As we talk about new regulation at the national or regional level, one of the concerns is that this might lead to the break up of the global internet.
This is sometimes referred to as a ‘Balkanization’ of the internet, but there has been some criticism of associating a regional name with a negative process (the Wikipedia article discussion provides a flavour of this debate).
Whatever the politics of ‘Balkanization’ as a term, it is anyway not a great analogy when we are discussing why national governments in 2021 might want to diverge from the global internet.
I would instead suggest that we think about what is happening as a ‘BREXITISATION’ of the global internet as, having watched this process play out over recent years, it better captures the dynamics in play.
[NB This post is certainly not at attempt to re-hash the substance of the Brexit debate as that is not useful right now but to learn something from the dynamics that shaped it.]
The key argument in favour of Brexit was that it would allow UK institutions to ‘take back control’ from supra-national bodies.
Core arguments in favour of staying in the EU were that it made trade simpler for businesses and travel easier for individuals.
Both supporters and opponents of Brexit recognised that leaving the EU would increase bureaucracy and reduce choice for some kinds of trade, but they differed (and still do) on how high that price would be and whether it would be worth it in the long run.
And both sides recognised that staying in the EU did compromise national sovereignty as important decisions were being made on a shared basis, but they diverged on how meaningful and useful the enhanced legislative freedoms would be that Brexit could deliver.
Lots of other lines of attack and defence were thrown into the mix but at the heart of the debate the British people were being offered a choice between sovereignty and convenience, and they chose the former by a slim majority.
The Internet Union
A driving force behind the development of internet protocols was a desire to overcome borders between computer systems and create a digital space that could be traversed with the minimum possible friction.
The developers of internet protocols wanted to enable the free movement of packets of data between systems that had previously been separate domains, either entirely inaccessible to each other, or connected via gateways that were difficult and expensive to use.
They have been wildly successful to the point where it has become very cheap and easy to introduce a new device into this ‘Internet Union’ and exchange data with any other device using the same protocols.
At least when it comes to the transfer of bits and bytes, the default position is one of frictionless, borderless, low bureaucracy, low cost global trade (with some notable exceptions we shall discuss shortly).
The rules of this Union are written and maintained by various supra-national committees that largely sit outside of any form of government control.
If you follow these rules, then you can expect to gain access to people in most countries around the world simply by turning on a service and advertising it as being accessible.
The default setting is to include everyone with a connection, wherever they may be physically, with extra steps having to be taken if you want to exclude people in certain countries.
These exclusionary steps can be taken by the owner of a server, which can be configured to deny access to clients based on where they appear to be located, or by the entities that provide user connections, who can configure their networks to block access to particular servers.
Some operators of servers do try to restrict access, especially where they are offering physical goods that can only be shipped to defined locations, but they generally prefer to have as wide a reach as possible.
And some governments do order local connectivity providers to block certain types of server, with China being the standout example, but the norm remains that the vast majority of public servers will be accessible when someone connects to the internet in most countries.
The protocols that enable this global union are on government agendas from time to time, with tussles over who gets to set technical standards and administer parts of the underlying architecture, but it is in the services that run over the internet where governments feel the greatest loss of control.
Without having taken conscious steps to enable this or negotiate the terms of trade, governments find that their citizens can access a wide variety of services that may be entirely outside of their jurisdiction.
The range of issues that may be in play when people use these internet services is huge including ones that are highly sensitive for any government, such as tax revenue and national security requirements.
The more ‘stuff’ that people in a country do on the internet, the more acutely will any loss of control be felt by its government.
If important internet services are largely domestically provided then this will be much less of a problem, but in many countries citizens will be highly dependent on foreign services.
Nation State vs The Internet Union.
As I was starting to think about the how the dynamics of this ‘Brexity’ choice might be applicable to internet regulation, an article popped up in my feed from a UK Conservative MP that was effectively an articulation of just this position.
It has been a regular lament of EU policy makers that there are not European equivalents to the large US-based tech companies that are so well-used by people living in the EU.
The UK has generally been more relaxed about dependency on the US, with whom we believe we have a ‘special relationship’, and it is interesting to see a UK legislator from a traditionally US-friendly political faction now making the case for breaking away.
The championing of regulatory sovereignty as worth it even if this means the loss of important services strongly echoes the arguments made in the Brexit debate, and it is assumed that new domestic businesses will emerge to replace the ones that may find it impossible to comply with new national regulation.
Those sceptical of the EU see it as making their country subservient to rules set by others, and there has often been a particular focus on Germany as it is perceived to be the most powerful rule-setter within the Union.
There is a similar scepticism that the Internet Union means accepting rules set in the US, the most powerful country in internet politics and host to many of the biggest global services.
There is a logic for British politicians who have chosen to follow a path that they see as allowing the UK to ‘break free of German dominance of the EU’ to want to take the next step and ‘break free of US dominance of the internet’.
What Happens Now?
We can imagine three scenarios playing out over the next few years :-
A – Business as Usual
There may be a continuation of something close to the status quo where the Internet Union stays as the dominant model and most countries remain open to most services.
This can happen even where some countries pass their own local regulation, if there is some consistency across the new requirements imposed on services and/or if these are not enforced too aggressively.
A key concept to consider is that of ‘regulatory interoperability’, ie how easy it would be for a transnational service to comply with each different national regime, and how far are conflicts of law are avoided.
If we look at privacy regulation between the US and EU as an example, we see there was an intent to make this interoperable by creating a system called the ‘Privacy Shield’ that established standards for US companies to apply to personal data of people from the EU.
This system was struck down by EU courts on the basis that some aspects of US law are incompatible with parts of EU law, leaving us in a state of uncertainty until some new interoperability mechanism is agreed.
There are again echoes in the Brexit situation, where the future relationship between the UK and the EU will depend in large part on the extent to which each side’s regulation remains consistent, even if differently worded, or diverges in significant respects.
As well as the letter of the law, enforcement practices can be a key factor in determining whether national rules cause a break-up with the Internet Union.
A country may pass quite divergent regulation that would be hard for foreign services to comply with but this may not create serious problems if it is not enforced aggressively.
For example, we see a lot of legislation being agreed by the Russian Duma that could create very serious compliance challenges for some platforms but there is a process of negotiation and phased introduction that has, at least to date, meant that major services have not been blocked from the market.
B – Have Your Cake and Eat It
A common refrain in the Brexit debate was that the EU would not allow the UK to ‘have its cake and eat it’, in other words to enjoy the benefits of the EU single market while being outside of the political entity.
[NB I was going to use the original version of this phrase which was ‘to eat your cake and have it’ rather than the one that is now in more common usage but I didn’t want to be accused of being the Unabomber.]
With the Internet Union, there is no such constraint and countries have been able to block access to their home markets while enjoying access to global services operating elsewhere.
Regulation in China has made it all but impossible for many global services to operate there without severely limiting and/or re-engineering the product they offer in other countries.
At the same time, Chinese businesses make extensive use of global platforms to market their goods to the wider world, and Chinese internet services have not been blocked from offering their services in other countries.
There may be a period where more countries follow this model of making their local environment inhospitable to foreign services while continuing to engage in the Internet Union as it suits them.
At some point this would likely create a strong enough sense of unfairness as to be unsustainable, and the Executive Order issued by the former US President in relation to TikTok and other Chinese services can be seen as an expression, albeit a clumsy one, of that frustration.
C – The Big Break-Up
Both of the previous scenarios depend on there being a critical mass of countries still preferring the original Internet Union model and keeping their national regulation ambitions in check so it can continue.
But if countries are really committed to sovereignty over convenience then preserving the old model may no longer be their goal, and its destruction may even be seen as desirable, echoing the sentiments of the UK MP I cited.
A likely candidate for driving such a break-up is the European Union, both because it has some differences of opinion with what it sees as a US-dominated prevailing model, and because it is large enough to feel confident about going it alone.
There is an expectation that there will be a comprehensive dialogue between the Biden administration and the EU on a broad range to tech policy questions with both sides saying this is a priority.
This process may lead to agreements that will preserve the Internet Union as these two large blocs sign up to a common approach and to keeping their regulatory regimes interoperable so businesses can easily serve customers on both sides of the Atlantic.
Or it could lead to the blocs ‘agreeing to disagree’ and heading down a path of significant regulatory divergence where businesses really need to have separate US and EU entities to comply with different and conflicting regimes for data protection, taxation, content moderation etc.
The evolving regimes of other countries, especially Brazil, Russia, India and Turkey, are also a critical factor and they will have their own bilateral discussions with the US administration.
We may see some new groupings emerge but for the next few years at least the key conversations will be these ones between the US on the one hand and the countries where US internet services have a significant footprint on the other.
Fight or Write
One of the reasons for rejecting Balkanisation as a description of the evolving internet regulation debate is that it carries a connotation that failure in negotiations leads to bloody conflicts.
Opponents of Brexit argued at times that leaving the EU could lead to renewed conflict in Europe and were criticised for overplaying their hand.
While it is still early, our experience of Brexit to date is that it is causing struggles that are fought with pens not swords as people have to deal with the increased bureaucracy that inevitably comes with being a ‘third country’.
So with a regulation-induced ‘break-up’ of the internet, we should resist the temptation to catastrophize this – it will create more paperwork (and cost) and reduce choice but not be the end of the world.
Some businesses may find it hard to survive in a more fragmented regulatory environment, but most will adapt by creating local entities (either physically or virtually) in the markets they care about.
Users of internet services may have less choice, especially if they live in a country that has a small population and high regulatory compliance costs, as services weigh up the cost of having a presence in particular markets instead of blithely opening up to everyone.
There may be more space for local businesses to develop where compliance costs (of all kinds not just cash) for foreign businesses are so high as to be prohibitive, as has been the case with China for some time.
The businesses that do thrive in this new world may find that they get increased protection from competition in regulated national markets as their users are less frequently courted by foreign competitors.
It will certainly be a different internet and we should be concerned about what may be lost along the way with the downsides likely to be felt most strongly in countries where governments are least trusted to act in the interests of their people.
I chose as the strapline for this blog ‘how to regulate the internet without breaking it’ thinking that most people would share the view that this is a good outcome, but it is perhaps time to reassess that assumption.
It may be that there is sufficient resentment towards the (US-dominated) Internet Union amongst important political groups around the world that its breakup is the desired outcome rather than an accidental by-product of regulation.
If that is the case then we should spend more time thinking about ‘how to cause as little damage as possible while we break up the internet’ – the lesson from Brexit is that these kinds of negotiations are inevitably tense and and that the details (including on things not obviously important) really matter.