subscription formulas both as a package as well as separately), there is a higher fear of channel spillover (a ratio of 74.8 percent on average) than newspapers that do not adopt any bundling strategy (a ratio of 44.7 percent on average). This seems logical; newspapers that choose mixed bundling apply price discrimination: they charge print subscribers a lower price for the online version. A higher fear of cannibalisation could be the reason for this.
Table 2: Relationship between bundling strategies and fear of cannibalisation. TYPE Ratio annual subscription prices (online/print) MIXED BUNDLING 45.8% 100.0% 31.8% 34.5% 100.0% 57.7% 89.6% 96.1% 59.9% 100.0% 122.1% 60.0% AVG 74.8% NO BUNDLING 72.4% 22.1% 34.0% 52.0% 40.2% 50.0% 42.5% AVG 44.7%
7.2: Price of the unbundled access options 
Table 3 presents a detailed overview of the prices for the unbundled access options. As can be seen, a day pass costs €1.37 on average and the median is €1.04. It is most expensive in the United Kingdom and Germany and cheapest in France . Furthermore Table 3 illustrates that the purchase of articles on a perunit basis costs €2.04 on average. Payperview is cheapest in Spain, whereas Italy is very expensive, as is France . Purchasing a PDF version on a perunit basis costs €1.22 on average. The PDF version is again cheapest in Spain and most expensive in the United Kingdom and Italy.
Table 3: Price of the day pass, payperview and one PDF version overview on a percountry basis (in €). DAY PASS PAYPERVIEW ONE PDF VERSION COUNTRY AVG N MEDIAN MIN MAX AVG N MEDIAN MIN MAX AVG N MEDIAN MIN MAX BELGIUM 1.20 5.00 1.00 1.00 1.50 1.13 4.00 1.00 1.00 1.50 THE NETHERLANDS 1.43 3.00 1.15 1.15 2.00 1.18 2.00 1.18 1.00 1.35 ITALY 5.00 2.00 5.00 5.00 5.00 1.60 3.00 1.80 1.00 2.00 LUXEMBOURG 1.00 4.00 1.00 1.00 1.00 FRANCE 0.90 1.00 0.90 0.90 0.90 2.36 6.00 2.00 1.50 4.19 1.16 5.00 1.00 0.90 1.68 GERMANY 2.20 1.00 2.20 2.20 2.20 1.50 1.00 1.50 1.50 1.50 1.00 1.00 1.00 1.00 1.00 SPAIN 1.04 1.00 1.04 1.04 1.04 0.38 3.00 0.20 0.20 0.75 0.91 5.00 0.90 0.50 1.20 UNITED KINGDOM 2.22 1.00 2.22 2.22 2.22 1.48 1.00 1.48 1.48 1.48 1.78 4.00 1.41 0.59 3.71 OVERALL 1.37 9.00 1.04 0.90 2.22 2.04 16.00 1.50 0.20 5.00 1.22 28.00 1.00 0.50 3.71
When comparing the prices of these unbundled alternatives, it is remarkable that overall payperview is generally much more expensive than a day pass or a PDF version. In France for example, payperview costs about €2.36 compared to €0.90 for a day pass and €1.16 for a single PDF. Spain and the United Kingdom are the only countries where payperview is cheaper than the day pass or PDF version. However, it should be stressed that what we report here are country averages. Hence, the results do not imply that individual newspapers set inconsistent prices.
7.3: The cutoff between payperview and subscriptions
To conclude this Section, we compare the prices summarised in 7.1 and 7.2, albeit now primarily on a newspaper basis. Indeed, as pointed out in Section 2, an important strategic challenge for online newspapers is to avoid mutual cannibalisation between online subscriptions and payperview. If individual articles are priced too low, this may result in fewer subscriptions, but high prices make payperview unattractive for occasional visitors.
In order to obtain an insight into how our newspapers approach this tradeoff, we calculated where the cutoff point lies for the 16 newspapers in our dataset offering both payperview and an online subscription. Concretely, we calculated how many articles a reader has to consume on a monthly basis for the subscription formula to be economical (Table 4). The underlying idea here is that the lower this number, the higher the fear for cannibalisation of the online subscription by payperview.
Table 4: Cutoff point between payperview and online subscriptions. NUMBER OF ARTICLES/MONTH THE NETHERLANDS FRANCE ITALY SPAIN UNITED KINGDOM OVERALL PDF SUBSCRIPTION 8.3 2 22.8 11.0 ONLINE SUBSCRIPTION 10.6 10.9 8.9 6.7 9.3
Table 4 shows that particularly Italian newspapers appear to have embraced payperview only halfheartedly, but without data on the average number of online articles a reader consults in a given time period, it is difficult to give further normative remarks on these figures. One thing that can be said, however, is that they are substantially lower than the 30 page views per month the Financial Times apparently considers to be the cutoff point between occasional and regular readers; see Section 2.
Figure 12: Relationship between offline size and cutoff point (online subscription payperview).
Again we have checked whether the offline size of a newspaper matters. Figure 12 shows that there is indeed a relationship, and that it is negative (ρ = -.727) . That is, the bigger the newspaper, the lower its cutoff point. This suggests that bigger newspapers are less payperviewminded and favour the subscription model.
Finally, we have also analysed the relationship between the price of a newspapers annual online subscription and the cutoff point between online subscription and payperview. Apparently, there is a positive relationship (ρ = .825) . That is, the more expensive an online subscription, the higher the cutoff point, favouring the payperview option (Figure 13). It has, however, to be stressed that the number of observations is limited here.
Figure 13: Relationship between annual online subscription price and cutoff point (online subscription payperview).
Although the present paper is mainly descriptive, it does yield interesting insights into the revenue models currently adopted by newspapers that enter the Internet arena.
When analysing the strategies of the 82 Western European newspapers in our dataset, we find on a first level that 19.5 percent of the sites are completely free, while 80.5 percent of the newspapers try to monetise (part of) their online content in direct ways. Within this second category, 28.8 percent offer only a chargedfor PDF version, while 71.2 percent try to generate revenue sources in other ways (too).
Concerning a second important strategic decision, the (nonexclusive) choice between online subscriptions and payperview, it can be noted that the newspapers in our dataset rely heavily on what one could call the traditional approach: 76.8 percent offer online subscription formulas, which is about 95 percent of the paying sites. As for the unbundled access alternatives, we find that 47.6 percent of all newspapers (i.e., 59.1 percent of the fee sites) offer at least one such option. Especially newspapers with a chargedfor archive are more likely to offer unbundled access options. As mentioned, the choice between subscription formulas and unbundled access options is not an either/orassessment. This is confirmed by the fact that 43.9 percent of the analysed Web sites offer both options, and thus target regular as well as occasional readers. A mere 3.7 percent of the newspaper sites offer innovative access options and nothing else. As explained, newspapers that offer online subscriptions as well as payperview have to be careful when setting their prices for the two options, as cannibalisation might arise. We find that overall the cutoff points between online subscriptions and payperview are relatively low, which seems to indicate that our newspapers have embraced payperview only halfheartedly. We also find that bigger newspapers are less payperviewminded and favour the subscription model.
A third issue that we can shed light on is the channel spillover between online and print. As explained, just as for the cannibalisation between subscriptions and payperview, newspapers can try to stave off this danger via bundling techniques and/or careful pricesetting. The first option requires what we have called metabundling decisions. Overall, 32.9 percent of the newspapers offer mixed bundling, 34.1 percent opt for no bundling and 1.2 percent choose pure bundling. Where the pricesetting is concerned, we use the ratio online/print subscription prices as an indicator of fear of channel spillover. We find that newspapers that adopt mixed bundling have a higher fear of channel spillover between the print and online subscription than those that do not adopt any bundling strategy. In this respect, it is interesting to point out that of the 32.9 percent of newspapers that adopt mixed bundling, 63 percent grant free online access to print subscribers (as does the one newspaper that opts for pure bundling). This suggests that these papers still focus strongly on the print edition. We also find that newspapers with a bigger market share overall have a lower fear of cannibalisation.
Finally, an important finding of our research is also that for each of the strategic decisions mentioned above, large differences exist between the countries analysed. Firstly, indirect revenue sources such as advertising are still frequently tapped in Luxembourg and Italy, a fact reflected in the provision of completely free Web sites. On the other hand, it is particularly newspapers in Spain, France and Belgium, which try to directly monetise their online content.
Online subscriptions are offered the most in Spain and the United Kingdom, while they are least present in Italy, the Netherlands and France. On the subject of metabundling, it is particularly newspapers in Germany, the Netherlands, France and Belgium, which opt for mixed bundling strategies, while those in the United Kingdom and Spain mainly refrain from bundling. Furthermore, judging from the fact that the proportion of the online subscriptions price is only 28.1 percent of the printed version on average, Spanish newspapers in particular do not fear channel spillover so much. The opposite is true for France and Italy (with a ratio of 75.1 percent, resp. 66.7 percent).
Where unbundled access options are concerned, the picture can be summarised as follows. Day passes are not offered at all in Luxembourg, the Netherlands or Italy, whereas they are frequently available in Belgium. Xdays passes are nonexistent in Luxembourg, the United Kingdom and Spain, yet are relatively common in the Netherlands and Italy. Creditscards are not offered in the Netherlands, but are in France, Italy and Luxembourg. Finally, payperview is not available in Belgium and Luxembourg, but well established in France and Spain. Overall, the unbundled options do not have a high penetration rate in Luxembourg, the United Kingdom and Italy. In 50 percent of cases archives are accessible without charge (especially in the Netherlands, the United Kingdom and Germany), whereas payment is required in slightly less than 50 percent of cases (particularly in Belgium and France). This might again also explain the strong presence of certain unbundled access alternatives in the latter two countries, as supported by the calculated correlations.
Taken together, our findings suggest that besides the cannibalisation issue, local market circumstances such as the relative importance of advertising revenues, Internet penetration, and even mimicking behaviour play a major role. We intend to analyse this in future research.
About the authors
ValérieAnne Bleyen is a PhD student at the Free University of Brussels, Belgium, at the Department of Economic, Social and Political Sciences and Solvay Business School. Her research focuses on digital goods and online payment instruments. Besides that, she is involved in an interdisciplinary research project on Flemish EPublishing Trends (FLEET).
Leo Van Hove is Associate Professor of Economics at the Vrije Universiteit Brussel (Free University of Brussels), where he teaches courses in monetary economics, Internet economics, and ecommerce. His current research interests include emoney, currency usage, and epublishing.
The authors wish to thank Steve Brown (The Newspaper Society, United Kingdom), Marc Thiltgen (TNSIlres, Luxembourg), Sabine Ozil (SPQN, France), Marlene Blonk (Cebuco, the Netherlands), Tamara van Langenhof (Vlaamse Dagbladpers, Belgium), Federico Megna (FIEG, Italy), Heike Laumer (BDZV, Germany), Maribel Matallanas (AEDE, Spain), Ruth Stoker (Media and Journalism University Centre Barnsley, England), Ellen Loix and Ralf Caers (VUB, Belgium) for their constructive comments as well as all the newspapers that helped to resolve more detailed and newspaperspecific questions.
1. NY Times to drop charges for website, Financial Times, 18 September 2007.
2. Murdochs choice: Paid or free for WSJ.com? Wall Street Journal Online, 19 September 2007.
3. FT.com pioneers change to charging, Financial Times, 1 October 2007.
4. Murdochs choice: Paid or free for WSJ.com? Wall Street Journal Online, 19 September 2007.
5. More media, less news, Economist, 24 August 2006.
6. Newspaper Association of America, Advertising expenditures, at, http://www.naa.org/TrendsandNumbers/Advertising-Expenditures.aspx, consulted on 2 October 2007.
7. Porter, 2001, p. 73.
8. This may be due to the fact that the authors analyse different time periods, different countries or different types of content (i.e., magazines and/or newspapers).
9. We primarily relied on the list of national daily paidfor newspapers in the report of the World Association of Newspapers (2006). For more specific data we contacted the national coordinating organisations for each country. For Spain, we collected data from the portal site (http://www.spain-newspaper.com) because we were unable to obtain a certified classification for the concept of national, daily, paidfor newspapers. The list of newspapers is available upon request from the authors.
10. According to the report of the World Association of Newspapers (2006), the number of regional and local paidfor newspapers with a daily circulation in 2005 amounted to 61 in France, 139 in Spain, 71 in Italy, 94 in the United Kingdom, 358 in Germany, 26 in the Netherlands, 0 in Luxembourg and 18 in Belgium. The market share of these newspapers, calculated as a proportion of the total circulation per country, amounts to 73 percent in France, 60 percent in Spain, 36 percent in Italy, 29 percent in the United Kingdom, 92 percent in Germany, 54 percent in the Netherlands, 0 percent in Luxembourg and 31 percent in Belgium. In other words, the relative importance of the regional and local newspapers varies drastically from country to country. Hence, the same is true for the newspapers in our dataset.
11. In France for example, Bilto, La Gazette des courses and Matin courses are intended for horselovers and Quoti is a newspaper for children.
12. Obviously, a whole series of alternative revenue sources exists for newspapers. However, in our analysis we do not consider extra services that online newspapers offer, such as stock quotes, personalised news (letters), email accounts, RSS features, news via SMS and so on.
13. In practice, newspapers often offer the news on their Web site in abridged form, providing, for instance, only the headlines or a selection of articles. In other words, in these cases the online content in HTML format does not completely cover everything in the print version of the newspaper. When a newspapers Web site is categorised as free and considered to be completely gratis, this thus does not necessarily imply that the newspaper is giving away its entire print content on the Internet.
14. In our analysis, 20 percent of the U.K. newspaper Web sites are completely free whereas Herbert and Thurman find that all the newspapers charge for something. However, in their context this can not only mean charging for news, columnists, the archive, the digital edition, but also for email alerts, mobile services, crosswords or games. Since we exclude the latter four monetising options (see note xiii), our results do not clash with those of Herbert and Thurman.
15. It is also possible that apart from the print version, there is a bundle available consisting of the offline and online version, while the online version is not for sale separately. For instance: bundle for €320 per year, only print for € 250 per year. Since we focus on the access possibilities to online content in this research paper, we classify this option as tying as well. Additionally, when there is only access possibility to the online facilities when one has subscribed to the print version, this is again categorised as tying. The online version is not available as a stand-alone product in the last case. The print version in fact encompasses the online facilities without a surplus being charged.
16. Newspapers that solely offer a type of subscription formula in the form of a PDF subscription, are located under no online subscription.
17. Again we stress that it is perfectly possible that not all the printed articles are available online.
18. This card is actually a kind of electronic wallet that keeps track of the online balance.
19. Customised bundling is a specific form of packaging whereby the firm determines only the size and the price of the bundle it will offer to customers, and the customers then decide on the contents. A music shop, for instance, can offer three CDs (chosen personally) for €15.
20. Some newspapers also offer a creditscard for the purchase of online PDF versions of the newspaper. This certainly is not payperview as it concerns the entire newspaper, rather than individual articles. This is again a form of customised bundling but now on a more aggregated level: one buys a bundle of X PDF versions at a discount, and this bundle can be redeemed at will.
21. We come to this conclusion both when we compare with the total number of newspapers, as well as with the paying sites.
22. Note that our result for the U.K. is identical to that of Herbert and Thurman (2007), who find in the same period that 40 percent of the analysed national newspaper sites have a chargedfor archive.
23. For the record, newspapers without an archive were excluded from the calculations.
24. Note that it seldom concerns one and the same newspaper (i.e., simultaneous offer of online subscription and PDF subscription) and in no case does it concern the same number of newspapers.
25. For Luxembourg and Germany, we have no data on the annual prices for online subscriptions, For the United Kingdom, no prices for the print subscriptions were available on the Web sites.
26. We also calculated the correlation between the ratio of the annual subscription price (PDF/Print) and the offline size of the newspaper, but it proved not significant (ρ = .12).
27. Regarding the Xdays pass, drawing conclusions as to price setting is extremely hard for several reasons. Firstly, only a limited number of the newspapers offer it. In addition, there are large variations in the duration of access authorised (e.g., 3 hours, 10 days, 50 days, etc.). A possible solution might be to convert the price for X days to a price for 1 day, but this would give a blurred result since the price generally decreases with the duration. Because of these difficulties, we have refrained from making price comparisons between the countries. For the creditscard too, it is difficult to analyse the prices, this time due to the large number of variations that exist in the quantity of credits offered at a certain price. In addition, the number of credits that has to be paid for the purchase of one article also varies from newspaper to newspaper. In many cases the number of required credits depends on the length of the article. Hence it is not straightforward to compare the prices of the creditscards for the analysed newspapers and countries.
28. Luxembourg, the Netherlands and Italy are missing here since newspapers in these countries do not offer a day pass.
29. Belgium and Luxembourg are missing here since none of the newspapers offer a payperview option.
30. We also calculated the correlation between the cutoff point (annual PDF subscriptionPPV) and the offline size of the newspaper, but it proved not significant (ρ = .438).
31. We also calculated the correlation between the cutoff point (annual PDF subscriptionPPV) and the annual PDF subscription price, but it proved not significant (ρ = -.494).
Yannis Bakos and Erik Brynjolfsson, 1999. Bundling information goods: Pricing, profits and efficiency, Management Science, volume 45, number 12 (December), pp. 16131630. http://dx.doi.org/10.1287/mnsc.45.12.1613
Yannis Bakos and Erik Brynjolfsson, 1997. Aggregation and disaggregation of information goods: Implications for bundling, site licensing and micropayment systems, Working paper, Graduate School of Management, University of California.
Hsiang I. Chyi and Dominic L. Lasorsa, 2002. An explorative study on the market relation between online and print newspapers, Journal of Media Economics, volume 15, number 2, pp. 91106. http://dx.doi.org/10.1207/S15327736ME1502_2
Barbara Deleersnyder, Inge Geyskens, Katrijn Gielen and Marnik G. Dekimpe, 2002. How cannibalistic is the Internet channel? A study on the newspaper industry in the United Kingdom and the Netherlands, International Journal of Research in Marketing, volume 19, number 4 (December), pp. 337348. http://dx.doi.org/10.1016/S0167-8116(02)00099-X
Lapo Filistrucchi, 2005. The impact of Internet on the market for daily newspapers in Italy, Working paper, Department of Economics, European University Institute, San Domenico, Italy, at http://center.uvt.nl/staff/filistrucchi/filistrucchi2.pdf, accessed 23 November 2007.
Peter C. Fishburn and Andrew M. Odlyzko, 1999. Competitive pricing of information goods: Subscription pricing versus payperuse, Economic Theory, volume 13, number 2 (February), pp. 447470. http://dx.doi.org/10.1007/s001990050264
Peter C. Fishburn, Andrew M. Odlyzko and Ryan C. Siders, 1997. Fixed fee versus unit pricing for information goods: Competition, equilibria and price wars, Working paper, University of Minnesota.
Matthew Gentzkow, 2007. Valuing new goods in a model with complementarity: Online newspapers, American Economic Review, volume 97, number 3 (June), pp. 713744. http://dx.doi.org/10.1257/aer.97.3.713
Jack Herbert and Neil Thurman, 2007. Paid content strategies for news Websites, Journalism Practice, volume 1, number 2, pp. 208226. http://dx.doi.org/10.1080/17512780701275523
Ulrich Kaiser, 2006. Magazines and their companion Websites: Competing outlet channels? Review of Marketing Science, volume 4, at http://works.bepress.com/ulrich_kaiser/1/, accessed 23 November 2007.
Ulrich Kaiser and Hans C. Kongsted, 2005. Do magazines companion Websites cannibalise the demand for the print version? Working paper, Centre for European Economic Research, Mannheim.
E. Noam, 2006. Strategies for newspapers when the free generation of readers meets the expensive generation of news, Presentation at the Thirteenth World Editors Forum, 47 June 2006, Moscow, Russia.
Koen Pauwels and Enrique Dans, 2001. Internet marketing the news: Leveraging brand equity from marketplace to marketspace, Journal of Brand Management, volume 8, number 4 (May), pp. 303331. http://dx.doi.org/10.1057/palgrave.bm.2540030
Michael Porter, 2001. Strategy and the Internet, Harvard Business Review, volume 79, number 3 (March), pp. 6278.
Carl Shapiro and Hal R. Varian, 1998. Information rules: A strategic guide to the network economy. Boston, Mass.: Harvard Business School Press.
Daniel H. Simon and Vrinda Kadiyali, 2007. The effect of a magazines free digital content on its print circulation: Cannibalisation or complementarity? Information Economics and Policy, volume 19, numbers 3-4, pp. 344361.
Florian Stahl, Marc F. Schäfer and Wolfgang Maass, 2004. Strategies for selling paid content on newspaper and magazine Web sites: An empirical analysis of bundling and splitting of news and magazine articles, International Journal on Media Management, volume 6, numbers 1 and 2, pp. 5666.
World Association of Newspapers, 2006. Trends in newsrooms 2006, Annual report, at http://www.worldeditorsforum.org, accessed 12 July 2006.
Paper received 27 October 2007; accepted 20 November 2007.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 2.0 Belgium License
Western European newspapers and their online revenue models: An overview by ValérieAnne Bleyen and Leo Van Hove
First Monday, Volume 12 Number 12 - 3 December 2007