This week CapX is republishing some of our favourite articles of the year. This piece first appeared on October 30.
In a now-famous essay, “What is Seen and What Is Not Seen”, the great economist Frederic Bastiat warned against judging the value of any activity in a vacuum.
Bastiat’s “broken window fallacy” brilliantly exposes a common tendency to focus on the visible, tangible benefits of an action – the “seen” – while neglecting the “unseen” penalties and long-term drawbacks associated with the same activity – the invisible cost of opportunities foregone.
Though writing in 19th century France, Bastiat’s insights have a timeless wisdom. We live with the consequences of reductive “broken window” thinking every day – especially where public money is concerned. Politicians often praise the visible benefits of public spending, e.g. the number of jobs “created”, without considering whether the funds could have been spent more wisely elsewhere – or even how the taxpayer might have spent the cash, had it remained in his or her pocket.
For my money, the fraught Brexit debate badly needs a dose of Bastiat.
So far, discussions of the gains and losses of Brexit have, understandably, tended to focus on the most obvious costs, like the amount Britain may pay in any “Divorce Bill”, the potential “Brexit hit” to companies exporting to the EU, and so on. Of course, these concerns are vitally important, but our focus on the immediate costs of EU departure risks blinding us to the very real costs of maintaining the status quo.
Membership of the European Union carries huge unseen penalties whose implications may not be immediately apparent. The EU’s Common External Tariff, for example, raises prices and so reduces the quantities of goods and services available to ordinary consumers. Since shoppers in the EU lack the counterfactual experience of trading at world prices, this penalty goes unnoticed, but it involves a misallocation of resources on a vast scale.
Negotiating the terms of our departure also comes with huge hidden dangers. In adopting the government’s proposed model for close customs cooperation and a common rulebook, we run the risk of finding ourselves with little scope to diverge from EU regulations on goods, and unable in practice to strike new trade deals with the rest of the world. It is often pointed out that the UK’s interests in trade agreements are primarily in services — but this makes it even more vital to maintain flexibility over what we can concede in goods, to incentivise potential trading partners to strike a deal. The status quo, or anything close to it, carries huge opportunity costs of its own.
Due to a combination of the precautionary principle, enshrined in the Lisbon Treaty, and the difficulties of getting 28 countries to agree on anything, the EU, intentionally or not, often stands in the way of innovation.
In particular, the precautionary principle, the preferred risk management strategy of EU regulators, places the onus on creators of new technologies to prove their invention is safe where some risk may exist — even if there’s no scientific consensus to suggest any actual harm will occur.
The result? It’s often too much bother to innovate.
During the 19th Century, many viewed the emergent railways with a great deal of suspicion. As recorded by cultural anthropologist Genevieve Bell, critics of early locomotives believed “that women’s bodies were not designed to go at 50 miles an hour,” and worried that their “uteruses would fly out of [their] bodies as they were accelerated to that speed”. Had Victorian Britain followed some version of the precautionary principle, it’s hard to imagine a single track of rail being laid, given the levels of contemporary railway fear.
Of course, moral panic over new technology is nothing new. Now, as in the 1850s, over-cautiousness risks hampering important drivers of future growth.
So far, the European Union has made only tentative steps towards regulating artificial intelligence and robotics, though they are currently consulting on the issue. Yet given the EU’s structure, history and current trajectory, the balance of probability suggests AI will be the latest in a long line of missed technological opportunities.
Take genetically modified crops. Since their commercialisation in many parts of the world during the 1990s, GM crops have raised the quantity and quality of the global food supply while lowering fuel and energy usage, requiring fewer pesticides, and reducing both soil erosion and carbon emissions — all with no scientifically-documented evidence of harm to human health. And yet, EU-wide precautionary thinking has meant a de facto ban on GM crops, only one variety of which has ever been approved and grown in Europe.
While farmers outside the EU continue to develop newer, better technologies, hysteria over man-made pesticides has kept European farming methods behind the times. Ironically, foregoing the GM revolution in insect-resistant plant breeding has left European farmers more reliant on pesticides than ever (as has the ECJ’s foolhardy ruling on genome editing earlier this year).
Just last week, the French Finance Minister claimed that EU member states are “very close” to agreeing a counterproductive tax on the turnover of tech companies, a policy likely to discourage new entrants and inflate costs for consumers.
Given all the above, how likely are the EU’s hyper-cautious regulators to pursue a different path when it comes to AI and robotics — or will it be “business as usual”, namely — when in doubt, tax and over-regulate? Certainly, initial signs, including misguided calls for a “robot tax” from the likes of Guy Verhovstadt, don’t inspire confidence.
The EU’s new General Data Protection Regulation (GDPR), implemented earlier this year, will almost certainly hinder the development of Artificial Intelligence by raising costs and limiting access to data. In particular, Article 22 creates a new requirement for humans to review certain algorithmic decisions — a restriction that will significantly raise labour costs, thereby creating a strong disincentive from using AI. After all, the whole point of developing AI is to automate functions that would otherwise be slower, costlier, and more difficult to complete if performed by humans. If you have to get a human to explain the logic, why bother investing in an AI solution in the first place?
These may seem like small concerns in the grand scheme of things, but taken as a whole — and the EU creates a whole lot of regulation — it adds up to an environment often hostile to innovation.
It’s no coincidence that Europe has lagged behind the US for decades when it comes to new inventions, innovations and entrepreneurship. There are of course important cultural differences between these continents, but much relates to the US government’s comparatively light-touch regulatory approach. Not for nothing are there no tech giants in Europe to rival Facebook, Google, Apple or Amazon.
Creating a competitive, innovation-friendly atmosphere is a huge potential hidden “win” of Brexit — with correspondingly huge opportunity costs from failing to do so. Indeed, with more leading universities than the rest of Europe put together, and an already thriving tech sector, Britain has much to lose compared to many of its neighbours.
One can only imagine what Frederic Bastiat would have made of things like robotics, AI and machine learning. But I suspect the spirit of his advice would be the same – consider the unseen, and don’t destroy the jobs of the future in a misguided attempt to protect the jobs of today.