I
You’ve heard of the climate crisis and the financial crisis. Internationally there’s the Ukraine crisis, the Syrian crisis, the Nigerien crisis. And in Britain there’s the NHS crisis and the relentlessly-mentioned cost of living crisis.
These are just the tip of the crisisberg.
Last week the Financial Times - the soberest of the British newspapers (which is not saying a lot) - ran stories on the Libyan government crisis, Europe’s water crisis, America’s drugs crisis, the Chinese property sector’s liquidity crisis (also in Vietnam), the energy crisis (which gets its own section, as does the pensions crisis), and the West’s crisis of courage (?).
Moving down the sobriety ladder, and ignoring the crises already mentioned, the Guardian ran stories on the crisis in public confidence in the police, the rural mental health crisis, the London air pollution public health crisis, and don’t forget their series titled ‘UK in crisis’.
Finally, at the drunken-yelling end of the spectrum, the Daily Mail covered Pizza Hut’s debt crisis, America’s antivaxx crisis, the celebrity-induced fertility crisis, the Los Angeles homelessness crisis, the Queensland premier’s crisis of confidence, Australia’s housing crisis, Manchester United’s left back crisis, and the Armenian humanitarian crisis (notable because protestors disrupted Kim Kardashian’s day).
Has the media always been this fond of calling things crises? It feels like a relatively new phenomenon, kicking up a gear over the last few years.
Unfortunately, at the time of writing, Google’s text corpora only goes up to 2019. But it charts a long-term escalation in crisis-mentions going back around a century, kicking up a gear in the 1960s, and then flatlining from 2013 to 2019. I predict that this will shoot up from 2020 onwards, but we shall see.
What explains the big 1960 upswing? It starts a little early, but the obvious candidate is 1962’s Cuban Missile Crisis. (‘Missile crisis’ takes off in 1958.)
The Cuban Missile crisis is a case in which the ‘crisis’ label is clearly warranted: if global thermonuclear war doesn't make the cut, then your standards for calling something a crisis are too high.
But today the standard is too low.
II
What’s wrong with setting the bar low? Why not talk of a climate crisis, a cost of living crisis, or Manchester United’s left back crisis?
Firstly, and most importantly, inflating too many issues to ‘crises’ hinders the identification of ‘real’ crises. If everything is a crisis, then nothing is.
I’m not claiming that people fed a media diet of cost-of-living crises are completely unable to grasp that nuclear war is a bigger deal than high inflation. People’s comprehension of a given issue doesn’t stop at the labels used to describe them. People, and language, are adaptable. If it becomes common knowledge that ‘crisis’ doesn’t mean what it used to, then other words will emerge to signal the distinction.1
Rather, I claim that constantly inflating everything to a crisis degrades our ability to discern between different degrees of severity. The media constantly crying wolf distorts people’s worldview on a range of serious issues. If we’re talking about even slightly more difficult comparisons than nuclear war versus inflation, then it would be useful to know whether Bad Thing has increased by 10% or 10,000%, at least in part so that we know how it stacks up against Bad Thing 2.
In other words, crisisism hinders prioritisation among various problems.
Secondly, and less importantly, calling everything a ‘crisis’ is just linguistically lazy. The English language has a bunch of specific words to denote specific bad things! Why call it the ‘Ukraine crisis’ when ‘war’ would be just as dramatic, and far more specific? Why call it the ‘cost of living crisis’ when you’re actually talking about ‘high inflation’? Why ‘Manchester United’s left back crisis’ and not ‘Reds’ roster riddle’ ?
I hope that the media can sensationalise stories well enough to get their clicks without having to hammer everything into a ‘crisis’, specifically. But I lack faith.
III
I’ve claimed that over-using the word ‘crisis’ to refer to negative events is a bad thing mainly because it degrades our ability to distinguish ‘real’ crises from fake crises.
But how do we demarcate the two? What counts as a ‘real’ crisis?
You might think that the Pizza Hut debt crisis is unworthy of the label, but that the climate and cost of living crises should count. You may even be suspicious that my objection to calling things crises, and my invocation of silly examples like Pizza Hut, are actually covert means by which to downplay the importance of the climate and cost-of-living crises.
If so, I have some bad news. I think that the Pizza Hut debt crisis might be more worthy of the ‘crisis’ label than the climate and cost-of-living crises.
Let me explain.
‘Crisis’ is a Greek word meaning a ‘decisive point in the progress of a disease’: an inflection point from whence the patient either recovers or dies. (Itself derived from the Proto-Indo-European ‘krei’ for ‘sieve’ or ‘discriminate’.) This etymology gives us a useful way to discriminate real from fake crises. A real crisis is (1) temporally restricted and (2) sufficiently important within a specific domain. A fake crisis lacks at least one of these criteria.
Climate crisis
The problem with ‘the climate crisis’ is that it doesn’t refer to a temporally restricted inflection point. Greenhouse gas emissions have been rising for many decades (at least), and are set to rise for a little while longer still. It’s possible (if unlikely) that certain levels of warming could trigger feedback loops leading to ‘runaway’ climate change. But, quite aside from such tipping points being highly uncertain, it would be the tipping point which would constitute the crisis. Not the incremental increases in GHG emissions decades previously. Even if those emissions predictably and metronomically caused the subsequent crisis, I maintain that they do not constitute the crisis. They remain a precursor.
It is possible (but entirely unproven) that rhetorically bringing forward a moment of crisis from the future to the present could spur necessary action today. If that were true, then I would grudgingly acknowledge the utility of the climate crisis terminology. But if you think that this is intuitively plausible, and therefore sufficient to justify the terminology, then remember that we can tell equally plausible-sounding stories with precisely the opposite valence. Perhaps talking of ‘a climate crisis’ spurs not action but despondency.2 Ultimately, if you care about climate change, then you should care to find out the actual consequences of the language you use.3 Until then, though, I say that ‘crisis’ is the wrong label.
And, to be absolutely clear, rejecting the crisis label is perfectly consistent with caring about climate change.
Cost of living crisis
The ‘cost of living crisis’ means high inflation, a term which is notably more concise and precise.
This one also lacks temporal specificity: prices rose, at varying rates of acceleration, over a period of months and years. Indeed, high inflation is generally a serious problem only if it drags out over a long time period. (Or hits extreme rates.)
But the main reason why we shouldn’t call this a crisis is because it’s simply not serious enough. Take a look at consumer price inflation over a few big countries since 1960:
This graph ends just as the ‘cost of living crisis’ was about to begin. In the UK the highest measure of (annualised) consumer price inflation hit a (brief) peak of around 11%, and today (September 2023) has fallen to around 6%. That is high relative to the last 30 years, but unremarkable relative to the 1970s-1990s. And it is positively trivial relative to bona fide hyperinflation (Did you know that Brazil hit ~2500% not too long ago?)
Remember that my second criterion of a real crisis is not just importance simpliciter, but importance within a specific domain. High inflation can lead to small-scale crises. People lose their homes, businesses and livelihoods collapse, some people even die. These are all real crises for those affected: temporally-restricted inflection points of high personal importance. But such personal crises happen every day. They’re part of the backdrop of life. So while high inflation increases the number of personal crises, that doesn’t ipso facto entail a national crisis, which is what is alleged by the ‘cost of living crisis’ moniker. Something more out of the ordinary should be required for that.
Pizza Hut debt crisis
This is why I claim that the Pizza Hut debt crisis may plausibly be more ‘real’ qua crisis than the climate or cost of living crises.
Unlike the climate crisis, Pizza Hut’s debt crisis may comprise a temporally limited inflection point: either they honour their corporate bonds by the specified date or they do not (with whatever consequences follow).
And while their debt issues are far less important than high inflation from an impartial view, it may be far more important relative to the domain of Pizza Hut. It could be existential to the company.
And that is why their debt crisis is more worthy of the label than the other two.
(I haven’t actually researched Pizza Hut’s debt position. For all I know, the Daily Mail could have invented the whole thing - the example is illustrative only!)
IV
I don’t think this blog will prevent the media from calling everything a crisis. But perhaps I’ve convinced you that their doing so is a bad thing. If you’d rather read things with a greater sense of perspective, then what alternatives are there?
First: find better media. This is tricky. Even if you dodge the stuff which is Bad on Purpose to Make You Click, the media cycle is not well-suited to sober perspective-offering, with even highbrow outlets like the FT succumbing to crisisism.
Nor is it enough to cut out the middle man, and take prestigious scientific-sounding bodies at their word. For example, you’d think that the Bulletin of Atomic Scientists would know what they’re talking about, and yet they hawk the Doomsday Clock, a completely nonsensical crisisist media machine.
There’s no general-purpose solution here. You just have to find high quality sources who actually care about perspective. Blogs can be your friend here (see eg my earlier piece on whether Cambridge will suffer a water crisis), but they’re no panacea.
(I like a lot of stuff in the Effective Altruism orbit because EAs tend to be numerate and care about scale. If an EA discusses a crisis, it’s likely to be a top-tier crisis: existential risks or non-existential but globally catastrophic risk, like an extremely severe pandemic. (And the difference between these is appreciated and noted!))
Second, you could replace (at least some) traditional-format media with better kinds of information source. For a long time Max Roser has been banging the drum about a lot of the things I’ve been complaining about here. The world has made enormous progress across a range of metrics like poverty, child mortality, and democracy, but most people are ignorant of this progress. (Which is not inconsistent with the claim that many things remain terrible.) But merely avoiding media misinformation is not a viable solution to being well-informed: personal experience is far too narrow. Instead, you need to seek out sources which actually attempt to provide some sense of perspective rather than the usual sensationalist tricks. And that’s why Max founded Our World in Data, the source of that graph above. Today it is the best one-stop-shop to get perspective on most of the world’s biggest issues.
Third is prediction markets, a solution beloved in my corner of the internet. (See link for details.) If you hear about a new Russo-Finnish crisis (don’t worry, I just made it up), and you are understandably concerned that it could trigger a full blown NATO/Russia war, then you can operationalise this into a question like ‘Will there be a NATO/Russia war before the end of 2023?’. People can offer their predictions, which convert to an implied probability of the event happening. (See, eg, these markets on nuclear weapon use before 2024.) And to figure out the magnitude of harm, not merely its probability, you could ask something like ‘Conditional on a NATO/Russia war starting before 2024, will there be more than 1m casualties before 2025?’ (mutatis mutandis for 10k, 100k, 10m, etc). (Or, if the platform allows continuous predictions, just ‘…how many casualties will there be?’.)4 And if you like the sound of this, then you can combine prediction markets with a newspaper-y landing page via the new and rather nifty Base Rate Times, which aggregates forecasts across several different prediction markets.
None are complete solutions to avoiding crisisism, but together they go a long way.
We usually see this going the other way: a stigmatic label is replaced by a friendlier euphemism → that euphemism quickly gains the stigma of the old label → we create a new and still-friendlier euphemism → it too becomes stigmatic → repeat ad infinitum. The classic example is terminology related to intellectual disability. Both ‘moron’ (a neologism) and ‘retarded’ (literally: ‘slower’) were terms introduced to avoid older stigmatic labels like ‘idiot’ or ‘imbecile’. But today those terms are considered slurs. And tomorrow the verboten term might well be ‘intellectual disability’, notwithstanding its governmental imprimatur, assuming that people continue to use these terms to be mean.)) Steven Pinker refers to this as the ‘euphemism treadmill’. Some groups have reclaimed previously-stigmatised terms (eg ‘queer’), but its a very dangerous game to be too early to the party, where it may be unclear to others whether you’re using the term because you’re an awful bigot or an enlightened vanguard. (Cf Scott Alexander on respectability cascades.) All-consuming irony is one pretty sad way to handle this.
Wikipedia mentions a couple of low-quality studies proposing that this is the case, but I wouldn’t consider this evidence either way.
David Deming: ‘Commitment to empiricism is more than just good scientific practice: it’s a moral imperative. Social programs… have extremely high stakes. There is only so much money available to help people in need. As a researcher I might advocate for programs because they comport with my prior beliefs about what works or because doing so enhances my scholarly reputation. But if the program is ineffective, I’m actively harming people who need help by misdirecting resources.’ Likewise for linguistic prescriptions, insofar as (a big caveat!) they affect action.
Metaculus has a conditional pair forecast feature.