Last week, the news from the Actuary magazine was that climate change could slash global GDP by 18%. This was based on a Swiss Re report, the economics of climate change, from which the analysis above is taken.

According to the report, “The current trajectory of temperature increases, assuming action with respect to climate change mitigation pledges, points to global warming of 2.0–2.6°C by mid-century.” It was unclear why they had decided to stop at 2050, when current commitments continue to push temperatures up until 2100. And the scenarios from the IPCC’s AR5 Synthesis Report (see below) show that the path we are currently on diverges far more considerably from the Paris agreements after 2050. Climate effects are very long-term and many of the impacts of a 2-3°C warming would be irreversible ones, ensuring continuing losses at similar or greater levels for decades to come, and that is before we even consider the much higher probabilities of feedback effects: from the melting of the permafrost, additional methane releases, loss of Amazonian carbon and the loss of the albedo reflectivity of Arctic ice. The Swiss Re report makes clear that is has not considered these.

You might notice that there is a separate column to the left, in a different colour, with the title “Well-below 2°C increases” and sub-title of “Paris target”. It is actually an agreement which 189 countries have signed up to, including the UK. As the Paris Agreement says (Article 2 Point 1):

This Agreement, in enhancing the implementation of the Convention, including its objective, aims to strengthen the global response to the threat of climate change, in the context of sustainable development and efforts to eradicate poverty, including by:
(a) Holding the increase in the global average temperature to well below 2°C above pre-industrial levels and pursuing efforts to limit the temperature
increase to 1.5°C above pre-industrial levels, recognizing that this would significantly reduce the risks and impacts of climate change;

There has been some debate over whether the Agreement is aiming for 1.5°C warming with a 50% chance of staying below it, or for “well below” 1.5°C similar to the 2°C goal with a 66% chance of avoiding more than 1.5°C warming, but the modelling used for the next IPCC report has adopted the latter definition. Either way, I cannot see why Swiss Re has decided to put the Paris Agreement targets in a different column from what it calls the “likely range of temperature gains” as if those we have committed to are no longer feasible to aim at.

In saying this, I do not underestimate the massive challenge of keeping to the Paris target. As Mark Lynas says in Our Final Warning, at the end of 2018 over 1,000 GW of additional fossil-fuelled electrical power generation capacity was planned, permitted or already under construction around the world, equivalent to adding an additional 188 Gt CO2 into the atmosphere to the 658 Gt already baked in from existing infrastructure, which gives a total of 846 Gt of CO2 not including impacts from deforestation, agriculture and future land-use change. This compares to a future carbon budget as estimated at the end of 2018 by the IPCC (although estimates of this vary considerably) of 420 Gt of CO2 (or 1,170  Gt of CO2 for 2°C warming). So an extraordinary change of direction is required and we should be very cautious of getting anywhere near these limits when we do not know precisely where they are.

Which brings me onto the modelling of economic impacts. The first thing to say is that modelling in terms of impact on GDP, while guaranteed to get the attention of the financial community, is perhaps not the best way of communicating the devastation of runaway climate change.

In the summary of Mark Lynas’ excellent book Six Degrees: Our Future on A Hotter Planet , which summarised the scientific consensus already arrived at by 2007, the three degree increase for which damages are being estimated is expected to lead to Africa […] split between the north which will see a recovery of rainfall and the south which becomes drier […] beyond human adaptation. Indian monsoon rains will fail. The Himalayan glaciers providing the waters of the Indus, Ganges and Brahmaputra, the Mekong, Yangtze and Yellow rivers [will decrease] by up to 90%. The [IPCC] in its 2007 report concluded that all major planetary granaries will require adaptive measures at 2.5° temperature rise regardless of precipitation rates.[and] food prices [will] soar. Population transfers will be bigger than anything ever seen in the history of mankind. [The feedback effects from the] Amazon rain forests dry[ing] out and wild fires develop[ing] [will lead] to those fires [releasing] more CO2, global warming [intensifying] as a result, vegetation and soil begin[ning] to release CO2 rather than absorb[ing] it, all of which could push the 3° scenario to a 4°-5.5° [one]. The recent update to this: Our Final Warning, describes “entering the three-degree world means we are now living in a hotter climate than any experienced on Earth throughout the entire history of our species”. These impacts, which are likely to pose existential risks for many, appear totally inconsistent with the economic loss modelling shown above.

In his 2020 paper, The appallingly bad neoclassical economics of climate change (apologies, Journal access required), Steve Keen says in the abstract:

Forecasts by economists of the economic damage from climate change have been notably sanguine, compared to warnings by scientists about damage to the biosphere. This is because economists made their own predictions of damages, using three spurious methods: assuming that about 90% of GDP
will be unaffected by climate change, because it happens indoors; using the relationship between temperature and GDP today as a proxy for the impact
of global warming over time; and using surveys that diluted extreme warnings from scientists with optimistic expectations from economists. Nordhaus has misrepresented the scientific literature to justify using a smooth function to describe the damage to GDP from climate change. Correcting for these errors makes it feasible that the economic damages from climate change are at least an order of magnitude worse than forecast by economists, and may be so great as to threaten the survival of human civilization.

There follows a demolition of the methodologies employed by Nordhaus and others in this field. To be fair to the Swiss Re report, some of the criticisms in Keen’s paper appear to have been borne in mind when constructing their model, eg:

A shortcoming of our model build so far is that some economic impacts are linearly estimated: non-linearities are not adequately captured. We use multiplicative factors of 5 and 10 to simulate the increasing severity of outcomes from nonlinearities… Importantly, the framework does not consider
tipping points, events such as the partial disintegration of ice sheets, biosphere collapses or permafrost loss, that pose a threat of abrupt and irreversible climate change. This is because it is thought that tipping points will materialise well after our model horizon of mid-century only.

And as the Swiss Re report also acknowledges:

It is likely that the estimated impacts of GDP damages from climate change will rise as existing modelling develops to incorporate economic linkages in trade, migration and other channels, and to generalise the results to multiple countries.

And they are getting criticisms from the usual suspects of climate denial, eg Bjorn Lomberg on Twitter here, that even their attempts to date to quantify the uncertainties caused by non-linearity are a step too far.

And yet there remains a problem with these analyses in that they fail to capture existential risk. One of the things Steve Keen points out in his paper is the different attitude Nordhaus found towards estimating damages from climate change in natural scientists as opposed to economists. Natural scientists typically estimated the damage at 20-30 times higher than economists and some refused to cooperate with the exercise at all:

I must tell you that I marvel that economists are willing to make quantitative estimates of economic consequences of climate change where the only measures available are estimates of global surface average increases in temperature. As [one] who has spent his career worrying about the vagaries of the dynamics of the atmosphere, I marvel that they can translate a single global number, an extremely poor surrogate for a description of the climatic conditions, into quantitative estimates of impacts of global economic conditions. 

But how do you calibrate what is clearly a complicated model that Swiss Re and Moody’s have constructed for this analysis? Obviously we all have a very recent GDP fall in our minds at the moment – here is a summary from the UK Commons Library of Economic Indicators as at 30 April 2021 (themselves sourced from OECDstat and Eurostat):

This shows an almost identical GDP fall of 10.5% year on year in Q2 2020 for the OECD as predicted in the event of a 3.2°C warming, although it has bounced back pretty quickly since. For a longer term view of the global data, Our World In Data have an Annual growth in GDP per capita graph which runs from 1961 to 2017 (see below).

One very large GDP fall which stands out in the data here is the 26.5% fall in China in 1961. This was towards the end of the China’s Great Famine, in which approximately 3 million people died of starvation over a 3 year period. This certainly qualifies as an existential event and Swiss Re’s modelling suggest something of similar proportions in Asia and Africa at 3.2°C warming.

The biggest danger in all of this is that rich countries will look at a 10.6% reduction in GDP (at 3.2°C warming) and think this liveable with and adaptable to for their populations. After all, Simon Wren Lewis calculates that the austerity policies between 2010 and 2018 in the UK reduced GDP by nearly half of this amount every year for at least the second half of this period, compared to where it would have been without these policies, with an estimated cumulative loss of 15.9% of GDP. An 18.1% overall world average loss, however, effectively means more than a 25% loss for the rest of the world outside the OECD, as the OECD accounts for around half of the world’s total GDP which, even if we did not allow for the acknowledged likelihood that these are underestimates, is still in the Chinese Famine category of disaster and neither liveable with nor adaptable to.

We are already seeing vaccine nationalism carve up the world between rich and poor countries, with up until last month only 0.3% of the vaccines administered around the world having gone to people in low-income countries. This is likely to reduce the ability of poorer countries to be represented properly at this year’s COP26 when it frames a global response to the climate change which will affect them so disproportionately. And the losses if we do not act will be measured in far more frequent floods and sea level rise, extreme storms and heatwaves, crop failures, water and food shortages and mass migration on a scale we have never seen before, not GDP.

Could climate change slash global GDP by 18%? It’s much worse than that.

 

There is a particular variety of We Know Zero graphs that look like this one – showing an experience of a steady increase in something (usually bad, but not always) up until now, followed by a projection of that thing falling in the future. My wife Marsha suggested I call them Hope-over-Experience graphs, which seems to suit them very well.

Such diagrams are often very comforting for those who want to maintain the status quo. Let’s look at three such curves in particular (the excellent Doughnut Economics by Kate Raworth has alerted me to the first two of these).

The Kuznets Curve

There is a considerable body of evidence, most notably from Kate Pickett and Richard Wilkinson, that inequality impacts most health and social problems adversely, to the detriment of all socio-economic groups, but what is to be done about it? Enter our first Hope-over-Experience graph. In this case the x-axis is actually income per capita, but to the extent that this is something expected to increase with time I don’t think this matters too much. The y-axis is inequality. It was originally proposed by Simon Kuznets (the inventor of GDP) in his 1955 paper Economic Growth and Income Inequality (my apologies, but you will need journal access to read this) based on data from England, Germany and the United States from 1875 onwards, and the belief that economic growth will automatically deal with inequality has been a powerful influence on economic policy at the World Bank and elsewhere since.

However, more recent data has shown the patterns suggested by this limited original data set are no longer correct, if indeed they ever were. Thomas Piketty and Emmanuel Saez, in their 2001 paper Income Inequality in the United States 1913-1998, state:

In particular, the evidence presented in this paper, together with the evidence on France by Piketty (2001a, 2001b) and the U.K. by Atkinson (2001),
strongly suggest that there was no such thing as a “spontaneous”, Kuznets-like decline of inequality in developed countries during the first half of the 20th
century. The inequality decline was to a large extent accidental (depression, inflation, wars) and amplified by political factors (progressive taxation). This does not mean that the current rise of inequality will not be followed by a mechanical downturn during the first few decades of the 21st century: this is simply saying that such a mechanical downturn apparently never occurred in the past.

Their data suggests a curve which looks like this instead:

The Environmental Kuznets Curve

This was first proposed by Gene Grossman and Alan Krueger in 1994 in their working paper Economic Growth and the Environment, which suggested that there was an eventual inverse relationship between pollution and income per capita, with a turning point mooted at around $8,000. Most of their graphs are not quite as U-shaped as the Kuznets Curve, but this nonetheless has come to be known as the Environmental Kuznets Curve.

However, in 2016, the international industrial ecology research community and United Nations Environment agreed on a comprehensive data set for global material extraction and trade covering 40 years of global economic activity and natural resource use, which led to several papers including the UNEP Global Material Flows and Resource Productivity: A Report of the International Resource Panel (again apologies but journal access needed). Their graph of material extraction instead looked like this:

The Human Development Index (HDI) is the geometric average of 3 indices: Gross National Income, Health and Education. An optimum score of 1 is achieved where life expectancy is 85 or more years, adult literacy is 100%, school enrolment is 100% and the Gross National Income is US$40 000 or more per person per year in purchasing power parity. So again, this is not very supportive of a reduction in material footprint with increased wealth.

Which brings us to the third graph, often cited as an argument for why one of the most obvious ways to reduce inequality rather than just focusing on average income per capita, ie make taxation more progressive, is pointless.

The Laffer Curve

The story of the Laffer Curve, dating from the 1970s, is recounted by Arthur Laffer himself here. It plots tax rates against tax revenues to indicate that there is a tax rate beyond which tax revenues actually reduce. As he says:

The Laffer Curve itself does not say whether a tax cut will raise or lower revenues. Revenue responses to a tax rate change will depend upon the tax system in place, the time period being considered, the ease of movement into underground activities, the level of tax rates already in place, the prevalence of legal and accounting-driven tax loopholes, and the proclivities of the productive factors. If the existing tax rate is too high…then a tax-rate cut would result in increased tax revenues. The economic effect of the tax cut would outweigh the arithmetic effect of the tax cut.

However, returning to Piketty, this time in the 2011 paper,  Optimal Taxation of Top Labor Incomes: A Tale of Three Elasticities by Piketty, Saez and Stefanie Stantcheva, the evidence underpinning this curve is again highly questionable. As they point out in the abstract (bold type added by me):

This paper presents a model of optimal labor income taxation where top incomes respond to marginal tax rates through three channels: (1) standard labor supply, (2) tax avoidance, (3) compensation bargaining…The macro-evidence from 18 OECD countries shows that there is a strong negative correlation between top tax rates and top 1% income shares since 1960, implying that the overall elasticity is large. However, top income share increases have not translated into higher economic growth. US CEO pay evidence shows that pay for luck is quantitatively more important when top tax rates are low. International CEO pay evidence shows that CEO pay is strongly negatively correlated with top tax rates even controlling for firm characteristics and performance, and this correlation is stronger in firms with poor governance. All those results suggest that bargaining effects play a role in the link between top incomes and top tax rates implying that optimal top tax rates could be higher than commonly assumed.

There are a number of charts which could be used from this paper, but I have chosen the plot of economic growth against changes in top marginal tax rate to illustrate most clearly the problems with the Laffer Curve idea:

This graph should show an inverse relationship if the Laffer Curve were true.

Why do I feel the need to debunk these simple so-called economic laws which are nothing of the sort? Because you will always prioritise economic growth over everything else if you believe that:

  • Growth will fix inequality;
  • Growth will fix pollution;
  • Trying to fix inequality through the tax system is counter-productive.

And these beliefs will then also have policy implications when faced with a different sort of curve.

This was an explainer from Grant Sanderson at 3Blue1Brown about COVID-19 from March 2020 setting out quite simply how it was likely to spread, and how different case numbers in different countries (eg between Italy and the UK) were as likely to be due to being at different time points since the start of the pandemic as reflecting the relative success of their containment policies. We now know the UK Government locked down too late, at least partly because they prioritised economic growth over containment policies in the first few weeks:

Those attitudes changed and we have had an incredibly successful vaccine rollout in the UK, but this has been at the expense of any idea of international cooperation in vaccine supply. Wealthy countries such as the UK have bought enough vaccinations to cover our populations almost three times over, while Covax, the global vaccine procurement scheme, only aims to vaccinate 20% of the populations of recipient countries this year.

This is very short-sighted if we think there might be an international issue even more threatening to life than COVID-19 which can only be combatted by unprecedented levels of international cooperation. And of course this is exactly what we have in the form of the climate emergency and our final graph (from the National Oceanic and Atmospheric Administration (NOAA) in the US showing the relentless rise in the level of carbon dioxide in the atmosphere as global emissions continue to increase:

 

Living in Hope-over-Experience may be very comfortable for some people for a limited time, but if it stops us engaging with the more implacable curves of the world we actually live in then none of us will be safe.

Source: Wikimedia Commons: Shattered right-hand side mirror on a 5-series BMW in Durham, North Carolina by Ildar Sagdejev. Cropped by Nick Foster

It starts in 2025 with a description of a horrific heatwave in India which will stay with me for a very long time. As well it should as, in the book, it kills 20 million people. In response, India send thousands of aircraft up to 60,000 feet to spray aerosol particulates of sulphur dioxide into the stratosphere, in defiance of the international conventions banning such activities, to deflect some of the solar radiation with the aim of reducing the probability of future heatwaves for a period. By how much or for how long or with what other consequences is unknown.

As we build up to COP26 in Glasgow in November this year, in the book we start with the results of COP29 in Bogota, where the organisation which would come to be known as The Ministry for the Future (and the title of the book by Kim Stanley Robinson) was set up “to advocate for the world’s future generations of citizens, whose rights, as defined in the Universal Declaration of Human Rights, are as valid as our own. This new Subsidiary Body is furthermore charged with defending all living creatures present and future who cannot speak for themselves, by promoting their legal standing and physical protection.”

The Indian crisis happens a few months later. The new head of this body, Mary Murphy, is briefly held captive by, Frank, one of the survivors of the heatwave in her own flat in Zurich (the book also feels like a love letter to Zurich) and challenged to do more:

It’s not enough. Your efforts aren’t slowing the damage fast enough. They aren’t creating fixes fast enough. You can see that, because everyone can see it. Things don’t change, we’re still on track for a mass extinction event, we’re in the extinctions already. That’s what I mean by not enough. So why don’t you do something more?

This has a profound impact on Mary, who keeps in touch with Frank and his troubled suffering life throughout the book. It also leans her towards effectively endorsing the involvement of her No 2 in “black” operations to ensure certain people are “scared away from burning carbon”.

Indeed the book is suffused with eco-terrorism. Technological progress has partly displaced the state monopoly of violence, with drone technology in particular meaning that no aircraft or ship or surface navy is safe from a well-enough organised group by the end of the book. People stop flying when aircraft start being shot down regularly, and those that still do fly use carbon-negative airships, where solar panels generate more power than the ships use. Davos attendees get taken hostage and given a compulsory seminar at one point. Tax havens become obsolete when all money becomes digital and tracked.

Mary’s interactions with central bankers are probably the closest this book ever comes to comedy. In the first, she tries to argue for a “carbon coin”, a digital currency which would be paid out to organisations and people who could prove they had removed carbon from the environment. This would be the incentive to work alongside the carbon taxes. The contemptuous response from the Federal Reserve and others at first is “not our purview”, however by the end they are on board with this and many of the other ideas developed along the way.

There are so many ideas in this book, far too many to cover them all here: some of them familiar to me from economics (carbon quantitative easing, Jevons’ Paradox, Modern Monetary Theory, Gini Coefficient – these each get a short chapter among many other ideas and interspersed with riddles) and others not so. The Indian techno fix is the first of many: some successful, like sucking out the meltwater under glaciers to slow them sliding into the ocean and others not so, like the billionaire wanting to refreeze the oceans. Russia dyes parts of the Arctic yellow to reflect more sunlight back. Huge areas of land are rewilded.

What strikes me most is that the arguments we tend to have here and now about which course to take (Freud’s phrase is quoted here in the book – “the narcissism of small differences”) seem largely moot in this one imagined near-future: all of them are tried there – it’s not techno-fixes or de-carbonisation of transport and heating, it’s both. It’s not carbon QE or re-wilding, it’s both. If something doesn’t work, it’s abandoned. By far the most important determinant of which of the IPCC future scenarios we end up on seems to be how quickly we start. Economists come in for particular ridicule there – whatever course of action is planned, they can find one group who thinks it will have one effect, one who think it will have the opposite effect and one which thinks it will make no difference at all. The difference is that the economists are no longer guiding policy there, but facilitating and post hoc rationalising it.

There is a wartime feel to the book throughout, with people doing what they feel needs to be done in desperate circumstances. The choices are all different levels of bad, but bad is almost incalculably better than worst. And the overall impression is of a world changing rapidly, with one of its herd animals belatedly getting into better balance with the others. Even at 560 odd pages the impressions are inevitably just that – one chapter is just a list of different organisations working on aspects of the climate emergency in different countries, described as about 1% of the total number active. It is like the shards of a smashed wing mirror picking out details from the vanishing world behind. I have never wanted to apply the word polymesmeric (which I first saw on the cover of Catch 22 by Joseph Heller) to a book as much as I have to this one.

The hoped-for outcome of all of this? In one conversation this is described as a “success made of failures” or a “cobbling-together from less-than-satisfactory parts”, which I think sums it up nicely.

And I definitely want to visit Zurich one day. Probably by airship.

 

There are many papers about model risk, and the dangers of blindly relying on algorithms or metrics without allowing for human judgement at any point in any subsequent analysis (in effect “baking in” whatever analysis was done at the time the computer model or algorithm was constructed as the final word), but these can often descend into the same level of technical impenetrability as the programmes they are attempting to critique.

I watched the film Sully: Miracle on the Hudson for the first time this week, on the anniversary of the landing on the Hudson. In the final scenes there is a hearing (spoiler alert!), where the evidence presented up until that point based on computer simulations, with and without pilots involved, was leading to the unanimous conclusion that Sully and Skiles could have turned back to La Guardia or Teterboro airports rather than landing on the Hudson River in January. However Sully had appealed to have the video recordings of the pilot simulations shown to the hearing, and these revealed the pilots responding to the catastrophic bird strikes which had taken out both engines (again something later confirmed when the actual engines were recovered, but which the simulations themselves did not accept because of the instrument readings on one of the engines from the aircraft) by calmly immediately setting course for La Guardia or Teterboro with no decision or response or recovery time needed at all. When a 35 second allowance for this was inserted into the simulations, the results were fatal crashes in both cases.

What struck me was how invisible this deficiency in the programming of the simulation would have been without a cockpit recording of the simulations. In many of the programmes we use to automate judgement-heavy processes, such as recruitment, many of the capital allocation decisions in financial institutions or even A-level grades, we do not have anything equivalent to a cockpit recording available to us. Perhaps we wait until either events prove us wrong (bad) or those on the receiving end of our automated decisions start to complain in sufficient numbers for us to reconsider (worse). What if quite a large proportion of the cost savings from automating these processes is in fact illusory as a result of our not putting enough time and attention into the original programming and/or not setting aside enough budget for maintaining it and challenging its decisions with parallel processes which do allow for human judgement? How much bigger is this problem going to become in the era of machine learning, where the programmes we are running are themselves several steps of abstraction away from those originally written by humans?

Our ability to programme machines to carry out billions of calculations in seconds would have been regarded as miraculous only a few decades ago and is still pretty astonishing to us now. We need to start thinking a lot more about how we can live alongside these ever more capable machines amicably over the long term. And it can’t be only programmers who get to see what the machines are doing – whatever the technical problems of allowing the equivalent of a cockpit recording to be made which can be understood by any of us, they need to be solved with as much urgency as the process automation itself. All of our decision-making processes need to be understandable and challengeable by the society in whose name they are carried out. It’s time to get serious now about our miracles.

“We won’t go back to normal, because ‘the normal’ was the problem.”

For me the turning point came on 12 March, when the FTSE 100 fell by 639 points or around 11% of its value in one day. What were the newspaper headlines that day?

Only the Times and the Financial Times had the stock market fall on their front page at all. Everyone else led with some variant on the Prime Minister saying that many families would lose loved ones. The attention switch was so complete that when KPMG published their UK Economic Outlook for March 2020 the following week – forecasting a main scenario for Gross Domestic Product (GDP) in the UK to fall by 2.6% in 2020 then grow by 1.7% in 2021, and a downside scenario for GDP to contract by 5.4% in 2020 and by another 1.4% in 2021, representing a slightly more severe recession than the downturn experienced in 2008-09 – nobody noticed that either (19 March and 20 March headlines here and here respectively), sandwiched as it was between the announcement that schools were to close and the Prime Minister saying that we had 12 weeks to turn the tide.

KPMG’s report was an example of damage function modelling of course: trying to model changes in economic activity due to some phenomenon and summarising that change in terms of a change in GDP. I have recently been quite exercised by similar considerations with regard to climate change damage functions and the inconsistencies of the ones in most current use with climate science. However it has become increasingly clear to me that I may have been missing the point. I realise I was focusing on damage functions because I felt they were leading to extreme optimism in the modelling of the impact of climate change on our economies and that it was this link which was most likely to get the attention of policymakers (and other actuaries!).

But of course GDP is only ever a proxy for some of the things we regard as important, rather than something that is important in itself, and a flawed one too. As Diane Coyle’s excellent book, GDP: A Brief But Affectionate History, makes clear. Its problems include:

  • It under-records growth by failing to capture fully the increase in the range of products in the economy;
  • It becomes a worse measure as the world economy consists less and less of material items, eg online activities; and
  • It can show positive growth caused by clearly unsustainable practices and those which deplete natural resources.

When KMPG released their economic outlook, it was as if they were trying to drag a weary world population away from the windows and balconies from which they are still trying to connect with each other and what is still real in the world back to the Monopoly game that they have set up in the front room.

It took a lot to get our behaviour to follow this change in attention. When Wuhan went into lockdown on 23 January, I was talking to Stuart McDonald, now a member of the COVID-19 Actuaries Response Group, about the talk he was planning to do at the University of Leicester on 18 March and deciding he would probably need to add a few slides about coronavirus. Italy went into lockdown on 9 March and yet on 12 March we had a second call where we still felt on balance that it might go ahead as long as we took sensible precautions, but by this time it was almost entirely about getting accurate messaging out about COVID-19. We called it off the following day. The UK finally went into lockdown on 23 March.

So perhaps it is no wonder that we have so far been unable to change human behaviour to anything like the same extent in response to climate change, which is a bit like COVID-19 in slow motion, progressing unseen with each stage of its development effectively locking us into the next steps in its relentless escalation. In the same way that movement restrictions may not slow down the increase in new cases for perhaps around a week, stopping carbon emissions now would still see us locked into further warming for 40 years. And even with the greater immediacy of coronavirus, it has only been when we have decided we care more about saving each other than maintaining our GDP that real progress has become possible.

My view is that some things that must be different post COVID are already clear. I think as a society we are going to demand more resilience, for example:

  • Resilience of our health service – this means much higher levels of spending, building deliberate over-capacity into the system in normal times;
  • Resilience of our food supplies, for example strengthening domestic supply chains;
  • Resilience of our population, so that we do not have 1.6 million food parcels needing to be given out in a year by the Trussell Trust, in the absence of a pandemic, for instance; and
  • Resilience of our infrastructure – to everything from floods to banking crises to pandemics to storms and heatwaves.

The Institute and Faculty of Actuaries (IFoA) has therefore shown great timing in its launch of its 2020 thought leadership campaign The Great Risk Transfer. The campaign aims to examine the trend of the transfer of risk from institutions to individuals, and how people can be better equipped to manage the financial risks they now face. I think the campaign rightly highlights the fact that risk transfer is all one way, but it clearly also goes way beyond the finance sector. Rail franchises never took on any real risk, it appears, even before the pandemic. Nor have PFI contracts, despite the price tag. By contrast the incremental removal of risk pooling by corporations for their employees and/or government for their citizens over the last 40 years has been relentless and in one direction only.

As Andrew Simms, one of the Green New Deal Group, said on Twitter yesterday about taking lessons for the climate emergency from the pandemic crisis:

Those roads with a fraction of the traffic, the drop in aviation, the economic shift to put public health & well-being first, the speed with which the brain adapts to the new normal: as someone said, these things are a postcard from the future we need to get to. Let’s take notes.

The War Room with the Big Board from Stanley Kubrick’s 1964 film, ”Dr. Strangelove”. Source: ”Dr. Strangelove” trailer from 40th Anniversary Special Edition DVD, 2004 Directed by Stanley Kubrick

In 1960, Herman Kahn, a military strategist at the RAND Corporation, an influential think tank which continues to this day, wrote a book called On Thermonuclear War. It focused on the strategy of nuclear war and its effect on the international balance of power. Kahn introduced the Doomsday Machine (which Kubrick used in his film “Dr Strangelove” alongside many other references from the book) as a rhetorical device to show the limits of John von Neumann’s strategy of mutual assured destruction or MAD. It was particularly noteworthy for its views on how a country could “win” a nuclear war.

For some reason Kahn came to mind as I was looking through Resource and Environment Issues: A Practical Guide for Pensions Actuaries, from the Institute and Faculty of Actuaries’ Relevance of Resource and Environment Issues to Pension Actuaries working party, which summarises the latest thinking on the climate change-related issues scheme actuaries should be taking into consideration in their work. I will come back to why.

The section which particularly caught my attention was called How might pensions actuaries reflect R&E issues in financial assumptions? This section introduces two studies in particular. First, we have the University of Cambridge Sustainability Leadership (CISL) report on Unhedgeable risk: How climate change sentiment impacts investment. This posits three “sentiment” scenarios (paraphrased slightly for brevity – see the report for details of the models used):

  • Two degrees. This is defined as being similar to RCP2.6 and SSP1 from the Intergovernmental Panel on Climate Change (IPCC) AR5. Resource intensity and dependence on fossil fuels are markedly reduced. There is rapid technological development, reduction of inequality both globally and within countries, and a high level of awareness regarding environmental degradation. It is believed that under this scenario global warming will not raise the average temperature by more than 2°C above pre-industrial temperatures.
  • Baseline. This is a world where past trends continue (i.e. the business-as-usual scenario), and there is no significant change in the willingness of governments to step up actions on climate change. However, the worst fears of climate change are also not expected to materialise and temperatures in 2100 are only expected to reach between 2°C and 2.5°C. This scenario is most similar to the IPCC’s RCP6.0 and SSP2. The economy slowly decreases its dependence on fossil fuel.
  • No Mitigation. In this scenario, the world is oriented towards economic growth without any special consideration for environmental challenges. This is most similar to the IPCC’s RCP8.0 and SSP5. In the absence of climate policy, the preference for rapid conventional development leads to higher energy demand dominated by fossil fuels, resulting in high greenhouse gas emissions. Investments in alternative renewable energy technologies are low but economic development is relatively rapid.

The modelled long-term performance for a range of typical investment portfolios is worrying:

CISL suggest quite different investor behaviour depending upon which climate change path they think the world is taking: moving into High Fixed Income if No Mitigation seems to be the direction we are heading, but adopting an Aggressive (ie 60% equities, 5% commodities) asset allocation if the Two Degrees scenario looks most likely.

Elsewhere the report suggests hedging via cross-industry diversification and investment in sectors with low climate risk. For example under No Mitigation, it is possible to cut the maximal loss potential by up to 47% by shifting from Real Estate (in developed markets) and Energy/ Oil & Gas (in emerging markets) towards Transport (in developed markets) and Health Care/ Pharma (in emerging markets). However over 50% of losses in all scenarios remain unhedgeable (ie unavoidable through clever asset allocation alone).

The second report (Investing in a time of climate change) from Mercer in 2015, focuses on the following investor questions:
• How big a risk/return impact could climate change have on a portfolio, and when might that happen?
• What are the key downside risks and upside opportunities, and how do we manage these considerations to fit within the current investment process?
• What plan of action can ensure an investor is best positioned for resilience to climate change?

The section I was drawn to here (it’s a long report) was Appendix 1 on climate models used, and particularly those estimating the physical damages and mitigation costs associated with climate change. The three most prominent models used for this are the FUND, DICE and PAGE models, apparently, and Mercer have opted for FUND. They have then produced some charts showing the difference between the damages exepcted for different levels of warming predicted by the FUND model compared to DICE:

The result of this comparison, showing lower damage estimates by the FUND model, led the modellers to “scale up” certain aspects of the output of their model to achieve greater consistency.

Both of these reports have been produced using complex models and a huge amount of data, carefully calibrated against the IPCC reports where appropriate and with full disclosure about the limitations of their work, and I am sure they will be of great help to pension scheme actuaries (although there does some to be some debate about this). However I do wonder whether as a profession we should be spending less time trying to find technical solutions in response to worse and worse options, and more time trying to head off the realisation of those sub-optimal scenarios in the first place. I also wonder whether the implicit underlying assumption about functioning financial markets and pension scheme funding is a meaningful problem to be grappled with at 3-4° above pre-industrial averages as some of this analysis suggests.

In the summary of Mark Lynas’ excellent book Six Degrees: Our Future on A Hotter Planet, the three degree increase for which damages are being estimated is expected to lead to Africa […] split between the north which will see a recovery of rainfall and the south which becomes drier […] beyond human adaptation. Indian monsoon rains will fail. The Himalayan glaciers providing the waters of the Indus, Ganges and Brahmaputra, the Mekong, Yangtze and Yellow rivers [will decrease] by up to 90%. The Amazonian rain forest basin will dry out completely. In Brazil, Venezuela, Columbia, East Peru and Bolivia life will become increasingly difficult due to wild fires which will cause intense air pollution and searing heat. The smoke will blot out the sun. Drought will be permanent in the sub-tropics and Central America. Australia will become the world’s driest nation. In the US Gulf of Mexico high sea temperatures will drive 180+ mph winds. Houston will be vulnerable to flooding by 2045. Galveston will be inundated. Many plant species will become extinct as they will be unable to adapt to such a sudden change in climate.

The [IPCC] in its 2007 report concluded that all major planetary granaries will require adaptive measures at 2.5° temperature rise regardless of precipitation rates.[and] food prices [will] soar. Population transfers will be bigger than anything ever seen in the history of mankind. [The feedback effects from the] Amazon rain forests dry[ing] out and wild fires develop[ing] [will lead] to those fires [releasing] more CO2, global warming [intensifying] as a result, vegetation and soil begin[ning] to release CO2 rather than absorb[ing] it, all of which could push the 3° scenario to a 4°-5.5° [one].

The last time the world experienced a three degree temperature rise was during the geological Pliocene Age (3 million years ago). The historical period of the earth’s history was undoubtedly due to high CO2 levels (about 360 – 440ppm – almost exactly current levels). I would suggest that our biggest problem under these conditions is not that over 50% of losses on pension scheme investments remain unhedgeable.

In his recent article for Social Europe, the unbearable unrealism of the present, Paul Mason presents two graphs. The first is the projection by the United States’ Congressional Budget Office of the ratio of debt to gross domestic product until 2048 in the United States.

The second is a chart from the IPCC showing how dramatically we need to cut CO2 emissions to avoid catastrophic and uncontrollable breakdown.

Mason feels that capitalism is too indebted to go on as normal and too structurally addicted to carbon. In his view Those who are owed the debt, and those who own rights to burn the carbon, are going to go bankrupt or the world’s climate will collapse. This feeling is echoed by George Monbiot here, where he cites a paper by Hickel and Kallis casting doubt on the assumption that absolute decoupling of GDP growth from resource use and carbon emissions is feasible and summarises some alternative approaches to the capitalism he feels no longer has the solutions.

Others dispute this, claiming that the Green New Deal is the only chance we have (here, here and here) to prevent irreversible climate change.

Whether you agree with any of these predictions or none of them, agree that we face a climate emergency or feel that is too extreme a description, it all brings me back to Kahn and Dr Strangelove. We seem to have replaced the MAD of the cold war with the MAD of climate change, except that this time we do not even have two sides who can prevent it happening by threatening to unleash it on each other. It is just us.

What we really cannot afford to be doing, via ever more complex modelling and longer and longer reports, is giving the impression that the finance industry can somehow “win” against climate change rather than joining the efforts to avert it as far as possible.

I have seen two very different pictures of the future of professional life over the last year or so. The first, which I wrote about over a year ago, was presented in The Future of the Professions by Richard and Daniel Susskind, and has been much debated since within the actuarial profession for what the implications might be for the future. In summary, the Susskinds set out two possible futures for the professions. Either:
• They carry on much as they have since the mid 19th century, but with the use of technology to streamline and optimise the way they work;
• Increasingly capable machines will displace the work of current professionals.

Their research suggests that, while these two futures will exist in parallel for some time, in the long run the second future will dominate. Indeed Richard Susskind has gone further in setting out what that future might look like for the legal profession, where he sets out future strategies for surviving in a world of increasingly capable machines as:

  • providing more for less (ie charging less (in particular the end of time cost fees), alternative billing arrangements such as “value billing”, making efficiencies and collaboration strategies where clients come together to share costs);
  • liberalising services (ie allowing a wider range of people to provide legal services); and
  • technology (ie online services in all of their forms to make the delivery of these cheaper, increasing use of data scraping, text mining, etc to replace what was previously done through expert judgement).

So far, so expected. The relentless increase in technological capability is bound to demand increased efficiency and leaner organisations competing ruthlessly in a pitiless market, right?

Enter an alternative vision for the future. Pointing out that we have been here before and that Keynes had speculated in 1930 that

In quite a few years – in our own lifetimes I mean – we may be able to perform all the operations of agriculture, mining, and manufacture with a quarter of the human effort to which we have been accustomed.

David Graeber, in his latest book Bullshit Jobs, points out that this never happened, despite pretty much all of the technological developments and income increases which Keynes predicted. He suggests that this future which the Susskinds are predicting is already happening in terms of needing fewer people to fill the meaningful roles within organisations but that, rather than employing fewer people, we are either creating “bullshit” jobs which even the people doing them can see no point to or bullshitizing existing roles for which the meaningful need has passed. It is as if the organisations themselves have attempted to maintain the outward appearance of the same structures by disguising the hollowing out of so many of their functions with simulated business.

It is an intriguing alternative vision of how the professional world might develop which has come in for some criticism, the most serious of which Graeber attempts to address in his book. One of the reasons he thinks the situation has been allowed to develop is that noone believed that capitalism could produce such an outcome. But that is only if you accept the rational profit maximising principle, which many economists have now abandoned as an explanation for corporate or individual behaviour. Graeber gives one particularly important example of this in the creation of Obamacare, where Barack Obama “bucked the preferences of the electorate and insisted on maintaining a private, for-profit health insurance system in America”, quoting him as follows:

I don’t think in ideological terms. I never have,” Obama said, continuing on the health care theme. “Everybody who supports single-payer health care says, ‘Look at all this money we would be saving from insurance and paperwork.’ That represents one million, two million, three million jobs [filled by] people who are working at Blue Cross, Blue Shield or Kaiser or other places. What are we doing with them? Where are we employing them?”

So which vision of the future is more likely? I think, at the moment, there is probably more evidence for the Susskind vision, mainly because he has been working in this area for 30 years and therefore many of his predictions, such as the use of email to provide legal advice, have had time to emerge. Many of the stories in Graeber’s book ring true for me and are similar to experiences I have had at times myself, but he has only obtained 300 of them. The YouGov poll which highlighted that 37% of working adults say their job is making no meaningful contribution to the world – but most of them aren’t looking for another one, was based on a sample size of 849. There was also a similar result (in this case 40%) from a survey in the Netherlands, for which I couldn’t easily find the sample size. However this does also lend some weight to one of Graeber’s other contentions in the book that the financial industry might be considered a paradigm for bullshit job creation, as the following graph (from a working paper on this issue by Stolbova et al) shows that the Netherlands and the UK are by far the most financialised economies in the EU.

There are other parts which ring less true for me. For instance, I do not recognise the alternative “non-managerial” university exam paper production process to that shown below (which is just the academic staff sending the exam to a teaching assistant to print and the teaching assistant confirming that he/she has done so) as ever having been remotely acceptable, but this may just reflect the fact that I have been working in academia for a far shorter time than Graeber. However there is no doubt that this is an interesting and useful field of enquiry and potentially concerning for all of us trying to support our graduates in negotiating a meaningful and rewarding entry into the workplace.

There is likely to be significant disruption over the next couple of decades in how we do things and it seems likely to me that there will be many seeking to protect familiar organisational and power structures along the way, as our assumptions about what we want and how we are prepared to have it provided to us are seriously challenged in sometimes unnerving ways. Of the sustainability of these protections ultimately, I am less sure.

The FTSE All-Share Index, originally known as the FTSE Actuaries All Share Index, along with the FTSE 100, represent nearly all of the market capitalisation and the top 100 companies by size listed on the London Stock Exchange respectively. They are mentioned in all BBC news bulletins. When they go up, we all feel better. When they go down, they are seen as portents of doom.

Let me show you a different actuaries’ index instead:

Figure 1 shows the ACI and each of the components. The composite ACI represents the average of the six components (with sign of change in cool/cold temperatures reversed). The ACI is increased by reduction in cold extremes, consistent with increased melting of permafrost and increased propagation of diseases, pests, and insects previously less likely to survive in lower temperatures. A positive value in the ACI represents an increase in climate-related extremes relative to the reference period.

The threat of climate change is real, independent of speculative trading and the news media cycle, and increasing with each degree of warming we are unable to stop. Alongside this are the increasing risks of extreme weather events, which is most neatly described for North America currently by the Actuaries Climate Index. This focuses on six components in particular which have the most impact on human societies:

  1. Frequency of temperatures above the 90th percentile (T90);
  2. Frequency of temperatures below the 10th percentile (T10);
  3. Maximum rainfall per month in five consecutive days (P);
  4. Annual maximum consecutive dry days (D);
  5. Frequency of wind speed above the 90th percentile (W); and
  6. Sea level changes (S).

It then tracks them all over time, as shown in the graph above.

It seems clear to me that we should be reacting much less to the booms and busts of economic cycles and much more to climate-related threats. This is for two main reasons:

1. More people are at threat of death or injury as a result of climate change than even the 2008 crash in our financial systems. The World Health Organisation (WHO) predicts that, between 2030 and 2050, climate change is expected to cause approximately 250,000 additional deaths per year, from malnutrition, malaria, diarrhoea and heat stress. However, the additional deaths are already here. Taking just two examples from the WHO:

  • In the heat wave of summer 2003 in Europe for example, more than 70,000 excess deaths were recorded, with the frequency of such events steadily increasing.
  • Globally, the number of reported weather-related natural disasters has more than tripled since the 1960s. Every year, these disasters result in over 60,000 deaths, mainly in developing countries, which means that 40,000 of those deaths pa can already be directly attributed to climate change.

On the other hand, the 500,000 additional cancer deaths and 10,000 additional suicide deaths since 2008 cannot be attributed directly to the 2008 crash, as the analysis shows. These are more a result of the austerity policies which have been applied since 2008. Unnecessarily.

Climate change on the other hand does not care whether we react to it or not. It will relentlessly change the chemistry and biology of everything around us as the Earth and the inhabitants of the Earth adapt. We may survive it, in reduced numbers, or we may not. The Earth does not care. Responding to the threat will not make more climate-related events happen unless our response is to, by and large, ignore it.

2. One depends on the other. We cannot base our economies on a FTSE-led GDP-growth-at-all-costs model because it is not physically possible to maintain it without losing the environment from which our growth originates. As Finbarr Livesey points out in his excellent From Global to Local, the circular economy which the overwhelming consensus of studies show would increase employment and contribute to economic growth is taking a long time to arrive. In Europe, where we consume around 16 tonnes of stuff each per year, figures from Siemens in 2016 suggest that 95% of it and its energy value is lost through the life cycle of the products themselves.  As Kate Raworth  and others have pointed out, we need to focus on different measures of success if we are going to direct our economies in a more sustainable, less volatile and doom-laden direction.

There are plans to extend the Actuaries Climate Index to Europe (including the UK in this instance!), with a recent feasibility study concluding “that the prospects for constructing an analogue to the Canada-US ACI over the European region are promising”. I hope we see such an index soon, because, as Randall Munroe illustrates here, we have not been here before.

I look forward to the day when a new global actuaries’ climate index is on every news bulletin, making us feel better when it goes down and seeing any rise as a portent of doom. Because this time it really would be.

Great infographic from futurism.com summarising the likely impacts of each additional degree in warming which I thought was definitely worth sharing!

 

Global Warming Scenarios

https://imgs.xkcd.com/comics/the_three_laws_of_robotics.png
This work is licensed under a Creative Commons Attribution-NonCommercial 2.5 License.

Daniel and Richard Susskind in their book “The Future of the Professions” set out two possible futures for the professions. Either:
• They carry on much as they have since the mid 19th century, but with the use of technology to streamline and optimise the way they work
• Increasingly capable machines will displace the work of current professionals

Their research suggests that, while these two futures will exist in parallel for some time, in the long run the second future will dominate. The actuarial profession is particularly vulnerable. As the Susskinds write:

Accountants and consultants, for example, are particularly effective at encroaching on the business of lawyers and actuaries.

Actuaries both here and in other countries are waking up to what is coming, but the response of the profession is a whole has been quite slow.

For the actuarial profession, we will see the extension of some trends which have already begun, eg:

  • Automation of processes not just leading to greater efficiencies but reconfiguring both what work is done and how it is done, eg propensity pricing and pensions valuations
  • Para professionalization, like CAA Global for instance
  • Globalisation
  • Specialisation
  • Mergers of businesses as markets consolidate
  • Flexible self employment

And the emergence of trends that have hardly started at all yet, eg:

  • The end of reserved roles for actuaries
  • Different ways of communicating advice (Richard Susskind got into trouble with the Law Society in the mid 1990s for suggesting that most legal communication between lawyers and their clients would be delivered via email in the future, which would strike us as an obvious observation now)
  • Online self-help for users of actuarial advice (ask discussed by the Pensions Policy Institute in their report last year)
  • The advance of roboactuaries and their assistants

Focusing on the last of these, a paper produced by Dodzi Attimu and Bryon Robidoux for the Society of Actuaries in July 2016 explored the theme of robo actuaries, by which they meant software that can perform the role of an actuary. They went on to elaborate as follows:

Though many actuaries would agree certain tasks can and should be automated, we are talking about more than that here. We mean a software system that can more or less autonomously perform the following activities: develop products, set assumptions, build models based on product and general risk specifications, develop and recommend investment and hedging strategies, generate memos to senior management, etc.

They then went on to define a robo actuarial analyst as:

A system that has limited cognitive abilities but can undertake specialized activities, e.g. perform the heavy lifting in model building (once the specification/configuration is created), perform portfolio optimization, generate reports including narratives (e.g. memos) based on data analysis, etc. When it comes to introducing AI to the actuarial profession, we believe the robo actuarial analyst would constitute the first wave and the robo actuary the second wave

They estimate that the first wave is 5 to 10 years away and the second 15 to 20 years away. We have been warned.

One of the implications of this would be far fewer actuarial students required and, in my view, a much smaller appetite amongst actuarial firms for employing students while they were sitting actuarial examinations, particularly the core rather than specialist ones. This in turn would suggest an expansion of the role of universities in supporting students through these stages of their actuarial education, massively increasing the IT and data analysis skills of the next generation of actuarial students and developing far more opportunities for students to develop skills more traditionally seen as “work-based”, such as presentation, project management and negotiation skills. Some universities, such as my own at the University of Leicester, are using the preparatory work in anticipation of the Institute and Faculty of Actuaries’ launch of Curriculum 2019 to do all of these things.

But universities and the education professionals in general face their own challenges from the rise of technology and increasingly capable machines:

  • The development of learning labs offering personalised learning systems
  • Online education networks, like Moodle, once used just to support traditional university teaching activities, but now starting to actively supplant them
  • Other online education platforms, like the Khan Academy
  • The rise of Massive Open Online Courses or MOOCs. For instance, more people have signed up to Harvard University’s MOOCs in one year than have enrolled at the University in its 377 year history

The actuarial profession and the higher education sector therefore need each other. We need to develop actuaries of the future coming into your firms to have:

  • great team working skills
  • highly developed presentation skills, both in writing and in speech
  • strong IT skills
  • clarity about why they are there and the desire to use their skills to solve problems

All within a system which is possible to regulate in a meaningful way. Developing such people for the actuarial profession will need to be a priority in the next few years.

Of course it is still possible to laugh at what Artificial Intelligence and Machine Learning (here and here) have not managed to do yet, despite their vast ambitions. But it should not blind us to the fact that those ambitions will be realised in our working lifetimes in many cases. And we need to start preparing now.