So, you’re telling me that a bill that provides disincentives for Facebook to cooperate disincentivised them from cooperating? And we’re supposed to be mad at Facebook for this‽ Anything to protect Rupert and Peter, I guess.
If news.com.au were to (hypothetically) place a link to iosef.org in one of their articles, on a webpage covered in ads, do they owe me money? That is, have they “stolen” the value of the ad unit I’ve placed on this page?
Even if most people read the article and choose not to click-through to my website, even if News earn thousands from their article, it would be utterly ridiculous for me to email Rupert and ask for a cheque.
If I do that, maybe their entirely hypothetical hagiography about this blog might receive an update. One that deletes the link to it!
With all of this in mind, the answer to my earlier question is an emphatic, capital-N, No.
In this hypothetical example, I would have made an unreasonable demand for compensation for News’ decision to create a hyperlink to my site and they decided not to deal with my chicanery.
In a similar vein, this week, Facebook blocked all users from sharing links to Australian news websites on their platform this week and hid foreign news sites from Australian users as the ridiculous news media bargaining code – or as I call it, the link tax – passed the House.
The code seeks to force Google and Facebook to compensate news media companies for the advertising that appears in their platforms that appears near hyperlinks on their platforms (which often include links to news articles) through individual negotiated deals with news companies at the threat of “final-offer arbitration” and further regulatory action.
If Facebook were to remove such links, as they did, the thinking goes that they wouldn’t be exposed to the code. A fact that they seized on with devastating effect.
Was Facebook’s reaction a little dramatic? Well, yes. A little is perhaps an understatement considering the number of government, heath, and NGO pages that saw their content suddenly pulled. They absolutely should have done more to prevent such pages being caught in their filter.
However, let’s not, as the media seems to have done, let this hissy fit distract us from the real problem which is the very concept of the bill itself.
This post will first examine the lie of big tech “news theft” which is the justification for the tax. It will then examine how Facebook’s effort to avoid “stealing the news” through links has led to actual news theft by its users, before examining a framework through which we can understand Facebook’s ban and applying it to a potential alternative.
The government-sanctioned, corporate media approved, myth of the great platform news heist
Imagine if Hachette or Penguin Random House complained of Google Search’s ability to resolve bibliographic information from ISBNs by saying that “Google steals books”. They would, I hope, be rightfully laughed out of the ACCC’s NewActon offices.
However, this is the exact line of harebrained moon logic on which the government’s proposal for the link tax relies. The only differences being that instead of ISBNs we’re talking about hyperlinks, and instead of publisher-provided bibliographic data we’re talking about publisher-provided metadata from news articles.
Both ISBNs on books and URLs to news articles are unique identifiers for copyrighted works which can be used to resolve metadata about the works they identify. In both cases, this metadata put there by publishers for explicit the purpose of making their works easier to find and acquire.
Readers do not get the full value out of a book by knowing its title, author, and miscellaneous publication data. They get the full value of books by reading their contents. One can read a small preview on Google Books, but it’s rare that they can read enough to replace the value of seeking out the full volume.
Similarly, news readers are unlikely to get the full values out of public interest journalism from a headline, the featured image, and (at most) two sentences.
Nonetheless, Nine and News Corp have disseminated the myth that Google and Facebook “steal” news. A myth that seems to be accepted by the entire media landscape uncritically because they all have something to gain: Google and Facebook’s billions.
Crikey’s Bernard Keane says the ACCC themselves debunked the myth of the news heist, at least in the traditional sense of “stealing” content (that is, copyright infringement), but which for some reason the agency continues to pursue anyway.
The report Keane cites to for evidence of this claim has this to say:
Digital platforms often reproduce headlines and snippets of content from original news media articles created by journalists and media businesses. Generally, digital platforms’ use of article headlines is unlikely to infringe copyright protections in Australia. This is because many headlines are concise statements of facts and therefore headlines alone are unlikely to be copyright protected.
Digital platforms reproducing a snippet of a copyright-protected news article does not infringe copyright protections if the snippet does not reproduce a substantial part of the article. [It usually doesn’t.]ACCC. 2019. Digital Platforms Inquiry : Final Report (CC-BY): p. 260. Emphasis added, footnotes and citations removed, square bracket text inserted by me.
Not exactly a ringing endorsement of “theft” theory.
The corporate media’s manufactured consent, if I may, has resulted in almost no articles being published which feature intelligent critique of the Bill in the media.
Apart from Crikey‘s running condemnation, the only criticism I’ve seen in the mainstream Australian media consists of three articles.
The first was an ABC article which discusses consternation from small business owners over Google’s now-abandoned threat to leave Australia.
The second was a Conversation article referring to World Wide Web inventor Sir Tim Berners-Lee’s Senate committee submission in which he sounds the alarm about the very concept of putting a price on links (as I am), before the article’s author bleats the same line of propaganda as outlined above.
The last was an op-ed in the Sydney Morning Herald from digital rights expert Samantha Floreani who argues, on similar lines to this article, that the news block is “just one of the possible bad outcomes of this proposal”.
Foreign media, such as American tech blog The Verge, has done a much better job, probably because foreign media companies have absolutely nothing to gain from the Code.
If links are not theft, then what are they?
The answer is what the name suggests: a link. I called it an identifier earlier.
When most people accessed news on Facebook (before it was blocked), it was through a link. When news companies
lie complain of having their news “stolen”, they never mean that Facebook took a substantial portion of the text of their articles without permission.
Rather, they mean at least one of two things, either:
- That Facebook loads a preview of articles when a link is included in a post which somehow devalues the article; and/or
- That because Facebook does not share the ad revenue from their ad units that appear next to every post – whether they link to news articles or not – this has the effect of “stealing” ad revenue.
The first is complete bullshit. Nested within the <head> tag in the HTML code of every single one of the news article web pages that are complained of are tags that contain the metadata Facebook gleans from the page to generate those “preview” cards you see when you paste a link into a post.
Those tags, whether they be the title tag, the various standard forms of <meta> tags, or extensions to the standard like Facebook’s own Open Graph tags, are all there for the explicit purpose of being resolved by web crawlers like those at Facebook and Google.
As I discussed earlier, it can’t reasonably be said that this metadata is the only valuable thing on the news companies’ webpages. When Facebook scrapes this de minimis data for their link cards, can they really be said to have “stolen” the content behind the link? No.
The second point seems to completely ignores that, for as long as large numbers of people use Facebook, the advertisements would be on the platform regardless of whether people are posting links to the news or posting pictures of their kittens.
The value of Facebook advertising is driven more by the fact that a lot of people use Facebook’s website and app more than it is driven by the fact that it happens to be a good news aggregator.
Facebook points to their claim, which for lack of rebuttal I’ll guess I’ll just have to accept as fact, that only 4% of their News Feed traffic is allegedly contains news as evidence of this.
The Guardian‘s version of this argument is that:
“The code is not intended to compensate news companies for the loss of advertising income due to digital disruption. It is intended to facilitate fair payment for the benefit the platforms receive from news.”“Why Google and Facebook are being asked to pay for the news they use – explainer“. (5 September 2020). The Guardian.
As addressed above, Facebook doesn’t “use” their news as their headline suggests. Its users link to it and Facebook displays the publisher-provided metadata. Also as addressed above, Facebook seems to get only a small benefit for what the news companies are demanding from them. (As an example, Nine who wanted the giants to pay $600 million to publishers, which was not enough for News who wanted them to pay $1 billion).
Though, what The Guardian denies is the reason for the link tax raises an interesting point.
If loss of potential ad sales to platforms like Facebook is the problem, then surely the solution is to redistribute some of the money that is said to be “tied up” in these platforms. Not as a condition of hosting users’ links to news content, but as a given.
This could either come through a revenue sharing program like YouTube’s partner program[⨇], or (as I will argue for later) a super advertisers’ tax, the revenue from which is to be used for the provisioning of grants for public interest journalism.
The government’s solution has only created very much undesirable consequences.
How stopping the Facebook “steal” has only created actual stealing
The Facebook’s ban of news links has spawned a wide range of creative workarounds amongst the site’s users. All of which, I imagine, would be absolutely horrifying to the publishers who have pushed so hard for the link tax.
The first I’ll address is linking to another, non-news, page which itself hosts a link to the desired news article.
ANU’s student newspaper Woroni does this by posting links to its posts on Twitter on its Facebook page.
I imagine a for-profit publisher would be horrified that this method doubles the alleged problem. Now it’s not just Facebook making money from ads next to users’ links to their content. Now whatever site people use to do this workaround could cash in too. On mobile, Twitter doesn’t display ad sunder tweets, but on desktop they show their promoted trending hashtags all the time.
It is not copyright infringement under the laws of any civilised country to post a hyperlink to an authorised reproduction of a copyrighted work – such as a newspaper article posted to that newspaper’s website.[*]
It is copyright infringement, however, to make an unauthorised reproduction of a copyrighted news article which has now become a popular means of sharing news on Facebook.
Despite this, I’ve seen plenty of people simply paste the full text of copyrighted articles into their posts and comment sections.
Even I’ve done this (before I noticed the trick of linking to Twitter). I posted screenshots containing a full copy of a Crikey article to the comment section of one of my posts.
Screenshots containing full articles, and literally just copying the text of articles, is copyright infringement.
Ironically for the media companies, Facebook themselves can’t be sued for this actual theft unless it’s shown that they actively participated in it.
Part V, division 2AA of the Copyright Act 1968, and part 6 of its regulations, holds service providers like Facebook immune from copyright actions where they have a notice-and-takedown procedure and a policy to deal with repeat offenders. Facebook has such both. As a result, affected news companies will have to pursue individual users, if they care enough.
Is there a better way?
Of course there is. Otherwise, I wouldn’t have bothered writing this post.
But first, some background (unless you click here or scroll to skip it).
Policymakers, whether they be departments, lobbyists, stakeholders, or politicians, should consider the calculations that could be made on the opposing side of the policies they draft.
I feel that, in this example, everyone involved in the drafting and passage of the link tax has just ignored this crucial step and just arrogantly assumed that everyone targeted by the tax will just play along.
As an example of considering calculations, look at income tax.
In Australia, individuals don’t calculate how much tax they owe themselves. They don’t because such a concept would be rife for abuse and there would be a massive incentive – that being not paying tax – to cheat by deliberately under-calculating and underreporting income.
Instead, the ATO makes income payers (for lack of a better term), such as employers, Centrelink and banks, report income earned and the ATO does the calculation on their end. This provides a paper trail and a projected tax liability to compare against when somebody does try to underreport their earnings.
The policy considers the possibility that people might try to avoid income reporting by not getting the (ostensibly) optional tax file number by forcing income payers to collect the highest rate of income tax until they provide one.
The example above, income tax administration, is a policy that assumes its targets will cheat (in the game theory sense) and increases the cost, while decreasing the incentives, of doing so.
So, what are the calculations for Google and Facebook?
Let’s use the framework described in the last section to look at what the respective calculations in place for the link tax would be from the perspective of Google and Facebook.
Remember, the core proposal of the code is that to continue showing links to news companies’ websites, Google and Facebook would have to, one way or another, pay up.
For Google, for whom a link tax makes a bit more sense (even if still antithetical to the infrastructure of the Web), a link tax means choosing between continuing to be easily the most useful and comprehensive index of the World Wide Web or losing a good chunk of the search results that people are looking for and, as a result, becoming useless.
For Facebook, the link tax creates this calculation: Go without regulated news links which they say only account for 4% of their traffic or pay an unreasonable amount of money to continue supporting an unreasonably small percentage of traffic.
Having considered the above, you can see why Facebook quickly dropped anything that could possibly be regulated while Google would whip out the chequebook.
You said there’s another way, so get to it already!
Right, sorry, got carried away.
Now, let’s consider an alternative policy.
If the problem is that all this ad money that could be going to media companies is now tied up in Google and Facebook ad-buys, then surely the policy response should have some basis in that advertising rather than in invented notions about the value of metadata gleaned from links.
Perhaps some kind of super profits advertising tax, administered similarly to GST, could work to recover some of the lost “advertising base” that media companies are concerned about.
I believe, though I invite constructive criticism in the comments, that this would decrease the cost of compliance to something below the cost of cheating.
Set some kind of before-tax ad sales revenue threshold – perhaps $500 million – after which point companies have to start putting aside a percentage of sales income – say, 10%, for the ATO.
I see it as unlikely that Facebook and Google would forgo billions of dollars in advertising revenue for a few million in taxes every year by shutting off the ads in the same way that Facebook shut off its delivery of new slinks.
And just imagine the ridicule they’d get from the public for threatening Australians with less ads.
I think Google and Facebook wouldn’t say a word about this hypothetical tax if it were backed by a mechanism to force ISPs to block requests to their advertising servers, similar to how copyright law allows the courts to block websites that facilitate copyright infringement.
Such a mechanism would incentivise compliance by making it a matter of economic survival. Advertising will always be desired by the market and the incentive to comply in such a system is that someone else will.
However such a tax is implemented, whether by replacing GST with this ad tax (requiring state/territory agreement), or by making a tax that is similar to GST specifically for these large advertising networks, it is surely a better option than a policy that entirely disincentivises distributing news links.[#]
The money collected from this tax could then go towards a public interest journalism grant scheme preferably run by an independent body like the Australia Council, though the Department has one that is okay I guess.
A 10% tax on only Facebook’s ad sales (not even considering Google’s) would have brought in roughly $67.4 million in 2019/20.[$] This is a little more in one year than the $60.4 million the government appropriated from general revenue to run ACMA’s Regional and Small Publishers Innovation Fund over three years in 2018.
Will the government make a better plan?
I’m not holding my breath that the government will back down on this ridiculous link tax. The stakes are now simply too high, and Facebook is already back at the negotiating table as I write this.
In the meantime, Australia remains in an internationally embarrassing game of chicken with Mark Zuckerberg: bickering over a policy that was supposed to give news publishers power but, thanks to drafters and stakeholders blinded by hubris, gave Facebook incentives to wield even more.
The government and the media are declaring that Facebook is “working outside the rules” when this news boycott is a logical consequence of the rules themselves. Rules which seek to tax links, rather than redistribute advertising wealth to embattled public interest journalism. Rules which one can avoid by not dealing with regulated links.
Plenty of people have been misled by mainstream media lies about article metadata being of the same value as an actual article in ways that would never apply to other forms of media.
Other, less ineffectual, policy options are available but now the government chooses to keep their fingers in their ears lest they look weak in Facebook’s wake.
I will die on the hill that the link tax is a silly policy that is based on a lie sowed by the media and mandated as truth by the government. Even if it looks like it will survive with near-unanimous support in Parliament and even if this remains an apparently deeply unpopular take.
Disclosure: iosef.org displays a Google AdSense ad unit that appears to users who are not logged in. Google does not have any editorial influence over the content I put up here.
[⨇] YouTube, unlike Facebook, hosts publishers’ content rather than simply linking to it which could be a crucial point of distinction in this debate.
[*] There don’t appear to be any Australian court cases on this matter, but the Federal Court of Canada case Warman v Fournier [at para 37] held that a website administrator who posted a hyperlink to a photograph hosted on the photographer’s own website did not infringe copyright.
Conversely, a Full Bench of the Federal Court of Australia found, in the case of, Cooper v Universal Music, that a website administrator whose website linked to unauthorised reproductions of copyrighted sound recordings did infringe copyright. (Kallenbach 2012)
[#] I focus on similarity to the GST because Google already pays it (or at least they claim to). No idea if Facebook does too.