WhatsApp and Encryption

Encryption is in the news; last night you may have noticed WhatsApp inform you that your messages are now “end-to-end encrypted”, and it’s hard not to link this to other events such as repeated attempts by the FBI to get Apple to let them inside its products.

The issue is one of privacy – should your messages to others on your phone be subject to warrants by governments and law enforcement agencies to access them? By adding this level of encryption, WhatsApp cannot see the messages we send, and hence cannot comply with any such warrant. No “wire tap” can be put on your WhatsApp messages, and nobody can “overhear” them, or “eavesdrop” on them any more.

For many people, that’s a great thing – nobody wants others muscling in on private correspondence. Equally, of course, it comes at the same time as another vast tranche of private correspondences regarding tax avoidance and offshore tax havens, were made publicly available in the Panama Papers. Which illustrates the tension: having privacy is great provided we are all nice and law abiding citizens doing nothing wrong. The moment we start doing things that are wrong, then our private correspondences are where we will discuss that wrongdoing.

And of course, on a much more sinister level, it is alleged that terrorists use apps like WhatsApp to communicate in order to avoid the attentions of law enforcement, and hence the desire of the FBI to get hold of the iPhone of a US terrorist. Hence should we afford terrorists more protection by allowing WhatsApp to encrypt like this, or should we, as the UK government has spoken of doing (although not in light of the Panama papers, it might be noted), in the interests of national security?

What could economics shed on to this? One insight economics often affords is that there is an optimal level of everything. We might assert the optimal level of illegal activity is zero, but when there is a private benefit to engaging in illegal activity, and a cost to law enforcement, then there must be a trade off such that the optimal level is not zero. It would take a practically infinite amount of resources to stamp out every kind of illegal activity and as such is impractical. Hence we cannot expect to stamp out terrorism completely, and we must accept that it will always exist. If end-to-end encrypted messaging services exist, they will be used by those engaging in illegal activity. But by and large there will be other ways in which to catch such people in the act of carrying out illegal activities such that impinging on the civil liberties of the masses need not be a necessary cost of making the job of law enforcement much easier.

Equally, a little reasoning from statistical or econometric thinking might help here, too. With any decision in econometrics, there is the risk of a Type I or Type II Error. These are false positives, and false negatives, respectively: incorrectly rejecting something that is true, and incorrectly not rejecting something that is false. With easier ability to eavesdrop on people, will law enforcement agencies use this to pursue the wrong people, people who are simply going about their day-to-day activities without engaging in any kind of terrorist activities? Of course greater surveillance means that catching terrorists will be easier, but will it mean that innocent people are caught up in the machinery set up to catch terrorists?

Terrorism and Economics

What is your reaction when you see the continued news stories surrounding the Paris terrorist attacks? Brussels today, for the third day in a row, is on lock down – army everywhere, most things shut down, people being told not to congregate in groups.

Your response, hopefully, is not necessarily with the economic impact of all of this – there’s a human impact first of all, and hopefully the very strong measures being taken in Belgium will have had the impact of saving some innocent lives.

In light of the Paris attacks, much commentary has been written; not least: why do we care more about Paris than Beirut? Here’s a piece written last week on the potential economic impact: the stock market effect was muted, but it may be that in the longer term we notice effects more in terms of economic efficiency in Europe – much production relies on cross-border transport links, which are being disrupted this year firstly via the refugee crisis, and now from the increased level of terror alerts across the continent.

Places like Egypt and Tunisia have undoubtedly suffered differently in the longer term after terrorist atrocities, via lost tourism income. It’s clearly a secondary impact compared to the initial impact of any attack on human life, but an impact nonetheless. Are we generally sufficiently risk averse that terrorist attacks will influence our consumption patterns, even if the likelihood of being caught up in a terrorist attack is incredibly small?