The rebound effect of 5G

The ‘rebound effect’: in the field of sustainability, this is known as the effect where you leave your energy-efficient heating on for longer, because it’s so efficient… and end up consuming the same amount of energy as you did with your old, inefficient heating system. When discussing 5G, the rebound effect – even more than the fears of unhealthy radiation – is also sufficient reason to temper enthusiasm. Is it wise to plunge head-first into 5G, enthusiastically embracing the next generation without first thinking about how we’ll need to use our internet connections in the future? A short essay by Socrates Schouten.

The rebound effect. Some people are familiar with the concept from relationships: it’s the moment when you and your lover just parted ways and you’re already flirting with a new person, feeling justified by a tangled mix of retaliation and projection. But in environmental science, the rebound effect represents a different pattern, which is more persistent and permanent: efficiency gains that are cancelled out by higher consumption of a commodity – such as leaving your energy-efficient heating on more often. In the roll-out of superfast internet, this rebound effect is not only expected in that environmental context, but also from a governance perspective: the notorious 5G technology.

What is happening in the rebound effect and why can it also occur on a ‘governance’ level? Let’s take a look at where it comes from.

The rebound effect first was first known as Jevon’s Paradox. The economist William Jevons stated in his 1865 book The Coal Question that the invention of a more efficient steam engine would not result in less coal being burned, but more. Once it was more efficient, the use of coal and steam power became economically attractive for many new applications, besides pumping water out of coal mines. This ultimately led to greater demand for coal to power the engines, and a major increase in coal consumption. According to Jevons, it was “a confusion of ideas to suppose that the economical use of fuel is equivalent to diminished consumption. The very contrary is the truth.”

Since then, the rebound effect has been identified in various parts of the economy. People with an LED lightbulb or low-flow shower head are likely to switch on the lights or take a shower for longer – since they’re already being so energy-efficient, right? And as Jevons described it, the same also applies at a group level: if everyone starts driving a hybrid, petrol will become cheaper and demand for petrol may increase – not only in other sectors of the economy, but also because people will start driving more, because it costs less. And car-sharing schemes seems to lead to more car kilometers and more traffic congestion, rather than less.

It is not only the environment that suffers from the rebound mechanism in consumers and in industries. The effect that a highly efficient technology encourages a greatly increased use of that technology plays a role almost everywhere. Certainly the current generation of ‘smart’ technology, such as algorithms and 5G, face these problems. Smart systems are still subject to soaring expectations for performance. It is not uncommon to see a civil servant be tempted into thinking that a technical, automated intervention will lead to efficiency and therefore less work.

Maybe we’re talking about a tax official who has to verify the legality of social benefits, and resorts to using a self-learning algorithm. Suddenly a whole field of new possibilities opens up. After all, when you’re holding a hammer, you’ll start to see nails everywhere. But efficiency gains achieved through smarter governance are often offset by the verifications that have to take place on those systems. And those checks are important, because mistakes are made; systems are fed by incomplete and/or biased information. And that often makes it more confusing for most of the people taking part, too: how did the self-learning algorithm reach this decision? Both the civil servant and the affected citizen are generally left without an answer until the IT company has woken up – or until a parliamentary inquiry has taken place.

What appears to be efficient, mainly leads to high costs and uncontrollable outcomes.

Just as clean technology can actually lead to more pollution on balance, smart technology can in fact lead to more complexity and more ‘hassle’ overall. What appears to be efficient leads to high costs and uncontrollable outcomes – and mainly elsewhere.

And what about 5G? Of course this superfast internet sounds too good to be true. The question is exactly what we need it for. After all, we already have WiFi at home, and you’re not allowed to use a smartphone while cycling, let alone use it to stream movies in super HD. The business case for 5G is substantiated with an image that is stuck at the level of an artist’s impressions of a city with self-driving cars, remote surgery and the buzzing background hum of an Internet of Things.

These self-managing devices may well be fun and convenient. But what are the costs of introducing them? Where will we discover hidden expenses and governance ‘rebounds’?

Part of the costs come from the use of energy and materials. Even just looking at the actual energy performance of 5G already merits a discussion. Yes, lab tests have shown that 5G is five hundred times more efficient than 4G. However, new applications will emerge, and the context in which they are used will also change substantially, so it no longer makes sense to attempt a ceteris paribus calculation of performance in practice. This makes it difficult to determine efficiency.

The reason for introducing 5G is not to do things more efficiently, but to make it possible to transport far more data throughout the system. Rough estimates are assuming somewhere between 100 and 1000 times more data traffic. But whether it will stay there is doubtful: remember Jevon’s coal paradox, and the way our data use has already spiked from Twitter and Tiktok, and we will be talking ‘giga’ rather than ‘mega’.

Besides energy consumption, 5G will also be claiming a large, as yet unquantified supply of raw materials: an antenna every hundred and fifty metres, accompanied by a an Internet of Things abundantly equipped with a plethora of cameras and sensors – the producers of gallium, silicon and lithium are already rubbing their hands in glee.

But that’s not even the biggest issue. The problem is the type of society that we are working towards once 5G is everywhere and data can be directed to and from everywhere. In order to legitimize the investment (of money, but also of a political profile as well as energy and raw materials), it will be necessary to ensure the emergence of the smart city, which analyses and manages city residents as well as cars. In such a data-driven, hyper-connected hall of mirrors, anything seems possible. And for the sake of innovation and returns on investments, it will absolutely have to happen – until eventually the whole thing turns out to be even more complex, more uncontrollable and more of an energy hog.

This is not a new argument, obviously.

The Research and Documentation Centre (WODC) already wrote in 2011:

“The tendency to apply a measure intended as a solution for one problem to another as well – even if it has not been established that the chosen remedy works at all – is very strong in political and policy circles. Three factors can be distinguished that lead to unproven ‘solutions’ becoming the focus of attention and a panacea for a wide range of problems. Firstly, there is the pressure for politicians to govern decisively and attract media attention by implementing concrete measures; secondly, often so much energy has been invested in a certain solution that a point of no return has been reached; and thirdly, other interests, including financial, play a role in choosing a certain policy measure.”

The possible financial interests involved in the roll-out of 5G are outside the scope of this article, but it is clear that an essentially public investment such as 5G involves various public risks. Let us therefore question the optimistic assumptions regarding 5G based on years of experience with new technology. Will 5G solve more problems than it introduces new issues? And how much can we expect to rebound in the smartest city? Or is it actually much smarter to first invest in our current relationship with technology and design a truly efficient city that can truly be governed effectively?