Here’s how my climate research got turned into fake news

A 2009 climate rally in Washington. Image: Getty.

Science is slow. It rests on painstaking research with accumulating evidence.

This makes for an inherently uneasy relationship with the modern media age, especially once issues are politicised. The interaction between politics and media can be toxic for science, and climate change is a prominent example.

Take the recent “deep freeze” along the US east coast. To scientists, it was one more piece of a larger jigsaw of climate change disrupting weather systems and circulation patterns. This includes dramatic changes seen in Arctic sea ice and the knock-on effect on temperatures elsewhere in northern latitudes – both warming and relative cooling. To President Donald Trump the cold snap was a chance to mock climate change, and some sceptics suddenly talked about an impending ice age.

Fiction. Image: Breitbart.

Colleagues and I experienced similar frustrations in late 2017, after we published a paper in the scientific journal Nature Geoscience, in which we concluded that there was more headroom than many had assumed before we breach the goals of the Paris Agreement. We found ourselves not only on the front page of the main British newspapers, but globally, as far-right website Breitbart ran with a story that a small band of buccaneering scientists had finally admitted that the models were all wrong – a fiction rapidly picked up by the more rabid elements in the media.

The essence of good science is to continually update, challenge, improve and refine, using as much evidence as possible. Single events rarely make for good science. And if every painstaking evaluation, updating work from years ago, may be portrayed as demolishing everything that went before – particularly at the whim of non-scientific agendas – then we have a major dilemma. The edifice of science is built with small bricks and this research was no exception.

We emphatically did not show that climate change was “less bad” or “happening slower” than previously thought. Our work built on the many previous scientific studies that had looked at the risks of unchecked emissions and the prospects for limiting warming to 2℃ above pre-industrial levels. The Paris Agreement went further, aiming to “pursue efforts” towards a more ambitious goal of just 1.5℃. Given we’re already at around 1℃ of warming, that’s a relatively short-term goal. Greater ambition therefore requires greater precision.

Our study took a microscope to that question. Where previous estimates were drawn from a range of mostly long-run models that looked at century-long changes, we instead focused on a precise definition and current starting point, and other factors which matter far less in the long term, but a lot if the goal is much closer.

Some of the earlier estimates seemed to imply a “headroom to 1.5℃” of less than a decade of current emissions – clearly unachievable given the long timespans and huge inertia. We estimated about 20 years – equivalent to global CO₂ emissions falling steadily from now until hitting zero in around 40 years – and made it plain that it still looks, to put it mildly, a formidable ambition. Other studies have since come to similar conclusions.

A (non-)story of revolution

The more detailed reporting by those correspondents who attended the scientific briefing was accurate enough (even if some of their headlines and lead-ins weren’t), but that was soon lost in the misrepresentations that followed. Doubtless we could have done more to explain how our conclusions arose from what were actually quite minor scientific developments. Some instead turned it into a story of revolution in climate science. Scientists are also human, and these sceptic reactions reinforced a natural initial inclination among other researchers to defend their previous numbers. Some took to Twitter to do so, but themselves seemed to confuse the media headlines with our actual conclusions.

Some challenges could yet be proved right. There could, for example, be more pent-up warming currently being masked by other pollutants or already lurking in the oceans. When the goal is close, other heat-trapping emissions (like methane) also matter a lot more. Our study – like earlier work – had its share of caveats and uncertainties.

Unfortunately, while good science embraces uncertainty, politics abhors it and the media seems confounded by it. That in turn pressures researchers to simplify their message, and treat existing estimates – often, from a range – like a position to be defended. It is a risky trap for scientists, however eminent and well-intentioned, to wield overnight reactions to parry months of painstaking peer review and refinement that lie behind analyses published in leading journals.

Science against spin

So how should science respond? The climate policy implications are easy: nothing significant has changed. We have but one planet, and both the physical and economic processes that are driving climate change have enormous inertia. If a big ocean liner were steaming into dense fog in polar seas, only a fool would maintain full speed on the basis that the technicians were still discussing the distance to the first big iceberg.

One underlying challenge is indeed around the communication of uncertainty. This is a well-worn track, but it bears repeating. The job of science is not just to narrow uncertainties, but to educate about the risks that flow logically from it. Like a medical prognosis from smoking, the fact that things might turn out better or worse than the average is not a good reason to keep puffing. You won’t know until it is too late whether the damage has been slight, or terminal.

But science also needs to embrace and embed another obvious feature of medical practice: a doctor would never look at just your temperature to diagnose your condition. So part of the problem stems from using a single indicator for complex processes. Too much debate treats temperature (and especially the most recent global average) as the sole indicator, whereas many other factors are at play including sea levels, ocean acidity, ice sheets, ecosystem trends, and many more.

The ConversationThese other trends need to be reported in context, just as economics news reports not only GDP but debt, employment, inflation, productivity and a host of other indicators. And scientists themselves need to improve the art of communication in a world where research can be spun, within hours, into a story of past failure, rather than the reality of continuous improvement.

Michael Grubb, Professor of Energy and Climate Change, UCL.

This article was originally published on The Conversation. Read the original article.


“Stop worrying about hairdressers”: The UK government has misdiagnosed its productivity problem

We’re going as fast as we can, here. Image: Getty.

Gonna level with you here, I have mixed feelings about this one. On the one hand, I’m a huge fan of schadenfreude, so learning that it the government has messed up in a previously unsuspected way gives me this sort of warm glow inside. On the other hand, the way it’s been screwing up is probably making the country poorer, and exacerbating the north south divide. So, mixed reviews really.

Here’s the story. This week the Centre for Cities (CfC) published a major report on Britain’s productivity problem. For the last 200 years, ever since the industrial revolution, this country has got steadily richer. Since the financial crash, though, that seems to have stopped.

The standard narrative on this has it that the problem lies in the ‘long tail’ of unproductive businesses – that is, those that produce less value per hour. Get those guys humming, the thinking goes, and the productivity problem is sorted.

But the CfC’s new report says that this is exactly wrong. The wrong tail: Why Britain’s ‘long tail’ is not the cause of its productivity problems (excellent pun, there) delves into the data on productivity in different types of businesses and different cities, to demonstrate two big points.

The first is that the long tail is the wrong place to look for productivity gains. Many low productivity businesses are low productivity for a reason:

The ability of manufacturing to automate certain processes, or the development of ever more sophisticated computer software in information and communications have greatly increased the output that a worker produces in these industries. But while a fitness instructor may use a smartphone today in place of a ghetto blaster in 1990, he or she can still only instruct one class at a time. And a waiter or waitress can only serve so many tables. Of course, improvements such as the introduction of handheld electronic devices allow orders to be sent to the kitchen more efficiently, will bring benefits, but this improvements won’t radically increase the output of the waiter.

I’d add to that: there is only so fast that people want to eat. There’s a physical limit on the number of diners any restaurant can actually feed.

At any rate, the result of this is that it’s stupid to expect local service businesses to make step changes in productivity. If we actually want to improve productivity we should focus on those which are exporting services to a bigger market.  There are fewer of these, but the potential gains are much bigger. Here’s a chart:

The y-axis reflects number of businesses at different productivities, shown on the x-axis. So bigger numbers on the left are bad; bigger numbers on the right are good. 

The question of which exporting businesses are struggling to expand productivity is what leads to the report’s second insight:

Specifically it is the underperformance of exporting businesses in cities outside of the Greater South East that causes not only divergences across the country in wages and standards of living, but also hampers national productivity. These cities in particular should be of greatest concern to policy makers attempting to improve UK productivity overall.

In other words, it turned out, again, to the north-south divide that did it. I’m shocked. Are you shocked? This is my shocked face.

The best way to demonstrate this shocking insight is with some more graphs. This first one shows the distribution of productivity in local services business in four different types of place: cities in the south east (GSE) in light green, cities in the rest of the country (RoGB) in dark green, non-urban areas in the south east in purple, non-urban areas everywhere else in turquoise.

The four lines are fairly consistent. The light green, representing south eastern cities has a lower peak on the left, meaning slightly fewer low productivity businesses, but is slightly higher on the right, meaning slightly more high productivity businesses. In other words, local services businesses in the south eastern cities are more productive than those elsewhere – but the gap is pretty narrow. 

Now check out the same graph for exporting businesses:

The differences are much more pronounced. Areas outside those south eastern cities have many more lower productivity businesses (the peaks on the left) and significantly fewer high productivity ones (the lower numbers on the right).

In fact, outside the south east, cities are actually less productive than non-urban areas. This is really not what you’d expect to see, and no a good sign for the health of the economy:

The report also uses a few specific examples to illustrate this point. Compare Reading, one of Britain’s richest medium sized cities, with Hull, one of its poorest:

Or, looking to bigger cities, here’s Bristol and Sheffield:

In both cases, the poorer northern cities are clearly lacking in high-value exporting businesses. This is a problem because these don’t just provide well-paying jobs now: they’re also the ones that have the potential to make productivity gains that can lead to even better jobs. The report concludes:

This is a major cause for concern for the national economy – the underperformance of these cities goes a long way to explain both why the rest of Britain lags behind the Greater South East and why it performs poorly on a

European level. To illustrate the impact, if all cities were as productive as those in the Greater South East, the British economy would be 15 per cent more productive and £225bn larger. This is equivalent to Britain being home to four extra city economies the size of Birmingham.

In other words, the lesson here is: stop worrying about the productivity of hairdressers. Start worrying about the productivity of Hull.

You can read the Centre for Cities’ full report here.

Jonn Elledge is the editor of CityMetric. He is on Twitter as @jonnelledge and on Facebook as JonnElledgeWrites

Want more of this stuff? Follow CityMetric on Twitter or Facebook