Illustration: Toby Morris
Illustration: Toby Morris

BusinessOctober 3, 2021

Google and Facebook are brilliant and dangerous and entirely out of control

Illustration: Toby Morris
Illustration: Toby Morris

Google and Facebook have transformed the world, made our lives better and brought enormous challenges. Yet despite their impact on everything from business to the mental health of teenagers, they are less regulated than your local takeaways, writes Duncan Greive.

Illustrations by Toby Morris and Archi Banal

But society soon realised its potential, and created infrastructure to support it. Roads to drive on, petrol stations to refuel at. Mechanics to fix them when they broke down. And, before long, expanded emergency departments to deal with injuries. Cemeteries became dotted with those who died driving, or hit by drivers. Some of this cost was borne by individuals, much of it by the state.

Over time we introduced tools to reduce the downside – speed limits, drink driving laws, better roads, emissions parameters, airbags, crumple zones, seatbelts – while still retaining the social and economic benefits of all that mobility. Debates sprang up about how to manage the tension between private road users and public transport systems, which the car mostly won. 

Some imagine a world in which the car’s victory was less comprehensive. They look back wistfully on the tram systems which got people around Tāmaki Makaurau in the 40s, and advocate for cycleways and light rail now. We wrestle still today with the long-term consequences of decisions or delays of decades ago – climate change caused by excessive reliance on fossil fuels, gridlock on our roads and cities which sprawl out rather than building up. Lives cut short. We live with the consequences of inaction for a long time. Sometimes forever.

We’re at a fateful intersection again

This is not a story about cars. Instead it’s an attempt to make a good faith reckoning with new technologies just as important as the private vehicle: social media and search engines. They have transformed communication and the distribution of information, and are at least as impactful to the way we live now as the car was to the 20th century. 

Both social media and search engines have provided extraordinary benefits to society, of which we’re well-appraised. We can connect and kōrero with friends and whānau all over the country and the world, without friction and essentially cost-free. We can bring up information with a few taps or keystrokes that would previously have required a trip to the local library, if it were available at all. Small businesses can set up shop and have a highly efficient advertising product to reach customers – either those actively looking for their goods or services (through search) or which are highly relevant (through social media). And on and on and on.

It’s unnecessary to point out that this has changed our lives, totally and irrevocably. One small window in: roughly half of those reading this likely came on a journey commenced on Facebook or Google. Yet these technologies are, by historic standards, very young: while neither quite has an outright monopoly in their respective areas, each is so dominant as to function as a verb.

Google was founded in 1998 in California, Facebook in 2004 in Massachusetts (though it moved to within a few miles of Google that same year). In some respects, this is the argument for treating them gently. To return to my analogy, if we compare the tech giants’ development timeline to that of the car, we’re in about WWI, when the Model T Ford was the height of motoring sophistication.

Image: Archi Banal

Social media and search engines are indeed young – founded within most of our lifetimes, and many of us remember life before they arrived. But they differ from the car in two crucial respects. One is that the infrastructure they rely on is already highly evolved and widely distributed. Hardware companies like Apple, Microsoft and Samsung put supercomputers in our bags and pockets. Our government presciently funded the rollout of a large broadband network, ensuring fast, relatively affordable broadband which is accessible to most (though by no means all). And companies like Spark and Vodafone built out mobile internet to ensure there were few places on the whenua where we couldn’t access their services.

Which is to say that we as a society funded an enormous infrastructure build which ensured that these services are accessible to a very large majority of us (those on the wrong side of the digital divide need help across, though that’s another story). And because they’re useful and free – a crucial and challenging element of their success to which I’ll return – they’re wildly popular, with a large majority of New Zealanders using them each day. All this means that while those companies are very young, they’re also ubiquitous – essentially they are the internet, to many of us.

This explains why, though still in their teens or early 20s, Facebook and Google’s combined market capitalisation – what it would cost to buy them – is over NZ$4 trillion. In recent years their combined annual revenues have blasted past New Zealand’s GDP. To put it another way, they’re respectively the fourth (Alphabet, Google’s parent company) and sixth (Facebook, which also owns Instagram and WhatsApp) most valuable businesses in the entire world. Which is to say that while they are young in years, they’re also powerful in ways few businesses ever have been. 

One more thing to know about Google and Facebook: for all their phenomenal, unprecedented scale and power, they are also almost entirely untaxed and unregulated in this country. In fact, we know very little about them. They don’t declare how much money they really make locally, and despite being among the most profitable companies in the world, they make almost no contribution to our tax base here, siphoning a massive and fast-growing fire-hose of money back to Northern California but leaving little here to pay for the public services we require to function as a society.

This brings me to the point of this essay. To try and assess the impact of these businesses on us as a country, and ask whether, as a community, this current arrangement makes sense. I will take it as a given that we enjoy tremendous benefits from freely accessible search and social media. As noted above, these are so widespread as to hardly need explaining.

Instead I’m going to try to grapple with what economists call the negative externalities, or even just odd elements that seem screaming for oversight. The social and search equivalent of the smog, car crashes and communities sprawled out and carved into pieces that came with the car. This will necessarily be just a tiny sliver of such impacts – their scale brings with it enormous complexity, which they use as a shield to resist any attempts at reform. 

All this will ultimately lead to asking if it makes sense for us to continue to allow two of the biggest and most profitable companies the world has ever known to operate here with less regulation and taxation than the average takeaways or taxi.

Vaccinations and information disorder

I write this from a locked down Tāmaki Makaurau, so naturally, this is front of mind. After an astonishing surge in vaccinations as we grappled with the early stages of the delta outbreak, we now watch a dispiriting plunge in the number of first vaccinations, despite our eligible population being well short of the 90% figure often touted as a key milestone toward greater freedoms. Those not receiving the vaccine are a complex group, containing a tiny number of people with legitimate health reasons to avoid it, and a far larger group who could safely take it but don’t want to.

They range from a persuadable group of the vaccine hesitant, to a very loud and determined group of “anti-vaxxers”, strongly opposed and committed to recruiting and retaining members. Despite the science being settled, much of this group is suffering from what is known as “information disorder”, an umbrella term for disinformation, misinformation and malinformation. We saw a particularly visceral example of this just last week, when The Spinoff reported that anti-vaxxers swarmed the prime minister’s stream of a 1pm press briefing, showering our most influential public figure’s feed with falsehoods regarding the vaccine, just as she was on stage exhorting people to get vaccinated. 

While Facebook is making a concerted effort to get its arms around anti-vax content, its efforts have been overwhelmed. Perhaps this is because the vast majority of its human moderators are focused on the US. According to the Wall Street Journal, “Facebook employees and contractors spent more than 3.2 million hours searching out and labeling or, in some cases, taking down information the company concluded was false or misleading… Only 13% of those hours were spent working on content from outside the US”. When 90% of Facebook’s users are outside American borders, that’s a major problem. Particularly given that large amounts of anti-vax content is in languages other than English.

I approached Facebook to ask where its moderators for New Zealand were located in 2020, and the company failed to answer. I asked again this week, along with questions about their proficiency in te reo Māori and Pasifika languages. A Facebook company spokesperson replied “we are committed to our diverse and vibrant community in New Zealand… Our content moderation approach is global, operates 24/7 across multiple time zones, and in many different languages.” Which neatly avoided answering either question. This is sadly fairly typical of its record on transparency, but it seems safe to assume that its moderation presence within Aotearoa and in tune with our values, language and culture is minimal.

Independent journalism takes time, money and hard work to produce. We rely on donations to fund our work. If you can help, donate to The Spinoff Members

While it’s near-impossible to know to what extent Facebook and its other platforms are responsible for the spread of anti-vax content in New Zealand, the platform is undeniably a hotbed of it still, as a chilling recent Charlie Mitchell feature for Stuff showed. It leaves our government in a bizarre position: having to devote an enormously costly public health and communications apparatus to persuade people to take vaccines, a basic measure to protect themselves and their community. To do this they have to use and pay many millions of public dollars to platforms that still rife with the anti-vaccine information disorder they’re trying to counteract. And do all this while individuals and the state are bleeding billions of dollars in a lengthy lockdown which might have ended by now had Facebook prevented the spread of anti-vax content on its platforms.

The cost, both fiscal and social, is carried by all of us. There are few consequences for Facebook for allowing this spread, and the profit from advertising wrapped around all this engagement goes offshore, largely tax-free.

Instagram and teens

This essay was in part inspired by a devastating series of reports, The Facebook Files, recently published by the Wall Street Journal. They were based on a massive trove of leaked information from within Facebook, some of it centred around research it has itself carried out on the impact of Instagram, which Facebook owns, on young people. 

Tech reporter Casey Newton described The Facebook Files as the biggest crisis the company has faced since the Cambridge Analytica scandal of 2018. While Facebook has firmly pushed back against the reporting (and, as Hal Crawford noted on The Spinoff, the news media has a not uncomplicated relationship with its platforms), it has also pushed pause on the development of “Instagram for Kids”, a product it had previously been very excited about but has a name which will make most parents feel nauseous.

Slide released by Instagram this month. (Image: Instagram / Archi Banal)

The crisis is largely due to incendiary slides leaked from Facebook, one of which is blunt headlined “we make body image issues worse for one in three teen girls”. The Journal’s reporting also cited another presentation which showed that “among teens who reported suicidal thoughts, 13% of British users and 6% of American users traced the desire to kill themselves to Instagram”. Facebook has said these numbers lack context, and released other slides suggesting that Instagram made more teens feel better about aspects of their lives than feel worse.

The post by Facebook’s researchers was published this to say “see it’s not all bad”, and it does give a more nuanced picture than what the reporting suggested. But it invites a few very obvious questions. One is a question of degree – does it make it slightly better in some cases, and a lot worse in others? The data does not say. More obviously, multiplied by tens of millions, that’s a lot of suicidal ideation and “made it worse”! Should our government – any government – just let this exist with an unregulated shrug?

This really matters, because Instagram is orders of magnitude more popular than Facebook among young people, with US teens more than four times as likely to be a daily user of Instagram than Facebook. The Spinoff’s own reporting revealed a plethora of local social media accounts which encouraged self-diagnosis of neuro-divergence, while Pew research showed a majority of US teens had experienced cyberbullying on social media. While there is a very fundamental sense of “information wants to be free” culture around the internet, the question is how free, and for whom.

None of this is to say that the solution is an outright ban, nor China-style access limits. It’s just that, as of today, we as a society have no way of quantifying how widespread these problems are, nor are there any legislative consequences to bad things happening on the platforms. Which seems quite strange.

Google and digital rent

So far I’ve focused on Facebook and Instagram. The lifeblood of social media – what makes it so wildly profitable – is that its users can instantly publish text, images and video which the social platforms then sell hyper-targeted ads around. This happens largely consequence free, thanks to Section 230, a mid-90s US law which means so-called “neutral carriers” who publish content without mediation online are not liable for that content – though increasingly AI and human moderators make the “without mediation” part questionable. Despite New Zealand never passing an equivalent law, we operate as if it governs us too. It is a candidate for the single most consequential piece of legislation ever enacted.

Google is in that business too – its sister company YouTube is by some measures the biggest TV channel in this country, and it gets almost all of its content free. The Spinoff has an active YouTube channel with more than 25,000 subscribers, publishing content funded by NZ on Air. YouTube has well-documented issues with radicalising its users, and the hoax “Plandemic” video was the origin of many Covid-19 conspiracies. It has banned anti-vax content, but did so only in the past few days, which seems a little redundant at this point in the pandemic. YouTube is nonetheless effective enough to be the starting point for any digital advertising campaign at scale, including our government’s vaccination campaign. 

Still, search-based advertising is a much larger business again. Google is imposingly dominant in search, with around 90% market share, and is the default way a vast number of transactions begin. Having a functional monopoly over something as impactful as search means many businesses have little choice but to engage with it. Few type in a URL anymore, simply searching for the business or service they seek in a browser. Search for any large New Zealand company on Google and you’ll find something odd: an advertisement for themselves. 

A small sampling of the hundreds of NZ businesses renting their own name. (Image: Screengrabs/Archi Banal)

This is because if they don’t buy that ad, a competitor likely will. This and other “must buy” search strings have been called “digital rent”, a functional tax on business simply to do what any Google user assumes would naturally happen when they use its products.

In the above examples, those businesses are paying the rent, and their competitors are staying away. Below is an example of what happens when a business or entity doesn’t pay the rent. 

What happens when you don’t pay the ‘digital rent’ (Image: Screengrabs/Archi Banal)

Crown is not paying the digital rent, and thus Google is selling that space to its competitors. In fact, even with a very specific search string, Crown is the fourth link displayed to consumers, and you have to scroll down to even view the link. People will have different views on whether this is acceptable – but it’s certainly rife with potential for abuse, and a very odd part of our business reality to have no rules associated with it.

Snippets and who pays for content

Along with its dominant role in the online ad market, Google also has an answer to almost any question you might have. It has assumed this role not by generating information, but by sorting and surfacing it. When Google started, two decades ago, you typed your query and then clicked a link to get the answer (hopefully). In so doing, you entered someone else’s online realm – maybe only once, or maybe you enjoyed the experience and returned for years after. Either way, for website owners, the transaction was mutually beneficial, in that after scraping a site, Google gave those sites a visitor that was theirs to profit from (or not) as they wished.

That era feels like a long time ago. Now Google is much more likely to either display the answer as grey text underneath a link, use machine learning to generate related questions with approximate answers or, increasingly, display the answer in bold type atop the search string.

How Google uses third party content. (Image: Screengrabs/Archi Banal)

This is an amazing experience for the consumer. But it means that the site which assembled that information receives no compensation for its creation. In fact, it likely doesn’t even know the search happened, and has had no opportunity to recoup whatever investment was required to gather it in the first place. Often the information is generic – the release date of an upcoming movie, say. But at other times it will be specific and costly to gather – policy intricacies from an interview with a politician, or data from analysis of a listed company which exists nowhere else.

Google treats all this the same way, and while it might only wrap lucrative advertisements around a relatively small proportion of its search strings, its market dominance is built on the breadth of its usefulness. This is precisely the kind of market imbalance which you might assume would require some rules and regulations or Commerce Commission intervention. Instead, it just rolls on unimpeded. At this point, through the endless expansion of their product suites, Google and Facebook profoundly shape people’s idea of what the internet is – underscoring that this is part of a very deliberate strategy, to keep people in their ecosystem, provide wraparound services for almost every imaginable need. To, as far as they can, become the two main doors into the internet.

The high cost of free

Silicon Valley’s tech giants have innovated across thousands of different areas of human life, and we as a species have rightly marvelled at the hardware and software they have created. Their innovation is hardly limited to those products though. They have also innovated in areas which straddle the legal and the financial. These fall into two broad buckets:

  • Firstly, our consumer protection laws are mostly written with the idea that the customer is also the person paying. But because Google and Facebook’s products are mostly free to use and funded by advertisers, they evade the natural reach of regulators that would otherwise supervise their impact. We are the product, bundled up and literally sold to the highest bidder – and our laws have not kept pace with this innovation.
  • Secondly, online transactions are much harder to track. Twenty years ago companies turning over hundreds of millions of dollars each year would need substantial local staff and likely be selling physical products or human-delivered services, so would naturally be taxed. But internet sales are less tangible and visible. Hence Facebook and Google can extract enormous revenues from Aotearoa while paying very little in tax. 

This is before we get to the immense government costs associated with dealing with the “free” products. The enormous growth in communications staff driven by the rise of social media. Expensive campaigns to combat vaccine myths propagated on YouTube and Facebook. Consumer warned recently about social media quizzes that exist to extract personal information hackers can use to access your bank accounts. There is an endless list of costs citizens, businesses and the government have to bear, ultimately tracing back to the wildly profitable, extremely complicated world we have made due to a complete absence of regulation of the search and social spaces.

What we do with other complicated products

Despite their enormous scale and influence, social media and search fall into a regulatory void that stands in stark contrast to many other areas of our economy. We have rules around toys, liquor, cars, advertising, drugs, cigarettes and takeaways. Some are relatively light touch, such as requirements that toys targeting young children don’t have choking hazards. Others are very punitive, like the packaging and sale restrictions on cigarettes, or the outright ban we place on many recreational drugs. These are expanded and complemented by excise taxes on fuel, alcohol and cigarettes which help fund infrastructure or social costs arising from these products. 

Social media and search are glaring absences from that list, in part due to the internet emerging and the companies being founded during an era when deregulation was globally on trend – but any clear-eyed assessment would mark them as screaming for rules, tax and oversight. Each has clear positive effects that many of us experience every day. They also have many problematic areas that impact everything from competition in business to the mental health of teenagers. And yet there is no concerted push toward a regulatory framework here.

In fact, despite the world-conquering scale of search and social media, we in Aotearoa appear a long way from any serious attempt to reckon with the impact of these firms. No businesses in history have ever grown so fast while retaining profit margins as high as 40% (by contrast, the average New Zealand retailer has a profit margin of 3.7%). Which is to say that Facebook and Google can easily afford to fund a regulatory regime and make positive tax contributions to the countries they operate in. 

For now they sit in a new middleground, part utilities (like Chorus, which operates our wholesale broadband), part publishers (like The Spinoff). Utilities are subject to intense oversight by the Commerce Commission. Media is overseen by multiple agencies including the censor, media council and Broadcasting Standards Authority. In between sit social media and search, exploiting the grey area unimagined by our laws, and exporting unfathomable amounts of cash in the process, leaving multiplying social problems in their wake. 

We can move fast, when we understand the issue

On Monday, The Spinoff reported Consumer as calling for regulation of the nascent “buy now, pay later” sector, where firms like Afterpay and Laybuy take advantage of low interest rates to offer free short-term loans. These, unsurprisingly, can have harsh fees for those who miss payments. The sector is a very recent arrival in Aotearoa, but has grown swiftly in popularity, to the point where it’s estimated that consumers have paid over $10m in penalties over the past year. No-fee lending is very new, but MBIE, the government’s business ministry, is already taking a look at it.

This shows that where the issue is clear, we can move quickly. Social and search are deeply complex and challenging, and grow more so by the day. This is brilliant business, but also brilliant strategy. The longer we take to address it, the more Google and Facebook appear too complex and unconquerable, bigger and more powerful than any nation state. Because they’re so transnational – indeed transcending borders is a huge part of their success – both citizens and government can feel like the idea of regulation is akin to trying to hold air in your hand.

This is particularly true for governments, which increasingly rely on them, and fear upsetting them. When MP David Clark, then in opposition raised the spectre of banning Facebook from New Zealand for its paltry tax payments a few years ago he was ridiculed. This is a perennial hazard of dealing with the tech giants, always itching to make politicians look old and out-of-touch (which, to be fair, is often not hard). Likewise, when Facebook briefly shut down news distribution in Australia earlier this year, it felt like a plotline from dystopian fiction, and it was hard to know whether the public would blame the social platform or Scott Morrison. 

That episode is instructive, though. After a few days Facebook blinked and turned the news back on. Google, after sabre-rattling with a threat to withdraw from Australia, caved and made deals with Australian news media after Microsoft adroitly indicated its excitement about taking over its search business in the country. Those deals were, in truth, not the best way forward – a bargain of convenience between the grizzled veterans of the Australian media scene and the tech companies, but they do show that the platforms will get around the table if they are made to. 

Here in Aotearoa, we can do things differently. We need to – with te Tiriti as our founding document, we have obligations here that exist nowhere else in the world. Google and Facebook, indeed all the tech giants, are ultimately extraordinary collectors, sorters and retailers of data. In Aotearoa, the idea that holders of data should act as its kaitiaki rather than thinking solely in commercial terms is gaining ground. Should Google or Facebook change the way they operate according to our specific social context? Given their ubiquity, it feels like a question we must ask ourselves.

Where to from here?

In truth, all of this is hard. “The best minds of my generation are thinking about how to make people click ads,” said a software engineer in a 2017 BusinessWeek story, and the complexity and pace at which this has evolved makes regulation difficult. This is compounded by the fact that some of the next-best minds work in government relations for Facebook and Google, and are adept at making politicians feel scared and small and intellectually incapable of approaching these issues. 

It’s hard, but not impossible. Germany passed a law requiring tech giants to monitor and cooperate with authorities on online hate speech. It worked well enough to prompt some users outside Germany to switch their location to avoid Nazi trolls, though has issues which have prompted an evolution. The EU has fined Google for anti-competitive behaviour, innovated around data privacy and introduced a “right to be forgotten” from search engines. These forays are imperfect, but should embolden us to test legislation and learn too. Even the companies themselves are starting to weary of this chaos, with Facebook itself calling for better regulations. Which, great – but they should on terms set by us, not by the platforms themselves.

Advertising run in Axios newsletters. (Image: Archi Banal)

The great thing is that actually grappling with this has major upside. There is immense tax revenue waiting there, which could compensate the government for the money it spends fixing issues like vaccine information disorder or paying for the production of journalism, which has become far less economic since Google and Facebook became the digital ad economy. Businesses would no doubt appreciate not having to rent their own names from Google. Parents would be thrilled if there were firmer guidance or even penalties around what happens to their tamariki on social media. 

Which is to say there are votes in figuring this out, as well as money.

Facebook and Google need not fear it, either. What it costs in tax revenue (which, morally, they should be paying anyway) they will make back in ever-more unassailable moats which protect their businesses. This will turn them into the safe, slightly boring utilities they aspired to be at birth. The good things which happen on social media – communities built, activists organised, small businesses started and grown – can still happen. Information can still be profitably organised and surfaced by search engines, even if those who created it are compensated for that work. Data gathered can be used in innovative ways even if it is cared for and its uses well-signalled.

We’ve done this before

It took years for us to start really thinking about cars. To get past the romance of them and into the reality. The below graph shows the horrific toll – more than 40,000 dead on our roads over the last 100 years, but tapering down since we really started to take safety seriously in the ’90s. We didn’t throw cars away, we made using them safer. Everything from mandatory seat belts to lowered drink-drive limits to speeding campaigns to better road engineering ultimately combined to save thousands of lives.

By then some of the other scars of an uncritical and unregulated relationship with cars were baked in. Motorways split communities and encouraged poor use of land. Air quality suffered. Public transport plans were abandoned. Now there is a broad political consensus that we need to build up in our cities, and focus on public transport hubs. It’s come a few decades late.

While social and search aren’t leaving bodies by the roadside, they do have a complex mixture of positive and negative impacts on us as a society. It feels like we’re at a turning point. We know the good they do. We also can’t ignore the harm. 

Maybe it’s time we did something about it?

Independent journalism takes time, money and hard work to produce. We rely on donations to fund our work. If you can help, donate to The Spinoff Members

Keep going!